Operational Aircraft

Performance and
Flight Test Practices
By Mario Asselin

Joseph A. Schetz
Editor-In-Chief
Operational Aircraft Performance
and Flight Test Practices
Operational Aircraft
Performance and Flight
Test Practices

Mario Asselin

AIAA EDUCATION SERIES

Joseph A. Schetz, Editor-in-Chief
Virginia Polytechnic Institute and State University
Blacksburg, Virginia

American Institute of Aeronautics and Astronautics, Inc.

American Institute of Aeronautics and Astronautics, Inc., Reston, Virginia

1 2 3 4 5

Library of Congress Cataloging-in-Publication Data

On file.
ISBN: 978-1-62410-592-0

Copyright # 2021 American Institute of Aeronautics and Astronautics, Inc. Printed in the
United States of America. No part of this publication may be reproduced, distributed, or
transmitted, in any form or by any means, or stored in a data-base or retrieval system,
without prior written permission.

Data and information appearing in this book are for informational purposes only. AIAA is
not responsible for any injury or damage resulting from use or reliance, nor does AIAA
warrant that use or reliance will be free from privately owned rights.
 AIAA EDUCATION SERIES

Editor-in-Chief
Joseph A. Schetz
Virginia Polytechnic Institute and State University

Editorial Board

João Luiz F. Azevedo Michael Mohaghegh

Comando-Geral De Tecnologia The Boeing Company
Aeroespacial
Conrad F. Newberry
Marty Bradley Naval Postgraduate School;
electra.aero California State Polytechnic
University, Pomona
James R. DeBonis
NASA Glenn Research Center Brett Newman
Old Dominion University
Kajal K. Gupta
NASA Dryden Flight Research Center Hanspeter Schaub
University of Colorado
Rakesh K. Kapania
Virginia Polytechnic Institute and Humberto Silva III
State University Sandia National Laboratories

Brian Landrum David M. Van Wie

University of Alabama, Huntsville Johns Hopkins University
À mon épouse, Sylvie Saint-Georges
CONTENTS

Acknowledgement xvii
Preface xix

Chapter 1 Atmosphere 1
Atmospheric Properties 1
Perfect Gas 4
Atmosphere to 65,617 Feet 4
Standard Atmosphere 5
Atmospheric Ratios 11
Temperature, Pressure, and Density Altitudes 13
Exercises 15
References 16

Chapter 2 Measuring Altitude 17

Introduction 17
Pressure Altitude 18
Radar Altitude 35
GPS Altitude 37
Exercises 37
Reference 37

Chapter 3 Measuring Airspeed 39

Introduction 39
True Airspeed, Speed of Sound, and Mach Number 40
Equivalent Airspeed 52
Calibrated Airspeed 53
Ground and Vertical Speeds 60

ix
x Operational Aircraft Performance and Flight Test Practices

Instruments: Airspeed Indicators 61

Exercises 66
Reference 66

Chapter 4 Altimetry Error 67

Indicated Airspeed/Altitude 67
Certification Regulations 68
Flight Testing for Air Data Errors 72
Probe Location: Manufacturing and Maneuver-Based Errors 114
Static Source Error Correction (SSEC) 121
Sensor Accuracy, Precision, Resolution, and Calibration 122
Flight Manual: Presenting the Information to the Flight Crew 125
Unreliable Airspeed (URA) 126
Concluding Remarks 128
Exercise 129
References 130

Chapter 5 Weight and Balance 131

Introduction 131
Weight Breakdown and Definitions 133
Aircraft Weighing: Theory and Practice 134
Center of Gravity: It’s a Balancing Act 137
Lateral CG 143
Vertical CG 145
Mass Moment of Inertia 146
Fuel Weight and CG 147
Passenger Weight 153
Cargo and Checked Baggage 155
Typical Weight and Center of Gravity Envelopes 156
Flight Testing and Flight Test Equipment and Ballast 156
Airplane Weighing Procedures and Documentation 160
Weight a Minute 164
Tracking the BEW In-Service 166
Impact of Weight on Performance: Examples 166
Exercises 169
References 169

Chapter 6 Lift and Stall 171

Basic Airplane Geometry 171
Lift and Lift Equation 176
 Contents xi

Lift Curve and Lift Distribution 177

Lift Limit: Aerodynamic Stall 186
Display of Angle of Attack in the Cockpit 192
Icing 193
Ice Protection Systems 204
Manufacturing Tolerance: An Aerodynamic Perspective 211
References 215

Chapter 7 Stall Testing 217

Importance of a Stall Program on Airplane Certification 217
Defining Stall from a Certification Point of View 218
Stall Speed Testing 250
Creating a Stall Speed Model 267
Presenting the Information to the Flight Crew 286
Exercise 287
References 288

Chapter 8 Thrust and Drag Modeling 289

Drag Model 289
Slow Flight Speed Domain 311
Thrust Model 313
Thrust Summary 349
Thrust-Drag Bookkeeping 350
References 353

Chapter 9 Flight Envelope 355

A Regulation-Based Limit 355
Low-Speed Limit 357
High-Speed Limit 386
Maximum Altitude 398
Temperature Envelope 400
Flaps Out and Gear Down Envelopes 401
Placard Speeds 404
Reduced Vertical Separation Minimum (RVSM) 404
Conclusion 410
References 410

Chapter 10 Cruise Performance and Flight Endurance 411

Defining Range 411
xii Operational Aircraft Performance and Flight Test Practices

Useful Parameters 412

Specific Air Range (SAR) 414
Range Equation: Turbojet/Turbofans 426
Range Equation: Reciprocating/Turboprop 437
Fuel Reserves 440
Impact of Wind on Range 444
Payload-Range Diagram 447
Flight Endurance 449
Electrical Propulsion: A Step Change? 451
Hybrid-Electric: An Evolutionary Change 461
The Case for the Reduced Energy Plane: The Hydrogen Plane 462
References 465

Chapter 11 Cruise and Drag Performance Testing 467

Creating a Drag Model from Flight Test Results 467
Testing for Low-Speed Drag 468
Testing for High-Speed Drag 487
Testing for SAR 493
Testing for Zero Fuel Level 499
Presenting the Information to the Flight Crew 502
Fuel (Energy) Management and Conservation 507

Chapter 12 Climb Performance 513

Developing a Climb Performance Model 513
Factors Influencing Climb Performance 527
Climb Schedules 538
Descending Flight 538
Load Factor and Flight Path Variation 539
Energy Height 544
Certification Requirements 547
Flight Test and Data Reduction 551
Presenting the Information to a Flight Crew 564
Exercise 570

Chapter 13 Climb Capability OEI 571

Impact of One Engine Inoperative (OEI) on Climb
Performance 571
Certification Regulations: Part 25 Transport
Category Airplanes 578
Flight Test and Check Climbs 590
 Contents xiii

Presenting the Takeoff Weight Limited by Climb

Requirements to a Flight Crew 596
Obstacle Clearance 598
Presenting the Climb Information for Obstacle
Clearance to a Flight Crew 609
Gliding Flight 617
Exercise 620
Reference 621

Chapter 14 Turn Performance and Maneuver Margin 623

Building a Basic Model 623
Special Cases for Turn Performance 638
Altitude Effects on Sustainable Turn Performance
Flight Envelope 647
Energy Maneuverability 649
Inner Wing Stall in a Turn 654
Maneuver Margin and Certification Regulations 657
Buffet Envelope 663
Unrestricted Turn Performance 672
Turn Performance under OEI Conditions 674
Exercises 675
Reference 676

Chapter 15 Takeoff Performance 677

Takeoff Distance Definition 677
Basic Takeoff Performance Model: Takeoff Segments 679
Takeoff Speed Summary 719
Takeoff Distance, Acceleration-Go 720
Certification Regulations: Part 25 Transport Category
Airplanes 722
Exercises 739

Chapter 16 Takeoff Performance Testing 741

Takeoff Distance Modeling 741
Minimum Control Speed: Air VMCA 750
Certification Regulations: Robustness of Takeoff Procedure 796
Takeoff Performance Testing: Risks and Mitigations 807
Exercise 811
Reference 811
xiv Operational Aircraft Performance and Flight Test Practices

Chapter 17 Rejected Takeoff Performance 813

Why Reject a Takeoff? 813
RTO Distance Definition 816
Stopway 845
Runway Safety Area (RSA) 846
Engineered Materials Arresting Systems (EMAS)
for Airplane Overruns 846
References 848

Chapter 18 Rejected Takeoff Performance Testing and

AFM Takeoff Field Length 849
Braking Performance and Testing RTO 849
Testing for Minimum Control Speed, Ground VMCG 850
RTO Testing 860
Maximum Kinetic Energy Testing (Max KE) 868
Maximum Wheel Speed 872
Airplane Flight Manual: Takeoff Field Length (TOFL) 872
References 891

Chapter 19 Landing Performance 893

Basic Modeling 893
Air Segment 896
Transition Segment 924
Ground Run Segment 929
Total Landing Distance 929
Operational Requirements 930
Exercises 933
References 934

Chapter 20 Landing Performance Testing and AFM 935

Preparing for a Landing Distance Test Campaign 935
Testing for the Air Segment 936
Testing for the Transition Segment 959
Testing for the Ground Run Segment 964
Total Landing Distance 965
Presenting the Information to the Flight Crew 971
Time of Arrival Landing Performance 982
Steep Approach 986
 Contents xv

Transport Canada Steeper Requirements 997

Reference 997

About the Author 999

Index 1001
Supporting Materials 1015
 ACKNOWLEDGEMENT
I would like to thank my wife Sylvie, a truly amazing woman, for her
continued support over more than the 31 years spent in the ups and
downs of the aviation industry which have brought us to all corners of
the North American continent, and to Europe. Throughout this journey,
she has reinvented herself to support our ‘adventure’, being a Director of
a language school during a time, organizing Art exhibitions for various
museums, and even working as a schoolteacher while we were stationed
at a Canadian Forces Bases for a year in Ontario. She then transitioned to
the aviation world when I convinced her that we needed to try this small
company down in San Antonio producing the SJ30-2, a Part 23 single
pilot, Mach 0.83, 49000 ft cruise altitude business jets with a 12-psi cabin
and 2500nm range. She gladly took on the challenge, helping the
company progress through certification and initial deliveries, working in
Operations and Customer Support, as well as with the transition team for
the company’s ownership change. When she takes on a new project, she
not only brings her great organizational skills to the table, she also brings
a human touch to the work, something I have found is often missing in
most leadership position in aviation.
In her latest venture, she jumped into the world of Flight Test Team
leader at the Bombardier Flight Test Center (BFTC) where she has
managed the daily activities of multiple certification flight test programs
including the Global 6000, the Learjet 70/75, the Challenger 350, the
CSeries. What is truly remarkable is that some programs had Transport
Canada as the primary certification agency while others had the FAA and
she expertly navigated the different approach to each primary certifica-
tion agency. Most recently, she has managed the sustaining flight test
fleets (CS100 and CS300) of the CSeries (now the Airbus A220) for
Airbus (as a Bombardier employee), and the Global 7500 fleet for Bom-
bardier at BFTC as these new airplane types progress through their initial
years of operation and continuous product improvements. I would not be
surprised to see her progress to Director of Flight Test in the coming
years!

xvii
xviii Operational Aircraft Performance and Flight Test Practices

She has also encouraged me to grow our small company and take on
interesting projects like non-symmetrical flying test beds (FTB), an electric
regional airplane, tilt-wing aircraft, and a flying motorcycle.
 PREFACE
This textbook represents a retrospective of an interesting career in avia-
tion and airplane certification. It is meant to complement AIAA’s An Intro-
duction to Aircraft Performance. This book has been formatted to meet the
needs of graduate students and professional engineers working in the field
of airplane design, airplane performance, flight test and certification. This
author has met many great young engineers eager to learn, having great
skills out of school, and just needing the extra “this is why we do it. . .”.
As with any engineering disciplines, it is understood that there is more
than one way to get the answer to a given problem; this book represents
this author’s experience on approaching airplane performance.
The book closely follows the format of the course of the same title
offered at Kansas University Lifelong & Professional Education, Aerospace
Short Courses. The book was conceived following discussions with several
practicing professionals and the many students at the Kansas University
course wishing to have a single source reference from the basic theory to
practical cases for certification flight testing and operational performance
monitoring. Class discussions identified that there was a need to provide
more real-life examples compared to what is typically offered in traditional
textbooks. Those discussions also generated summary charts like Figures
14.27 and 15.30 that provide a quick summary of the certification regu-
lations impact on the design requirements.
To this end, the first part of every subject will follow a similar format as
did the book An Introduction to Aircraft Performance with the development
of the basic performance equations based on a given set of assumptions.
Following this, we decompose the maneuver being analyzed to validate
its components (create small flight test packages to validate models and
expected airplane behavior). Then, flight test considerations will be dis-
cussed such as required instrumentation, flight test risk and risk mitigation,
data scatter, data reduction and presentation of the performance infor-
mation to the flight crew. Several examples of flight test results have
been added to help the reader better understand what (s)he may be faced
with when collecting data to create performance models.

xix
xx Operational Aircraft Performance and Flight Test Practices

Performance Maneuver based

theory
Test plan/procedures Identify risk Program based
instrumentation Mitigate risk (cost and schedule)

Performance Adjust testing based

Minimizing data scatter
testing on aircraft characteristics

Showing airplane
Handling data scatter Data
performance compliance
Performance models reduction
to regulations

Showing performance
Apply regulatory limits Data
models compliance
to performance models expansion
to regulations

Be clear and concise If it is too complicated,

Presentation to
to minimize the flight crew will not
flight crew
reading errors use the performance data

There is a major focus on FAA 14 CFR Part 25 certification require-

ments has a basis for discussion in this textbook since the format of this
set of regulations is publicly available and clearly shows the basic items
that go into demonstrating that an airplane is safe to operate. The
general approach is also valid for certification under Part 23 or military spe-
cifications and we may, from time to time, bring in some of these additional
requirements to show variance into an airplane certified performance.
As with any regulations, they evolve over time to include the lessons
learned by the aviation industry; so the regulations listed in this textbook
may be out of date when the reader goes through the material, these are
the regulations that were in effect at the time of printing. The FAA main-
tains a very good link to historical regulations under their main web site
(www.faa.gov), including all previous regulations and the notice of rulemak-
ing with final ruling that shows why the regulation changed. Therefore, the
info presented is meant to capture a standard at a given time knowing that it
will most likely evolve as we learn more.
The book also includes many websites (links) that overtime may disap-
pear, but the info generally is only relocated, and the links provided may
provide a good starting point for a search on the subject.
Throughout this book, we will review accidents and incidents that have
occurred on various airplanes. It is not the purpose of this book to assign
blame or liability, but rather to share experiences and lessons learned to
hopefully help with the prevention of future accidents and incidents.
Those accidents and incidents have often led to changes in regulations to
continue improving aviation safety.
Chapter 1 Atmosphere

Chapter Objective
A good knowledge of the atmosphere is very important for the flight of atmos-
pheric vehicles. The flow of air over the surface of the airplane generates the
necessary lift to allow it to fly. The airplane’s engines need air to generate
thrust. The airplane is shaped in such a way as to maximize lift and to mini-
mize drag. Many airplane instruments also need the atmosphere to work (air-
speed, altitude, etc.).
This chapter will introduce the student to the basics of atmospheric
properties modeling.

Atmospheric Properties

T
he atmosphere is a mixture of gases, which can be considered for air-
plane performance analysis purposes as a single gas called air. This
assumption is valid for flows at speeds lower than about Mach 5;
above these speeds, dissociation of air particles begins. The air consists, by
weight, of 76% nitrogen, 23% oxygen, and 1% of other gases (such as neon,
freon, helium, water, etc.). By volume, the composition of air is 78% nitrogen,
21% oxygen, and 1% other gases. Although water is not a major element of
the atmosphere, it is the small amount of water in the air that creates all of
the weather (clouds, rain, storm, etc.). To facilitate our calculations, we will
consider the atmosphere as being dry air, but the presence of water (humid-
ity) will decrease the actual density of air for the same temperature and
pressure.

Air Pressure
Pressure is a force per unit area. Also, pressure is isotropic at a given point
(its value being the same in every direction). Static pressure (p or ps ) can be
viewed as the weight per unit area of a column of fluid above the point of
measurement. It is the potential energy of the fluid. In the case of a fluid in
motion, an additional pressure, called impact pressure qc , will arise due to
the velocity of the fluid. It is the kinetic energy of the fluid in movement.

1
2 Operational Aircraft Performance and Flight Test Practices

Total pressure (po or pT ) is the summation of the static and impact

pressures.
pT ¼ ps þ qc (1:1)
Dynamic pressure q is defined as one half the air density times the
airspeed squared.
1
q W rV 2 (1:2)
2
Most instruments measure the gauge pressure Dp, which is a relative
pressure in that the pressure indicated by the instrument is the difference
between the measured pressure and a reference pressure. Relative pressure
can be negative or positive, but static pressure is always positive (vacuum
or zero pressure being the lower limit).
Common units of pressure are Newtons per square meter (N/m2 ) or
Pascal (Pa), pounds per square inch (lb/in.2 or psi), pounds per square
foot (lb/ft2 or psf), inches of mercury (inHg), and atmosphere (atm). The
conversion factors among the various units are as follows:

1 atm ¼ 101,325 N=m2 ¼ 101,325 Pa

1 atm ¼ 2116:217 psf ¼ 14:69595 psi
1 atm ¼ 29:92126 in Hg
1 psi ¼ 144 psf
The units favored for this course for air pressure are psf.

Air Temperature
Temperature T is a measure of the average kinetic energy of the particles
in the gas. This means that the higher the temperature of the gas, the
higher the speeds of the molecules in it. Absolute temperature is always posi-
tive and is given in Kelvin (K) for SI units and in degrees Rankine (8R) for the
imperial units. Relative temperature can be positive or negative and is given
in degrees Celsius (8C) for SI units and degrees Fahrenheit (8F) in imperial
units.
The absolute scales start at zero, and the ratio of Kelvin to degrees
Rankine is 1.8.
1 K ¼ 1:88R
The melting temperature of ice under 1 atm is 08C (328F). This corre-
sponds to an absolute temperature of 273.15 K. In imperial units, this
temperature is 459.678R. The conversion factors for temperature are
½8F ¼ 32 þ 1:8½8C
 CHAPTER 1 Atmosphere 3

The temperature at a given altitude is called static air temperature (SAT)

or outside air temperature (OAT). If the air is in movement or the airplane is
moving through the air, there will be a temperature rise. This new tempera-
ture will be known as the total air temperature (TAT or To or TT ). This
temperature will be covered in more detail in Chapter 3, Measuring Airspeed.
The units favored for this course will be K and 8C.

Air Density
Density r is the mass of a substance per unit volume. Units are usually
kilograms per cubic meter (kg/m3 ), pounds mass per cubic foot (lbm/ft3 ),
or slug per cubic foot (slug/ft3 ). The conversion factors are

1 slug=ft3 ¼ 515:379 kg=m3

1 slug=ft3 ¼ 32:174 lbm=ft3
As a note, because the slug is not a commonly used unit, consider that

1 lbf ¼ 1 slug  1 ft=s2

This is similar to when one uses the SI system, where

1 N ¼ 1 kg  1 m=s2
Whereas, when using lbm, one needs to divide by a constant
1 lbm  32:174 ft=s2
1 lbf ¼
32:174 lbm ft=lbf s2
The units favored for this course for air density are slug/ft3 .

Air Viscosity
Absolute (or dynamic) viscosity m is a measure of the internal resistance
exhibited as one layer of a fluid is moved in relation to another layer. It is one
of the most important properties in fluid dynamics because all real fluids
have a nonzero value of viscosity. The units are kilograms per meter
second [kg/(m s)] or pounds per foot second [lb/(ft s)]. We can also define
the kinematic viscosity n as the ratio of the absolute viscosity to density of
the fluid. Its units are meters squared per second (m2 /s), feet squared per
second (ft2 /s), or the equivalent in other units.
m
n¼ (1:3)
r
The viscosity of the air varies as a function of the temperature. The
following equation represents the change in absolute viscosity m with
4 Operational Aircraft Performance and Flight Test Practices

temperature T (Kelvin), the constant derived from experimental data:

b T 3=2
m ¼ (1:4)
T þ S
where
m ¼ absolute viscosity at temperature T
b ¼ dynamic viscosity constant [1.458  1026 kg/(s m K1=2 )]
S ¼ Sutherland’s constant for air [110.4 K]

Perfect Gas
The air, under normal atmospheric conditions, is considered a perfect gas
(where the intermolecular forces are negligible). The relation among the
pressure, the density, and the temperature is then
p ¼ rRT (1:5)
where R is the specific gas constant. For air, R is
J ft lbf ft lbf
R ¼ 287:053 ¼ 1716:5 ¼ 53:35
kgK slug8R lbm 8R
Air can be considered a perfect gas at temperature up to 2500 K (23278C).
At temperatures higher than 2500 K, oxygen begins to dissociate.
Another important value for the study of atmospheric properties is the
ratio of specific heats g; more specifically, the ratio of the specific heats for
constant pressure cp and constant volume cy . Its value for air is
cp
g¼ ¼ 1:4
cy
Finally, Carnot’s law stipulates that
cp  cy ¼ R

Atmosphere to 65,617 Feet

The atmosphere is divided into many distinctive layers. Those of interest
are below 65,617 ft above the mean sea level (MSL), because most airplanes
have a normal operating ceiling lower than this value (with a few exceptions
such as the SR-71). The majority of the world’s airplanes operate at altitudes
below 51,000 ft.
The troposphere is the lowest layer. This is where most atmospheric
flights occur. It is characterized by a negative temperature gradient (or temp-
erature lapse rate). This means the temperature decreases as the altitude
increases (dT/dh , 0). The height of the troposphere will vary according
 CHAPTER 1 Atmosphere 5

to the geographic location, from approximately 28,000 ft at the poles to about

55,000 ft at the equator. It also varies between summer and winter.
The tropopause is the upper limit of the troposphere. At this
point, the temperature gradient gradually changes to zero (dT/dh ¼ 0).
The next layer is the stratosphere, where the temperature gradient remains
zero up to 65,617 ft (20 km). In this layer, the water vapor is almost
nonexistent.
Within these two layers, the gravitational acceleration is approximately
constant. We will cover the variation of gravitational acceleration with alti-
tude in the next section.

Standard Atmosphere
Because the density, the pressure, and the temperature of the air depend
on many variables, such as date, position, humidity, and so on, it is usual in
aerodynamics to postulate certain standard values for these fundamental
quantities. This is necessary in order to make the comparison between the
performances of different airplanes possible.
One of the first technical reports to mention a standard atmosphere was
NACA TM-15, released in 1921 [1]. Several standards now exist, the most
common being the 1976 U.S. Standard Atmosphere [2], the International
Standard Atmosphere (ISO 2533:1975), and the International Civil Aviation
Organization (ICAO) International Standard Atmosphere (Doc 7488-CD,
3rd ed., 1993). These are based on average atmospheric properties at midla-
titudes north. The one favored for this book will be the International Stan-
dard Atmosphere (ISA).
A series of assumptions are used to mathematically define a standard
atmosphere. These assumptions are:
• The air is a perfect gas.
• The air is dry.
• Hydrostatic equilibrium exists (dp ¼ 2 r g dh).
• The gravitational acceleration g is constant.

Geopotential vs Geometric Altitude

The assumption of constant gravitational acceleration is appropriate
when dealing with small changes in altitude such as the variation from sea
level to 65,617 ft (20 km). Over this relatively small change in altitude, the
actual gravitational acceleration decreases by approximately 0.62%. The
relationship of the gravitational acceleration g at a given altitude as compared
to the gravitational acceleration at sea level gSL is
   
g 6356:766 km 2 20,855,531 ft 2
¼ ¼ (1:6)
gSL 6356:766 km þ h 20,855,531 ft þ h
6 Operational Aircraft Performance and Flight Test Practices

where h is the geometric height above the mean sea level and 6356.766 km
(20,855,531 ft) represents the mean radius of the Earth. The value of the
gravitational acceleration at sea level gSL is 32.174 ft/s2 (9.8066 m/s2 ).
Accounting for altitude changes, the gravitational acceleration would be
about 0.1% lower at an altitude of 10,000 ft as compared to the sea level
value, about 0.35% lower at 36,000 ft, and 0.62% at 65,617 ft.
The modeling of the standard atmosphere requires, per the assumptions,
a constant gravitational acceleration. The variation of gravitational accelera-
tion can be accounted for in our modeling by defining a geopotential altitude
H. This geopotential altitude is a measure of the potential energy of a unit of
mass at a given altitude assuming a constant gravitational acceleration. It can
be defined mathematically as follows:
gSL dH ¼ g dh (1:7)
Equation (1.7) can be integrated to find the relationship between h and H.
The difference between the two altitudes is about 5 ft at h ¼ 10,000 ft
(H ¼ 9995 ft) and 202 ft at h ¼ 65,000 ft (H ¼ 64,798 ft). Therefore, the
default altitude to be used when discussing standard atmosphere in this
course will be the geopotential altitude H.

Modeling the Standard Atmosphere

The properties of air change with altitude. As the air gets thinner, the
pressure will decrease and so will the density. The temperature will vary from
layer to layer. Figure 1.1 indicates how the atmospheric properties, the ratio of
property at altitude to the sea level value, vary with the altitude. Note that the
troposphere extends from 0 to 11 km (36,089 ft) and the stratosphere from
11 km (36,089 ft) to 20 km (65,617 ft) for the 1976 U.S. Standard Atmosphere.
The values of temperature, pressure, density, and gravitational accelera-
tion at mean sea level, in the standard atmosphere, are
TSL ¼ 288:15 K ¼ 518:678 R ¼ 158C ¼ 598F

pSL ¼ 101,325 N=m2 ¼ 2116:216 lb=ft2 ¼ 14:696 psi ¼ 29:9213 inHg

rSL ¼ 1:225 kg=m3 ¼ 0:0023769 slug=ft3 ¼ 0:0765 lbm =ft3

gSL ¼ 9:8066 m=s2 ¼ 32:174 ft=s2
The standard atmosphere concept came from the need for scientists to
compare performance between airplanes and to estimate the performance
of any airplane at any altitude. A. Toussaint proposed the first formula for
the decrease of temperature with height in 1920. One of the first reports to
mention a standard atmosphere is NACA TM-15, published in 1921.
T ¼ 15  0:0065 H (1:8)
Where T is in degrees Celsius and H is in meters.
 CHAPTER 1 Atmosphere 7

70,000

60,000

Temperature
50,000
Geopotential altitude (ft)

Air density
40,000

30,000

Air pressure
20,000

10,000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Ratio of atmospheric properties to sea level reference

Fig. 1.1 Standard atmosphere.

Modeling the Troposphere

The troposphere is characterized by a linearly
decreasing temperature with altitude. This model Rule of thumb: The air
has a temperature lapse rate aLR of 6.5 Kelvins temperature decreases by
decrease per kilometer increase in geopotential alti- approximately 28C per
1000 ft pressure altitude in
tude.
the troposphere under
standard conditions or 6.58C
T ¼ TSL þ aLR H (1:9) per km.

where aLR ¼ 20.0065 K/m ¼ 20.00658C/m ¼

20.0019812 K/ft ¼ 20.0035668F/ft ; temperature
lapse rate. The temperature in the standard atmos- F2
phere troposphere ranges from 158C (598F) at sea
level to 256.58C (269.78F) at the tropopause. dy
Small column of air

Starting from the hydrostatic equilibrium prin-

ciple, combined with the perfect gas assumption,
W
one gets dh

p gSL
dp ¼ r g dh ¼ r gSL dH ¼  dH
RT
p gSL dH
¼
R ½T1  aLR ðH  H1 Þ F1
8 Operational Aircraft Performance and Flight Test Practices

Subscript 1 indicates the atmospheric conditions at altitude H1 . Integrat-

ing from an initial altitude H1 to a second altitude H2 ,
H
ð2
dp gSL
¼ ½lnð pÞH 2
H1 ¼  a ½lnðT1 þ aLR fH  H1 gÞH2
H1
p LR R
H1

Simplifying,
   
p2 gSL T2
ln ¼ ln
p1 aLR R T1

  gSL
p2 T2 aLR R
¼
p1 T1
The reference atmospheric conditions (condition 1) are the mean sea
level values as defined earlier in the text. Substituting the sea level conditions
for condition 1 in the previous equation and subscript H for the conditions
at a given altitude in the troposphere, one gets the following expression
for the change in pressure with altitude:

  gSL
TH aLR R
pH ¼ pSL (1:10)
TSL
Replacing the temperature at altitude with Eq. (1.9),

  gSL
aLR
pH ¼ pSL 1 þ H aLR R (1:11)
TSL
The density will vary with both the pressure and the temperature.
Combining Eqs. (1.5) and (1.11) gives
   
      gSL     gSL  1
r2 p2 T1 T2 aLR R T1 T2 aLR R
¼ ¼ ¼
r1 p1 T2 T1 T2 T1
Therefore,
 
 gSL  1
TH aLR R
rH ¼ rSL (1:12)
TSL
 
  gSL  1
a
rH ¼ rSL 1 þ LR H aLR R (1:13)
TSL
 CHAPTER 1 Atmosphere 9

It should be noted that the value of the exponent is the same in both sets
of units. These values are dimensionless and equal to
–gSL =ðaLR RÞ  5:2559
– ½ðgSL =faLR RgÞ þ 1  4:2559

Modeling the Stratosphere

The same exercise is repeated to develop the equations for the properties
of the air in the first part of the stratosphere where the temperature gradient
is zero. The steps are the same as for the troposphere with the exception that
the air temperature does not vary with altitude. Therefore,
p
dp ¼ r gSL dH ¼  g dH
R TS SL
where TS is the temperature in the low stratosphere, 256.58C (269.78F), that
is, the temperature at the tropopause. Integrating the equation between two
altitudes, one gets
H
ð2 H
ð2
dp g
¼  SL dH
p R TS
H1 H1
 
p2 g
ln ¼  SL ðH2  H1 Þ
p1 R TS
 
gSL
p2 ¼ p1 exp  ðH2  H1 Þ
R TS
Let condition 1 be the air properties at the tropopause (H1 ¼ 36,089 ft or
11 km), and the temperature used must be an absolute temperature
(216.66 K, or 389.998R). This equation becomes
 
gSL
p ¼ p11,000 m exp  ðH  11,000 mÞ
R TS
(1:14)
¼ p11,000 m expð0:00015768 ðH  11,000 mÞÞ
¼ p11,000 m expððH  11,000 mÞ=6341:9Þ
in SI units, and
 
g
p ¼ p36,089 ft exp  SL ðH  36,089 ftÞ
R TS
(1:15)
¼ p36,089 ft expð0:000048079 ðH  36,089 ftÞÞ
¼ p36,089 ft expððH  36,089Þ=20,800Þ
10 Operational Aircraft Performance and Flight Test Practices

in imperial units. Knowing that the temperature remains constant in the low
stratosphere, one can compute the variation of air density with altitude as
follows:
   
p p
¼
r R TS 1 r R TS 2
or
p2 r2
¼
p1 r1
and
 
gSL
r2 ¼ r1 exp  ðH2  H1 Þ
R TS
Therefore, in SI units
 
gSL
r ¼ r11,000 m exp  ðH  11,000 mÞ
R TS
(1:16)
¼ r11,000 m expð0:00015768 ðH  11,000 mÞÞ
¼ r11,000 m expððH  11,000 mÞ=6341:9Þ

and in imperial units

 
g
r ¼ r36,089 ft exp  SL ðH  36,089 ftÞ
R TS
(1:17)
¼ r36,089 ft expð0:000048079 ðH  36,089 ftÞÞ
¼ r36,089 ft expððH  36,089Þ=20,800Þ

Example 1.1
Compute the atmospheric properties, in SI units, at the tropopause.

Solution: The tropopause is the upper limit of the troposphere; it is at an

altitude of 11 km. Using Eq. (1.9), one can compute the temperature
T11 km ¼ TSL þ aLR ð11,000 mÞ
T11 km ¼ 288:15 K þ ð0:0065 K=mÞ ð11,000 mÞ
T11 km ¼ 216:65 K ¼ 56:58C

(Continued)
 CHAPTER 1 Atmosphere 11

Example 1.1 (Continued)

For the pressure, Eq. (1.11) is used.
  gSL
aLR
p11,000 m ¼ pSL 1þ ð11,000 mÞ aLR R
TSL

 5:2559
ð0:0065 K=mÞ
p11,000 m ¼ ð1,01,325 PaÞ 1 þ ð11,000 mÞ
288:15 K

p11,000 m ¼ ð1,01,325 PaÞ ð0:7519Þ5:2559 ¼ ð1,01,325 PaÞ ð0:22336Þ

p11,000 m ¼ 22,632 Pa

Finally, the air density value at 11 km altitude is

 
  gSL  1
a
r11,000 m ¼ rSL 1 þ LR ð11,000 mÞ aLR R
TSL

 4:2559
  ð0:0065 K=mÞ
r11,000 m ¼ 1:225 kg=m3 1 þ ð11,000 mÞ
288:15 K

   
r11,000 m ¼ 1:225 kg=m3 ð0:7519Þ4:2559 ¼ 1:225 kg=m3 ð0:29707Þ

r11,000 m ¼ 0:36391 kg=m3

Atmospheric Ratios
Instead of carrying the units in the computation of the atmospheric
properties, one can simply use a ratio of the property at altitude to the sea
level value. The temperature ratio u is defined as the ratio of the temperature
at altitude over the temperature at mean sea level. The pressure ratio d is
the pressure at altitude over the pressure at mean sea level. Finally, the
density ratio s is the ratio of the density at altitude over the density at
mean sea level.

TH pH rH
u¼ d¼ s¼ (1:18)
TSL pSL rSL
12 Operational Aircraft Performance and Flight Test Practices

The temperature ratio equation for the troposphere is obtained by divid-

ing both sides of Eq. (1.9) by the temperature at sea level TSL
aLR
u¼1þ H
TSL

u ¼ 1  ð2:2557  105 ÞH ¼ 1  ð6:875  106 ÞH (1:19)

H ! meters H ! feet
The pressure ratio equation in the standard atmosphere troposphere is
obtained from Eq. (1.11)
d ¼ u5:2559 (1:20)

The air density ratio equation is obtained from Eq. (1.13)

s ¼ u4:2559 (1:21)
Now for the standard atmosphere model in the stratosphere, knowing
that the temperature is constant, we get the following equations from Eqs.
(1.14) to (1.17) (for the first equation of both d and s, H is in meters; for
the second, H is in feet):
u ¼ 0:7519

d ¼ 0:2234 eð0:00015768ÞðH11,000Þ ¼ 0:2234 eð0:000048079ÞðH36,089Þ

s ¼ 0:2971 eð0:00015768ÞðH11,000Þ ¼ 0:2971 eð0:000048079ÞðH36,089Þ

(1:22)

Example 1.2
At what altitude is the air density half the value at sea level in a standard
atmosphere?

Solution: From Table 1.1, it can be noticed that the air density ratio at the
tropopause is 0.2971 or 29.71% the value of the sea level air density. Therefore,
the altitude at which the air density is 50% that of sea level must lie within the
troposphere. Combining Eqs. (1.19) and (1.21), one gets
 4:2559
s ¼ 1  6:875  106 H

from which one can find the following equation for altitude in feet:

1  s1=4:2559
H¼
6:875  106

(Continued)
 CHAPTER 1 Atmosphere 13

Example 1.2 (Continued)

Now, s ¼ 0.5 for the air density to be half the value of the sea level air density.
The resulting altitude is

1  ð0:5Þ1=4:2559
H¼ ¼ 21861:2 ft
6:875  106

Table 1.1 was created using these equations and Tip: Using ratios allows you
is provided for quick reference to the atmospheric to quickly verify if the airplane
properties variation with altitude. is in the troposphere or
stratosphere. For values
greater than 0.7519 or 0.2234
or 0.2971 for the temperature,
Temperature, Pressure, and Density pressure, and density ratios
respectively, then the altitude
Altitudes is less than 36,089 ft. It is
much easier to remember
If you know the air pressure or air temperature in than using values with units.
a standard atmosphere, using any of the equations Note that the temperature
developed in the section “Modeling the Standard ratio will not be less than
Atmosphere” would allow you to compute the air- 0.7519 in a
standard atmosphere.
plane altitude as was done in Example 1.2. In
reality, standard atmosphere conditions are rarely
encountered, the standard atmosphere being an
average of the atmospheric conditions at midlatitudes north of the equator.
Three new types of altitudes are now defined: the pressure Hp , the temp-
erature HT , and the density Hr altitudes. The pressure altitude is the altitude
in the standard atmosphere where the air pressure is the same as the one
measured by the airplane at the current altitude. For example, if the static
pressure being measured by the airplane air data system was 95 kPa
(13.778 psi), the altitude in the standard atmosphere, based on this pressure,
would be
95 kPa
d¼ ¼ 0:9376
101:325 kPa
 5:2559
d ¼ 1  6:875  106 Hp

1  d1=5:2559
Hp ¼ ¼ 1772 ft
6:875  106
Following the same principle, the temperature altitude is the standard
atmosphere altitude where air temperature is the same as the one measured
by the airplane current altitude. The air density altitude is the standard
14 Operational Aircraft Performance and Flight Test Practices

Table 1.1 Standard Atmosphere Ratios

Altitude Altitude
(ft) Theta Delta Sigma (ft) Theta Delta Sigma
0 1.0000 1.0000 1.0000 33,000 0.7731 0.2586 0.3345
1000 0.9931 0.9644 0.9711 34,000 0.7663 0.2468 0.3220
2000 0.9863 0.9298 0.9428 35,000 0.7594 0.2353 0.3099
3000 0.9794 0.8962 0.9151 36,089 0.7519 0.2234 0.2971
4000 0.9725 0.8637 0.8881 37,000 0.7519 0.2138 0.2844
5000 0.9656 0.8321 0.8617 38,000 0.7519 0.2038 0.2710
6000 0.9588 0.8014 0.8359 39,000 0.7519 0.1942 0.2583
7000 0.9519 0.7716 0.8107 40,000 0.7519 0.1851 0.2462
8000 0.9450 0.7428 0.7860 41,000 0.7519 0.1764 0.2346
9000 0.9381 0.7148 0.7620 42,000 0.7519 0.1681 0.2236
10,000 0.9313 0.6877 0.7385 43,000 0.7519 0.1602 0.2131
11,000 0.9244 0.6615 0.7156 44,000 0.7519 0.1527 0.2031
12,000 0.9175 0.6360 0.6932 45,000 0.7519 0.1456 0.1936
13,000 0.9106 0.6114 0.6714 46,000 0.7519 0.1387 0.1845
14,000 0.9038 0.5875 0.6500 47,000 0.7519 0.1322 0.1758
15,000 0.8969 0.5644 0.6293 48,000 0.7519 0.1260 0.1676
16,000 0.8900 0.5420 0.6090 49,000 0.7519 0.1201 0.1597
17,000 0.8831 0.5204 0.5892 50,000 0.7519 0.1144 0.1522
18,000 0.8763 0.4994 0.5699 51,000 0.7519 0.1091 0.1451
19,000 0.8694 0.4792 0.5512 52,000 0.7519 0.1040 0.1383
20,000 0.8625 0.4596 0.5328 53,000 0.7519 0.0991 0.1318
21,000 0.8556 0.4406 0.5150 54,000 0.7519 0.0944 0.1256
22,000 0.8488 0.4224 0.4976 55,000 0.7519 0.0900 0.1197
23,000 0.8419 0.4047 0.4807 56,000 0.7519 0.0858 0.1141
24,000 0.8350 0.3876 0.4642 57,000 0.7519 0.0817 0.1087
25,000 0.8281 0.3711 0.4482 58,000 0.7519 0.0779 0.1036
26,000 0.8213 0.3552 0.4325 59,000 0.7519 0.0742 0.0987
27,000 0.8144 0.3399 0.4173 60,000 0.7519 0.0708 0.0941
28,000 0.8075 0.3251 0.4025 61,000 0.7519 0.0674 0.0897
29,000 0.8006 0.3108 0.3882 62,000 0.7519 0.0643 0.0855
30,000 0.7938 0.2970 0.3742 63,000 0.7519 0.0613 0.0815
31,000 0.7869 0.2837 0.3606 64,000 0.7519 0.0584 0.0776
32,000 0.7800 0.2709 0.3473 65,000 0.7519 0.0556 0.0740
 CHAPTER 1 Atmosphere 15

Example 1.3
You are flying at a pressure altitude of 20,000 ft where the temperature is
228F. What is the air pressure value? What are the temperature ratio and
temperature altitude? What are the air density ratio and air density altitude?

Solution: A pressure altitude of Hp ¼ 20,000 ft corresponds to a pressure

ratio d ¼ 0.4596 (Table 1.1), and therefore a pressure of
 
p ¼ pSL d ¼ 2116:216 lb=ft2 ð0:4596Þ ¼ 972:6 lb=ft2

A temperature of 228F (457.698F) corresponds to a temperature ratio of

457:698R
u¼ ¼ 0:8824
518:698R
Using Eq. (1.9), one can find the temperature altitude
ðTH  TSL Þ ð457:698R  518:698RÞ
HT ¼ ¼ ¼ 17,106 ft
aLR 0:0035668R=ft
Note that this is a different altitude than the pressure altitude specified in
the question. The air density ratio is
0:4596
s ¼ d=u ¼ ¼ 0:5208
0:8824
and the air density altitude is

1  s1=4:2559
Hr ¼ ¼ 20,671 ft
6:875  106
All three altitudes, based on the measured air properties, differ from one
another. This is typical of what should be expected during normal flight.
Which one represents the real geometric height (tape measure distance)
above the mean sea level? Maybe none of the three is representative!
Therefore, how does one know the current airplane geometric altitude?
This subject will be addressed in Chapter 2, Measuring Altitude.

atmosphere altitude where the air density is the same as the one at the
current airplane altitude.

Exercises
1. Produce a small computer program to generate the standard atmosphere
values of temperature, pressure, and density ratios as well as the speed of
sound for altitudes varying from 0 ft to 45,000 ft, printing the results every
1000 ft. Present the results in a table format.
16 Operational Aircraft Performance and Flight Test Practices

2. The Federal Aviation Administration (http://

www.faa.gov) regulation 14 CFR §91.211— Note: This program will be
used as a building block for
Supplemental Oxygen requires under paragraph an altimetry and airspeed
(a)(2) that the flight crew use oxygen when flying tool as we progress through
at or above 14,000 ft pressure altitude for an all chapters.
unpressurized airplane. What is the pressure in psi
at that altitude? What is the pressure ratio?
3. You are at Denver International Airport in Colorado, elevation 5435 ft
(geometric). The pressure reading at the airport is 12.1 psi. What is the
pressure altitude? The outside air temperature is 258C. What is the density
altitude?

References
[1] Grimault, P., “On the Definition of the Standard Atmosphere,” NACA TM-15, 1921.
[2] “U.S. Standard Atmosphere, 1976,” NASA TM-X-74335, NOAA S/T 76-1562,
October 1976.
Chapter 2 Measuring
Altitude

Chapter Objective
The objective of this chapter is to introduce you to the various sources of alti-
tude in aviation and the means to measure them. Operational restrictions and
temperature corrections for barometric altimeters are also presented. The
material is supported by the presentation of several incidents involving the
use of the wrong altitude source.

Introduction

K
nowing the current flight altitude is essential at the lower altitudes to
avoid obstacles such as mountains and at all altitude to avoid other
airplanes. When flying in a standard atmosphere, measuring the
pressure or temperature and using Eqs. (1.11) or (1.9) from Chapter 1
would provide the airplane’s altitude above mean sea level (MSL) in the tro-
posphere. In the stratosphere, only the measurement of the pressure can
provide a source of altitude measurement, because the air temperature
remains constant. Unfortunately, the atmosphere is rarely if ever standard.
For example, Fig. 2.1 shows a variation of temperature with pressure altitude.
The temperature values were obtained during a climb to cruise altitude on a
warm summer day in the American Midwest.
Points of interest to note from the graph of Fig. 2.1 include the following:
• There is a temperature inversion in the lower altitudes. A temperature
inversion is a condition in which the temperature in the troposphere
increases with altitude instead of the normal decrease with altitude.
Several processes such as the passage of a weather front can create
temperature inversions.
• The temperature at a given pressure altitude is on average 15.58C above the
standard atmosphere temperature for the same pressure altitude. The
temperature difference, in this example 15.58C, between the international
standard atmosphere (ISA) and the actual temperature is often expressed
as DISA ¼ 15.58C.

17
18 Operational Aircraft Performance and Flight Test Practices

35,000

30,000

25,000
Pressure altitude (ft)

20,000

15,000

10,000

5000

0
–50 –40 –30 –20 –10 0 10 20 30 40
Temperature (deg C)
Actual temperature Std atmosphere

Fig. 2.1 Temperature variation with altitude as measured on a hot summer day in the
American Midwest.

• The average temperature lapse rate from start of climb to cruise altitude
on the chart is 2.058C per 1000 ft, whereas the standard atmosphere
temperature lapse rate is 1.988C per 1000 ft, but the temperature profile
is far from being standard.
• At a pressure altitude of 20,000 ft, the temperature altitude is 12,875 ft
(i.e., 210.58C instead of the 224.68C expected for ISA conditions at
20,000 ft).
Air pressure, on the other hand, always decreases with altitude. The rate of
change of air pressure with altitude depends on the local temperature lapse
rate. The pressure altitude is used as the airplane’s reference altitude.

Pressure Altitude
When flying in a standard atmosphere, the measured air pressure can be
converted into a pressure altitude Hp . In the troposphere, the following
equation [derived from Eq. (1.11) for imperial units] can be used:

h i h i
1  ð p=pSL Þ1=5:2559 1  d1=5:2559 Note: One of the
assumptions used in deriving
Hp ¼ ¼
6:875  106 ft1 6:875  106 ft1 Eq. (2.1) was that the
temperature lapse rate
(2:1) is standard.
 CHAPTER 2 Measuring Altitude 19

ps

Orifices

Fig. 2.2 Static-pressure orifice (top) and tube (bottom).

In the stratosphere, the equation is [derived from Eq. (1.15) for imperial
units]:
n  h io
Hp ¼ 36,089  20,800 ln 4:4763 p=pSL ft (2:2)

The units favored for this course for altitude will be feet.

Measurement of Pressure Altitude

Static air pressure is measured through an orifice oriented perpendicular
to the local airflow so as not to measure impact pressure qc . Typically, on an
airplane, these orifices can be simple vents on the side of the fuselage or side
openings on static pressure tubes (see Figs. 2.2, 2.3, 2.4, and 2.5).

Fig. 2.3 Static pressure port on the side of a Super Constellation airliner.
20 Operational Aircraft Performance and Flight Test Practices

Fig. 2.4 Static pressure port on the side of an MD-11.

A body in movement through an air mass will generate a pressure field

with locally varying static air pressure. For a static pressure probe to be
used as a source of pressure for altitude computation, it must be located
in a region where the local pressure will be as close as possible to the
ambient air pressure. Any difference between the local air pressure and the
free stream air pressure will create a reading error when the pressure is con-
verted into an altitude. This will be discussed in more details in Chapter 4,
Altimetry Error.

Static ports

Fig. 2.5 Pitot-static probe (static ports on side of probe).

 CHAPTER 2 Measuring Altitude 21

AoA = 0 deg

eg
= 20 d +1.0
AoA AoA = 0 deg

⌬p
qc 0

–1.0 AoA = 20 deg

Fig. 2.6 Delta pressure between free stream and local static pressures.

Figure 2.6 shows how local static pressure could vary as the air mass flows
around an airplane at two different angles of attack (AoAs).

Barometric Altimeter and Barometric Setting

The static pressure reading is channeled to an instrument that converts
air pressure to an altitude. This instrument is called a barometric altimeter.
This altimeter can vary in complexity depending on the desired accuracy
of the altitude presentation.
The altimeter is usually an airtight instrument case vented to the
static pressure source; Fig. 2.7 shows the details of a typical barometric alti-
meter for a general aviation plane. This example contains sealed aneroid cap-
sules. As the air pressure changes in the compartment around the sealed
capsule, the capsule will expand or contract. This movement is captured by a
calibrated gearing and translated into an altitude reading on the altimeter face.
Mechanical altimeters will typically have an altitude range of 35,000 ft or
less (within the troposphere limits), and the mechanical gears translating the
movement of the aneroid capsules into an altitude reading will be calibrated
using Eq. (2.1), but have also been used to much higher altitudes (as high as
354,200 feet on the X-15 experimental airplane).
Alternatively, modern jetliners, business jets, and military airplanes are
equipped with air data computers (ADCs) on which the air pressure
source is connected and read by a pressure sensor. These computers translate
the sensed air pressure into an altitude. The ADC can be programmed to
switch from one equation to the next based on the sensed pressure, with
the tropopause being at a pressure ratio of 0.2234 in a standard atmosphere.
The barometric altimeter, mechanical or ADC, cannot distinguish
between a change in atmospheric pressure due to a change in altitude and
that due to the passage of a weather front. For this reason, barometric alti-
meters are equipped with a barometric setting window (also called an
22 Operational Aircraft Performance and Flight Test Practices

Aneroid capsules

ps

ps

Fig. 2.7 Barometric altimeter.

altimeter setting window) that allows the pilot to dial in an equivalent mean
sea level pressure (MSLP) (see Fig. 2.8). By changing the barometric setting
(or altimeter setting), the pilot effectively shifts the altitude curve up or
down for a given air pressure reading (see Fig. 2.9).
Consider the following example while referring to Fig. 2.9. An airport is
located at a geometric height of 6650 ft. The actual air pressure as measured
at the airport is 22.24 inHg (0.7433 pressure ratio). This corresponds to a
pressure altitude of 7985 ft in a standard atmosphere. If the pilot of a
parked airplane at the airport dials in a barometric setting of 28.22 inHg
instead of the standard atmosphere value of 29.92 inHg, the altimeter
reading would become 6650 ft, the geometric height of the airport. Then,
28.22 inHg would be the MSLP.
The equivalent mean sea level pressure (MSLP) can be determined as
follows. Starting from an indicated pressure altitude Hi and the standard
atmosphere pressure altitude Hp ,
@H
Hi ¼ Hp þ Dp (2:3)
@p
 CHAPTER 2 Measuring Altitude 23

Barometric setting

Fig. 2.8 Barometric setting window on various altimeter displays.

20,000
18,000 Standard atmosphere
Actual altitude/geometric height (ft)

with barometric setting

16,000 of 29.92 inHg at sea level
14,000
12,000
10,000
hp = 7985 ft
8000 Example: H = 6650 ft

6000
Changing the barometric
4000
setting shifts the curve
2000
0
10.00 15.00 20.00 25.00 30.00 35.00
22.24 28.22
Atmospheric pressure or barometric setting (inHg)

Fig. 2.9 Barometric setting impact on pressure altitude reading.

24 Operational Aircraft Performance and Flight Test Practices

From hydrostatic equilibrium,

 
dH R DT R TSL  aLR Hp
¼ ¼
dp pSL gSL pSL gSL
Let,
Dp ¼ pSL  MSLP
Therefore,
 
TSL R aLR R Hp
Hi ¼ Hp þ ðMSLP  pSL Þ  (2:4)
gSL pSL gSL pSL
The usual units for sea level pressure for barometric altimeters are
inches of mercury (inHg) or hectopascal (hPa), and the altitude units are in
feet. Using the inHg and ft units, Eq. (2.4) takes on the following form:
 
Hi ¼ Hp þ (MSLP  29:92 inHg) 924:8  0:0063586 Hp (2:5)

where Hi is now the indicated pressure altitude (units of feet) for a given
nonstandard day MSLP. This correction to the pressure altitude assumes
that the temperature lapse rate aLR is identical
to the one in the standard
atmosphere. Tip: A typical rule of thumb
The indicated altitude on an altimeter is an on the error between the
estimation of the airplane true altitude. If the indicated pressure altitude
and the actual pressure alti-
proper barometric setting is dialed in and tempera- tude is as much as 1000 ft/in.
ture conditions are close to ISA, the estimation is of mercury error.
good. A wrong setting will yield a wrong estimate.

Example 2.1
An airplane is descending in preparation for a landing. The airport elevation is
3600 ft (geometric). The proper barometric setting for the current conditions
at the airport is 28.95 inHg. By mistake, the pilot dials in 29.95 inHg (1 in. of
mercury difference). What will be the indicated altitude when the airplane
touches down at the airport? The temperature conditions are ISA.
Solution: If the proper barometric setting were dialed in, the indicated alti-
tude would be 3600 ft. Using Eq. (2.5), one can determine the actual pressure
altitude of the airport for the proper barometric setting of 28.95 inHg.
Hi  924:8 ðMSLP  29:92Þ 3600  924:8 (28:95  29:92)
Hp ¼ ¼
1  0:0063586 ðMSLP  29:92Þ 1  0:0063586 ð28:95  29:92Þ
¼ 4469 ft
(Continued)
 CHAPTER 2 Measuring Altitude 25

Example 2.1 (Continued)

This higher pressure altitude as compared to the field elevation is to be
expected for a low barometric setting (lower than 29.92 inHg). The impact
of the pilot mistakenly dialing in a setting that has a 1-inHg error
(29.95 inHg instead of 28.95 inHg) can be determined as follows. Knowing
the pressure altitude at the airport, as determined previously, one gets
 
Hi ¼ Hp þ ðMSLP  29:92Þ 924:8  0:0063586 Hp
¼ 4469 þ ð29:95  29:92Þ ½924:8  0:0063586 ð4469Þ
¼ 4496 ft

Therefore, the indicated altitude would be 896 ft higher than the actual
field elevation at the time of landing at the airport field elevation of 3600 ft.
This misleading information can be dangerous for airplanes flying in instru-
ment flight rules (IFR) conditions where visibility can be severely compro-
mised and can give a false sense of security to the pilot.

The following is an example of an incorrect altimeter setting incident

from the Aviation Safety Reporting System (ASRS, http://asrs.arc.nasa.gov).
ACN 941349: “We were cleared to descend to 14,000 FT while in a holding
pattern. Descending through FL180 we set the altimeter to the local setting.
The local altimeter setting was noted on the clipboard as 29.94. However the
correct altimeter setting was 29.34. A CRJ was ahead of us was leveled at
13,000 FT following the same clearance about 1/2 mile ahead of us. Upon
reaching 14,300 FT we received an RA stating ‘Monitor vertical speed’. I dis-
connected the autopilot and adjusted the flight path to remain in the TCAS
command bars. We notified ATC of the deviation and RA compliance. ATC
queried the CRJ about their altitude and they confirmed that they had
received an RA as well and that they were level at 13,000 FT at the time of
the advisory. Upon hearing this we checked the altimeter setting in the
AFIS and noted the discrepancy. We corrected the altimeter setting to
29.63 then climbed and leveled at 14,000 FT.”

The following incident occurred in 2006:

A Gemini Air Cargo Boeing MD-11F descended to 720 ft above ground in
darkness while still at 7 nm (13 km) from the airport; the normal height
for this distance is 2,200 ft. As the airplane was lining up with the runway,
air traffic control (ATC) cleared the airplane to descent to 2,000 ft on an alti-
meter setting of 974 mb (28.76 inHg). The pilots were distracted and did not
enter the proper setting, leaving the altimeters at 1013 mb (29.92 inHg).
They maintained 720 ft (2,000 ft indicated) until they intercepted the glide-
slope and landed normally.

Operational Regulations
Operationally, airworthiness regulations throughout the world provide
specific rules for adjusting the barometric setting of an airplane. In the
26 Operational Aircraft Performance and Flight Test Practices

United States (14 CFR Part §91.121, http://www.faa.gov) and in Canada

(CAR Part VI, Subpart 2, para 602.35, http://www.tc.gc.ca), air regulations
require that, for an airplane flying below 18,000 ft, the altimeter setting be
readjusted every 100 n miles (150 n miles in Canada) traveled with the
MSLP from the nearest weather station along the route.
The following are incidents and accidents from the Canadian Aviation
Safety Reporting System that relate to the setting of altimeters and the dis-
tance of the reporting source.
The first incident is number A80W0001:
The pilot stated that he was lined up [on approach] for the runway and that
the altimeter was reading 300 feet when the nose wheel struck the ice. The
pilot applied power . . . and flew back to [his departure airport over 100 n
miles south], where the landing was uneventful. The pressure in the area
was lower than the point of departure, sufficient to make the altimeter
read 250–300 ft high if not properly reset.

The second was an accident, number A80C0079:

On a night VFR flight, the pilot encountered deteriorating weather as he
approached his destination. He received an IFR clearance . . . During a pro-
cedure turn, the aircraft started to strike the tree tops. The aircraft stalled
and crashed in the trees. Because the airport had been closed for the night,
no altimeter setting was available. The FSS operator gave the pilot the
setting for XYZ (29.68) [approximately 90 n miles south] and for ABC
(29.87) [approximately 90 n miles east]. The aircraft altimeter was set at
29.94.

There are three possible barometric settings for normal airplane operation.
They are:

1. If the barometric setting is dialed to indicate airport elevation at

landing, the setting is given the code QNH. This is the code used for
navigation, terrain avoidance, and obstacle clearance.
2. If the barometric setting is dialed to indicate zero altitude at landing,
the setting is given the code QFE.
3. Finally, if the barometric setting is dialed to the standard atmosphere
sea level pressure of 29.92 inHg (1013.25 millibars), the setting is given
the code QNE.

Altimeter setting regions (ASRs) are regions where, below the transition alti-
tude, the altimeter must be set to QNH. The transition altitude for most of
Canada and the United States is 18,000 ft. Therefore, in the United States
and Canada, the MSLP is adjusted to the QNH code while en route below
18,000 ft and for takeoff and landing. Once the airplane exceeds the tran-
sition altitude of 18,000 ft pressure altitude, the barometric setting is set to
the QNE code.
The transition altitude varies throughout the world. For example, most of
Europe uses transition altitudes varying between 4000 and 6000 ft, Australia
 CHAPTER 2 Measuring Altitude 27

uses 10,000 ft, and Japan uses 14,000 ft. In South America, the transition alti-
tude varies between 3000 ft for Buenos Aires, Argentina, and 18,000 ft for La
Paz, Bolivia. In all cases, once the transition altitude is crossed while climbing,
the barometric setting must be set to the QNE code. Above the transition
altitude, the altitudes are referred to as flight levels (FL), where the altitude
is divided by 100. An FL of 370 (or FL370) corresponds to a pressure altitude
of 37,000 ft.
Standard pressure regions are regions where the altimeter setting is
always set to QNE. In the United States and Canada, the standard pressure
region includes all airspace at or above 18,000 ft MSL. As well, Canada’s
northern region (see Fig. 2.10) requires all airplanes to select 29.92 inHg
from low altitudes. For takeoff from an airport in the standard pressure
region, the departure airfield altimeter setting is used until terrain clear-
ance is assured, and then 29.92 inHg is selected. In preparation for
landing in the standard pressure region, the altimeter is kept to QNE
until final descent is started; at that time, the altimeter is set to the field
setting.
North America uses inches of mercury (inHg) to report barometric
settings, whereas Europe and several other places worldwide use hectoPascal
(hPa) or millibars (mb) to report barometric settings. The conversion factors
between the three units are:

29:92 inHg ¼ 1013:25 hPa ¼ 1013:25 mb

Fig. 2.10 Canadian altimeter regions. Source: IVAO Canadian Division.

28 Operational Aircraft Performance and Flight Test Practices

Of course, navigating from a region using inHg to one using mb or hPa

sometimes leads to missetting of the barometric altimeter, as can be seen
in this ASRS report:
The copilot who had copied the ATIS gave me 29.97 when I asked for QNH.
Gusty winds and [the controller] thick . . . accent weren’t helping things.
[Obstructions] seemed unusually close to our altitude. [The] copilot had
assumed 9-9-7 to be 29.97.

Because 997 mb actually corresponds to 29.49 inHg, the airplane had an alti-
meter setting 0.48 inHg too high, which actually put the airplane below its
intended altitude by almost 500 ft.

Cruising Altitudes: Operational Restrictions

The airspace above most countries is strictly controlled to allow safe
operation and proper airplane separation. In the United States, 14 CFR
Part §91.179 states:

a) In controlled airspace. Each person operating an aircraft under IFR in

level cruising flight in controlled airspace shall maintain the altitude or
flight level assigned that aircraft by ATC. . .
b) In uncontrolled airspace . . . each person operating an aircraft under IFR
in level cruising flight in uncontrolled airspace shall maintain an
appropriate altitude as follows:
(1) When operating below 18,000 feet MSL and—
(i) On a magnetic course of zero degrees through 179 degrees, any
odd thousand foot MSL altitude (such as 3,000, 5,000, or
7,000); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any
even thousand foot MSL altitude (such as 2,000, 4,000, or 6,000).
(2) When operating at or above 18,000 feet MSL but below flight level
290, and—
(i) On a magnetic course of zero degrees through 179 degrees, any
odd flight level (such as 190, 210, or 230); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any
even flight level (such as 180, 200, or 220).
(3) When operating at flight level 290 and above in non-RVSM airspace,
and—
(i) On a magnetic course of zero degrees through 179 degrees, any
flight level, at 4,000-foot intervals, beginning at and including
flight level 290 (such as flight level 290, 330, or 370); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any
flight level, at 4,000-foot intervals, beginning at and including
flight level 310 (such as flight level 310, 350, or 390).
(4) When operating at flight level 290 and above in airspace designated
as Reduced Vertical Separation Minimum (RVSM) airspace and—
(i) On a magnetic course of zero degrees through 179 degrees, any
odd flight level, at 2,000-foot intervals beginning at and including
flight level 290 (such as flight level 290, 310, 330, 350, 370, 390,
410); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any
even flight level, at 2000-foot intervals beginning at and
including flight level 300 (such as 300, 320, 340, 360, 380, 400).
 CHAPTER 2 Measuring Altitude 29

Table 2.1 Lowest Usable Flight Level with Altimeter Setting

Current Altimeter Setting Lowest Usable Flight Level

29.92 (or higher) 180
29.91 to 29.42 185
29.41 to 28.92 190
28.91 to 28.42 195
28.41 to 27.92 200
27.91 to 27.42 205
27.41 to 26.92 210

14 CFR §91.121 (2)(b) further specifies that the lowest usable flight level is
determined by the atmospheric pressure in the area of operation, as shown
in Table 2.1.

Reduced Vertical Separation Minimum (RVSM)

Reduced vertical separation minimum (RVSM) airspace extends from
FL290 to FL410 over most of the globe (see Fig. 2.11). It is a region within
which air traffic control (ATC) separates airplanes by a minimum of
1000 ft vertically between flight levels.

RVSM Worldwide implemented

as of November 2011

Canada North
Canada South Europe 1/02 Japan/Korea
4/02
1/05 9/05
Caucasus
Area
Domestic US 3/05
IRAQ FIR
1/05 NAT 3/97
3/11 EURASIA
China Pacific
11/11 2/00
11/07
Mid East
Pacific EUR/SAM 11/03 Western Pacific
2/00 WATRS 11/01 Corridor 1/02
Asia/Europe South China Sea
South of Himalayas 11/03 2/02
Africa
9/08
EURASIA
CAR/SAM Australia
1/05 Russian 11/01
Federation
Afghanistan
Kazakhstan
Krygyzstan
Mongolia
Implemented Tajikistan
Uzbekistan

Fig. 2.11 RVSM airspace. Source: Federal Aviation Administration, https://www.faa.gov/

air_traffic/separation_standards/rvsm/
30 Operational Aircraft Performance and Flight Test Practices

To operate in RVSM airspace, the operator and the airplane used by the
operator must be approved by a national agency (FAA, Transport Canada,
EASA, etc.). Under the FAA Part 91 regulations:
14 CFR §91.180 FAA
(a) Except as provided in paragraph (b) of this section, no person may
operate a civil aircraft in airspace designated as Reduced Vertical
Separation Minimum (RVSM) airspace unless:
(1) The operator and the operator’s aircraft comply with the minimum
standards of appendix G of this part; and
(2) The operator is authorized by the Administrator or the country of
registry to conduct such operations.
(b) The Administrator may authorize a deviation from the requirements of
this section.

Appendix G to Part 91, Operations in Reduced Vertical Separation Minimum

(RVSM) Airspace, defines the minimum requirements that must be met by
the airplanes and the operators to be allowed into RVSM airspace without
special handling. The appendix is located at http://www.gpo.gov/fdsys/
pkg/CFR-2012-title14-vol2/pdf/CFR-2012-title14-vol2-part91-appG.pdf.
We will cover the subject of RVSM in more detail in a dedicated section
on RVSM in Chapter 9.

Temperature Effects on Pressure Altitude Measurement

The barometric correction allows the pilot to adjust the displayed alti-
tude to account for the equivalent mean sea level pressure in the region
of flight and provide a barometric altitude that is a close estimate of the
geometric altitude of the airplane; however, it does not account for large
temperature deviation from standard. A temperature that is significantly
different from the standard temperature for a given pressure altitude
will induce an error between the pressure altitude indicated and the geo-
metric altitude of the airplane. If the temperature is higher than ISA,
the airplane geometric altitude will be higher than the altitude indicated
by the altimeter; if the temperature is lower than ISA, the airplane will
be lower than the indicated altitude. For the colder temperature, the
difference between the barometric altitude and the geometric altitude
can be on the order of a 4% height increase per 108C below standard
temperature.
Some airworthiness authorities, such as Transport Canada, provide
operational procedures where an altitude correction is applied to the baro-
metric altitude based on temperature deviation from standard conditions.
These procedures apply to the instrument flight rules (IFR) minimum
altitudes when the temperature is below standard atmosphere values. This
correction is required to ensure terrain and obstacle clearance during an
approach.
 CHAPTER 2 Measuring Altitude 31

The International Civil Aviation Organization (ICAO) provides an

equation to correct for off-standard temperature conditions (see [1]). Trans-
port Canada has adopted this means for applying corrections to published
minimum IFR altitudes. This equation produces results that are within 5%
of the accurate correction for altimeter-setting sources up to 3000 m
(10,000 feet) and with minimum heights up to 1500 m (5000 feet) above
that source. The equation is
ð15  T0 Þ
DH ¼ h (2:6)
ð273:16 þ T0  1=2 aLR ðh þ hSS ÞÞ
where
DH ¼ Altitude correction due to temperature deviation from
standard.
h ¼ Minimum IFR height above the barometric setting source.
The setting source is normally the aerodrome.
T0 ¼ The aerodrome (or specified temperature reporting point)
temperature (in 8C) adjusted to sea level. It is equal to

T0 ¼ Taerodrome þ aLR haerodrome (2:7)

aLR ¼ 0.00658C per meter or 0.001988C per foot, standard temp-

erature lapse rate.
hss ¼ Barometric setting source elevation (again, normally the
aerodrome).
Taerodrome ¼ Aerodrome (or specified temperature reporting point)
temperature.
haerodrome ¼ Aerodrome (or specified temperature reporting point)
elevation.
See Fig. 2.12 for more explanation.

Minimum IFR altitude

h, minimum authorized obstacle clearance height

(geometric) above baro setting source

Aerodrome
and altitude
setting point hss and haerodrome
MSL

Fig. 2.12 Parameters of Eq. (2.6).

32 Operational Aircraft Performance and Flight Test Practices

Example 2.2
An airplane is approaching an airport with a field elevation of 2500 ft. The area
around the airport is mountainous with peaks at 5000 ft elevation. The
minimum clearance height authorized is 1000 ft above the highest peak; there-
fore, the minimum IFR altitude is 6000 ft. The airport temperature is 2408C,
and the MSLP, as provided by the airport tower, is 29.55 inHg. What is the
pressure altitude at the airport, and what is the required altitude correction
that must be applied to the minimum IFR altitude to ensure the airplane
will not descend below a geometric height of 6000 ft?

Solution: The airport pressure altitude can be found using Eq. (2.5), where
Hi is replaced by the geometric field elevation (h ¼ 2500 ft)

h  924:8 ðMSLP  29:92Þ 2500  924:8 (29:55  29:92)

Hp ¼ ¼
1  0:0063586 ðMSLP  29:92Þ 1  0:0063586 ð29:55  29:92Þ
¼ 2835 ft

For this pressure altitude, a standard atmosphere temperature would

be 9.48C. The actual temperature of 2408C is a deviation from standard of
249.48C (i.e., DISA ¼ 249.48C). The temperature correction to be added
to the indicated altitude to ensure the airplane does not descend below the
minimum geometric height of 6000 ft can be computed as follows:
T0 ¼ Tairport þ aLR hairport ¼ 408C þ ð0:001988C=ftÞ 2500 ft ¼ 358C

The minimum height above the airport h is 3500 ft (6000 ft minimum

IFR elevation minus the airport elevation). The resulting temperature correc-
tion to the indicated altitude is, therefore, from Eq. (2.6)

(15  T0 )
DH ¼ h
(273:16 þ T0  1=2 aLR (h þ hss )
ð15  ð35ÞÞ
¼ ð3500 ftÞ  
273:16 þ (35)  1=2 0:001988 C=ft ð3500 ft þ 2500 ftÞ
¼ 754 ft

Therefore, the minimum indicated altitude the pilot should follow to

ensure proper obstacle clearance should be 6754 ft.

Table 2.2 was generated from Eq. (2.6) assuming a sea-level aerodrome
and reporting source. The height correction was rounded to the higher
10-ft value. This table clearly shows the magnitude of the correction required
for cold temperature.
 Table 2.2 ICAO Altimeter Correction with Temperature Deviation from Standard

Aerodrome Height above the Elevation of the Altimeter Setting Source [feet]
Temperature
[88 C] 200 300 400 500 600 700 800 900 1000 1500 2000 3000 4000 5000
0 20 20 30 30 40 40 50 50 60 90 120 170 230 280
210 20 30 40 50 60 70 80 90 100 150 200 290 390 490
220 30 50 60 70 90 100 120 130 140 210 280 420 570 710
230 40 60 80 100 120 140 150 170 190 280 380 570 760 950

CHAPTER 2
240 50 80 100 120 150 170 190 220 240 360 480 720 970 1210
250 60 90 120 150 180 210 240 270 300 450 590 890 1190 1500

Measuring Altitude
33
34 Operational Aircraft Performance and Flight Test Practices

The following is an example of an incident in which temperature correc-

tions were not applied during an approach for landing.
A southern operator was making a nonprecision approach to an airport in the
interior of British Columbia in an MD-80 with an airport temperature of
2278C. The airplane was cleared for an approach by Vancouver Center
and told to contact the tower for landing clearance. The crew then aborted
the approach following a Ground Proximity Warning System warning.
The crew performed a different nonprecision approach and landed safely.
Questioned after the incident, it was clear that the crew had not applied the
temperature correction to the procedure turn altitude. The procedure turn
altitude for the airport was 4900 ft above field elevation, so for the 2278C
airfield temperature, the crew should have added 800 ft to the published
procedure turn altitude to account for the low temperature. It is estimated
that the airplane missed the top of the mountain in the region by about 150 ft.

Equation (2.6) assumes a standard temperature lapse rate and that the proper
barometric setting is used. The actual lapse rate may vary considerably from
the assumed standard, depending on latitude and time of year. However, the
corrections derived from the linear approximation are satisfactory for general
application at altitudes up to 4500 m (about 15,000 ft).
The impact of a small variation in the temperature gradient on the rate of
change of pressure with altitude can be illustrated as follows: Fig. 2.13 com-
pares the variation of the pressure ratio of the standard atmosphere lapse rate
to a lapse rate half the value of the standard atmosphere lapse rate. The figure
assumes that both altimeters would have the same setting at sea level. Alti-
meters are calibrated for a standard lapse rate; therefore, an error would
result between an atmosphere with a lapse rate of half the value of the stan-
dard atmosphere and the reading of the altimeter. Figure 2.14 presents the

40,000
35,000
30,000
Altitude (ft)

25,000
20,000
15,000
10,000
5000
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Pressure ratio

Standard temperature lapse rate Half standard temperature lapse rate

Fig. 2.13 Impact of temperature lapse rate on pressure ratio.

 CHAPTER 2 Measuring Altitude 35

35,000

30,000

25,000
Altimeter altitude (ft)

20,000

15,000

10,000

5000

0
–2000 –1800 –1600 –1400 –1200 –1000 –800 –600 –400 –200 0

Altitude error (ft)

Fig. 2.14 Pressure altitude error due to other than standard temperature lapse rate.

resulting error of the altimeter. The figure indicates that in the present
example, the altimeter would read low; that is, the airplane is actually
higher than what the altimeter indicates. The error at the low altitude is
small, but grows rapidly with altitude; for example, at 1000 ft indicated alti-
tude, the altitude error is approximately 2 ft for this example, but at an indi-
cated altitude of 30,000 ft, the error is greater than 1700 ft. With the majority
of the world’s airports below 6000 ft altitude, the effects of temperature lapse
rate variation represent a small enough error for navigation purposes. For
navigation purposes, in regions where the temperature lapse rate could
have an important impact, the air traffic controllers will provide additional
altitude corrections.

Radar Altitude
Barometric altimeters determine the altitude of the airplane with respect
to a reference MSL based on a standard atmosphere barometric setting cor-
rected. This is the standard reference for altitude for normal flights such as
climb, cruise, or descent conditions, and for takeoff and landing.
For flights at low altitudes, such as approach and landing in low visibility
conditions, an altimeter that can provide a precise height above ground
level (AGL) is often required. A radar altimeter can perform this task. A
radar altimeter consists of a transmitter, usually with a dedicated antenna,
and a receiver, again with a dedicated antenna, located in the region that is
usually the lowest part of the airplane in the normal takeoff and landing atti-
tudes (see Fig. 2.15). The transmitter emits a radio wave toward the ground,
and the receiver picks up the echo. The travel time to and from the ground is
then computed. The airplane height above ground is then the radio wave
36 Operational Aircraft Performance and Flight Test Practices

Radar
signal

Fig. 2.15 Typical radar altimeter antenna location.

travel time multiplied by the speed of light, divided in half (because the radio
signal travels to and from the ground, covering twice the distance separating
the airplane from the ground).
Although a radar altimeter (also called a RADALT) can provide precise
height over the current terrain (typical precision of about 1 m at low altitude,
about 1–3% at higher altitudes), it usually does not provide a view forward of
the airplane. Therefore, the pilot may see rapid variation of radar altitude
while flying at a constant flight path angle, such as during an approach to
land where the pressure altitude would vary in a more predictable fashion
(see Fig. 2.16). Figure 2.17 provides a comparison of radar altitude variation
as a function of pressure altitude variation during a takeoff from a given
airport. Note that, for the first part of the takeoff, the radar altitude and
pressure altitude variation are a close match. This is to be expected
because the terrain along the centerline of a runway will be cleared and
leveled for a given distance past the end of the runway.
Radar altimeters have a limited envelope, usually being limited to
maximum pitch and roll values of the order of 30 deg to 60 deg. Maximum
height above ground (i.e., radar range) is also typically limited to 5000 ft or
less, the range being proportional to the fourth root of the transmission
power. These values are sufficient for normal takeoff and landing envelopes.
Some strike fighters may have an expanded radar altimeter envelope com-
bined with forward-looking radar to allow for low-level flights.

Flight path Radar altitude

Runway Runway elevation

Fig. 2.16 Rolling terrain on the approach path.

 CHAPTER 2 Measuring Altitude 37

800

700 Radar altimeter

(smoothed)
600
Radar altitude (ft)

500

400

300 Pressure altitude

increase from liftoff
200

100

0
1100 1200 1300 1400 1500 1600 1700 1800 1900
Pressure altitude (ft)

Fig. 2.17 Example of radar vs pressure altitude variation on takeoff.

GPS Altitude
Another source of altitude is the Global Positioning System (GPS). GPS is a
satellite navigation system developed and controlled by the U.S. Department of
Defense (DoD). It consists of a constellation of satellites, nominally 24 oper-
ational satellites, in orbit around the earth in six different orbital planes
inclined at 55 deg with respect to the equator. The satellites emit a coded
signal that can be processed by GPS receivers. The signals from four different
satellites are required for the GPS receiver to compute a three-dimensional
position and GPS time. The position computed by the GPS receiver is in Earth-
centered, Earth-fixed x, y, z (ECEF XYZ) coordinates. These coordinates are
transformed to a position in terms of latitude, longitude, and height above a
reference altitude by using a model Earth. The model Earth on which the
GPS is based is WGS-84. The MSL altitude for a GPS receiver will be based
on this model Earth. The altitude provided by this model is geometric altitude.

Exercises
1. You are flying at a pressure altitude of 16,000 ft. Your altimeter barometric
setting is 30.55 inHg. What altitude is displayed on the altimeter? (Assume
zero static pressure error.)
2. You are flying in a standard pressure region on a heading of 65 deg. Would
you expect to be flying at FL250 or FL240?

Reference
[1] ICAO, Procedures for Air Navigation Services, Aircraft Operations, Vol. I, No. 5, 8168-
OPS/611, 2006.
Chapter 3 Measuring
Airspeed

Chapter Objective
The objective of this chapter is to introduce you to the various sources of air-
speed and the means of measuring them. This chapter also compares the air-
speeds to each other with varying atmospheric properties to show trends and
characteristics. Finally, some airplane incidents and accidents are presented to
show the importance of reliable airspeed presentation to the crew.

Introduction

A
irspeed is an important parameter of airplane performance. A pilot
must know the airspeed to take off and land safely. There are also
appropriate airspeeds for the steepest climb, fastest climb, longest
flight, best glide, descent, stall speed, and so forth. Airplane designers and
aerodynamicists must know the airspeed in a wind tunnel to evaluate a
model’s aerodynamic coefficients from which its performance will be estab-
lished. The following incident shows the impact of unreliable airspeed on the
flight of a modern airliner.

NTSB DCA09IA064, Northwest Airlines Airbus A330 –323, June 23,

2009. The aircraft was in normal cruise flight at high altitude, in the vicinity
of convective weather conditions, and the flight crew was aware of the
weather. Shortly before the unreliable airspeed events, the airplane encoun-
tered a visible effect of the weather conditions—the crew reported seeing
moderate rain and hail. Within seconds, the airplane’s autopilot and auto-
thrust disconnected, master warnings and cautions were indicated, primary
airspeed display was lost or fluctuated, and the airplane transitioned into alter-
nate flight law. According to Airbus documentation and analysis, if an air-
speed discrepancy of more than 20 kt or an altitude discrepancy of more
than 400 ft is detected between one ADR and the two others, the subject air
data reference (ADR) is rejected. Then if a discrepancy occurs between the
two remaining ADRs, all autoflight functions are lost, and autopilot, flight
director, and autothrust disconnect. Applicable procedures call for autopilot
and autothrust to be off, and for the pilot to pitch to 5 deg and set thrust
levers to the CLB detent. The flight crew turned the airplane to exit the

39
40 Operational Aircraft Performance and Flight Test Practices

weather area, and the airspeed indications returned to normal within a

short time.
The National Transportation Safety Board determined the probable
cause(s) of this incident to be:
Brief and temporary blockage of the pitot probes in cruise flight,
most likely due to ice crystals aloft, leading to erroneous airspeed indi-
cations and airplane automation degradation as designed.

Contributing to the incidents were design features of the Thales AA probes

that left them more susceptible to high-altitude ice crystal icing than other
approved pitot probe designs.

Usual units of speed in aerospace are knots (kt), which represent nautical
miles per hour (n mile/h). These are the units favored for this text. A nautical
mile is 6076 ft long, as compared to a mile, which is 5280 ft long.
1n mile ¼ 6076 ft
1n mile ¼ 5280 ft
1n mile ¼ 1:6093 km
The conversion factors among various speed units are
1n mile=h ¼ 1 kt
1n mph ¼ 0:8690 kt
1n km=h ¼ 0:5404 kt
1n ft=s ¼ 0:5921 kt
1n m=s ¼ 1:944 kt

True Airspeed, Speed of Sound, and Mach Number

The speed of the airplane relative to the air mass is called the true air-
speed (TAS). The symbol used in equations is V, and the usual units of air-
speed in aviation are knots, represented as KTAS for knots true airspeed.
The speed of sound is the speed at which disturbances travel in the
medium they were created in. The speed of sound for air can be determined
as follows (assuming isentropic flow):
rffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffi
gp
a¼ ¼ gRT (3:1)
r
or, in the format used most throughout this book,
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
a ¼ gRu TSL (3:2)
 CHAPTER 3 Measuring Airspeed 41

where g is the ratio of specific heats and is equal to 1.4 for air, as described in
Chapter 1. From this, it is interesting to note that the speed of sound in air
depends essentially only on air temperature; because we have a model for
the variation of air temperature with altitude, we can also model the variation
of the speed of sound with altitude. For example, the speed of sound under
standard day sea level conditions would be

aSL ¼ 1116:4 ft=s ¼ 761:2 mph ¼ 661:4 kt ¼ 1225 km=h

The speed of sound is a reference value for atmospheric flights. For low
airplane airspeeds with respect to the speed of sound, one can neglect the
effects of compressibility on airplane drag, lift, and thrust. As the airplane air-
speed approaches the speed of sound, compressibility effects are more appar-
ent, and small shock waves will develop on some parts of the airplane. When
the airplane airspeed reaches the speed of sound (V ¼ a), any disturbances
created by the airplane will pile up in front of the source to form a normal
shock wave. When the airspeed exceeds the speed
of sound (V . a), the entire airplane flies within a Note: Capt. Charles
cone of disturbance, as illustrated in Fig. 3.1. E. Yeager was the first to
exceed the speed of sound in
The Mach number, first used in 1929 in honor level flight. He achieved this
of Ernst Mach (1838– 1916) for his contribution feat on 14 Oct. 1947 while he
to the study of supersonic flow, is the ratio of the was flying a Bell X-1 (see
airplane true airspeed to the speed of sound. Fig. 3.2).

V V V
M¼ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ pffiffiffi (3:3)
a g R u TSL aSL u

Alternatively, the true airspeed is a function of the Mach number and the
temperature at altitude. Equation (3.3) can then be written as
pffiffiffi
V ¼ M aSL u (3:4)

V<a V=a V>a

Fig. 3.1 Disturbances (circles) generated by the nose of the airplane at regular intervals [1].
42 Operational Aircraft Performance and Flight Test Practices

Fig. 3.2 Bell X-1. Source: National Air and Space Museum.

The usual units for true airspeed are knots (KTAS), and the usual units of
temperature in aviation are degrees Celsius. This allows us to rewrite Eq. (3.4)
as follows:

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
KTAS ¼ 38:97 M 273:15 ½K þ SAT½8C (3:5)

The Mach number, in addition to the Reynolds number, is a governing

parameter in airplane aerodynamics. Usually, the flow is considered incom-
pressible for Mach numbers lower than approximately 0.3. For Mach
numbers greater than 0.3, each case must be investigated individually to
determine the effects of compressibility. For supersonic flights, the cone of
the oblique shock wave (Fig. 3.3) has the following angle, with respect to

Fig. 3.3 Shock waves as seen by Schlieren photography. Source: NASA.

 CHAPTER 3 Measuring Airspeed 43

its axis of symmetry:

 
1
m ¼ arcsin (3:6)
M

Thus, at Mach one, the angle is 90 deg (normal shock wave); at Mach
numbers greater than one, the angle is less than 90 deg (at Mach 2,
m ¼ 30 deg).

Example 3.1
During wind tunnel testing of a model under supersonic flow, the picture
shown in Fig. 3.4 was taken. Based on the shock cone, estimate the test
Mach number.

Solution: The cone angle as measured from the photo is 43 deg. Using
Eq. (3.6),
1
M¼ ¼ 1:466
sinðmÞ

Fig. 3.4 Wind tunnel test photo. Source: NASA.

44 Operational Aircraft Performance and Flight Test Practices

TAS vs M: Altitude Effects

For a fixed Mach number, an increase in altitude will result in a decrease
in true airspeed in the troposphere and a constant airspeed in the strato-
sphere for a standard atmosphere. This is illustrated in Fig. 3.5.
The equations for KTAS vs altitude can be derived by combining
Eqs. (3.5) and (1.9). In the troposphere of a standard atmosphere, one can
determine the true airspeed for a known Mach number as
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
KTAS ¼ 38:97 M 288:15½K þ aLR H (3:7)
where
aLR ¼ – 0.0019812 K/ft, the temperature lapse rate
H ¼ altitude in ft
In the stratosphere of a standard atmosphere, the true airspeed in knots is
KTAS ¼ 573:6 M (3:8)

TAS vs M: Temperature Effects

When the conditions are not standard, the previous equations will need
to be adjusted. Although the temperature can vary greatly at sea level, by the
time an airplane reaches about 20,000 ft, the temperature will pretty much be
within 208C of standard. If you express the relationship between Mach

70,000
100 KTAS 200 KTAS 300 KTAS 400 KTAS 500 KTAS

60,000

50,000
Altitude (ft)

40,000

30,000

20,000

10,000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Fig. 3.5 True airspeed vs altitude and Mach in a standard atmosphere.

 CHAPTER 3 Measuring Airspeed 45

number and true airspeed as a function of the temperature deviation from

standard DISA, one gets in the troposphere

V
M ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
   ffi (3:9)
1:8 g R TSL 1  6:875  106 H þ DISA

and in the stratosphere

V
M ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (3:10)
1:8 g R ð216:65 ½K  þ DISAÞ

where
V ¼ true airspeed (ft/s)
H ¼ pressure altitude (ft)
TSL ¼ 288.15 K
g ¼ 1.4
DISA ¼ deviation from standard (8C)
R ¼ 1716.5 (ft lbf/slug 8R)
From Eqs. (3.9) and (3.10), it can be seen that a warmer than ISA condition
will result in a lower Mach number for a given true airspeed. This is shown in
Fig. 3.6.

70,000
400 KTAS 500 KTAS
ISA + 20 ºC
ISA + 20 ºC

60,000
ISA
ISA

50,000
Altitude (ft)

40,000

30,000

20,000

10,000

0
0.5 0.6 0.7 0.8 0.9
Mach

Fig. 3.6 Temperature effect on Mach number for given true airspeed.
46 Operational Aircraft Performance and Flight Test Practices

Example 3.2
An airplane flying at Mach 0.80 at 39,000 ft in a standard atmosphere encoun-
ters a front where the temperature increases by 108C. If the airplane remains
at a pressure altitude of 39,000 ft and at a Mach number of 0.80, what will be
the impact to the true airspeed?
Solution: At 39,000 ft in a standard atmosphere, a Mach number of 0.80
corresponds to a true airspeed of
KTAS ¼ 573:6 M ¼ 458:9 kts
This is the airspeed under standard temperature conditions in the strato-
sphere [from Eq. (3.8)]. Under standard conditions, the temperature in the
stratosphere would be – 56.58C. Now, for the problem at hand, the tempera-
ture is actually 108C warmer, or –46.58C. Then, the true airspeed under these
conditions would be
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V ¼ M 1:8 g Rð216:65 K þ DISAÞ ¼ 792:13 ft=s ¼ 469:3 kt
Therefore, the higher temperature resulted in a higher true airspeed
(10.4 kt) for the same Mach number.

Impact Pressure vs Dynamic V2 = 0

V1
Pressure
The impact pressure qc rep- p1 p2
resents the rise in pressure due
to a fluid in movement. The com-
bination of this pressure and the Fig. 3.7 Slowdown of the airflow.
ambient static pressure ps is
equal to the total pressure pT of the fluid. This was defined by Eq. (1.1)
and is rewritten here for convenience:
pT ¼ ps þ qc (3:11)
For a compressible fluid such as air, the total pressure of the air can be
determined by slowing the airflow down isentropically from the freestream
airspeed (condition 1 of Fig. 3.7) to a complete stop (condition 2).
For an adiabatic flow, the following equation applies:
V2
cp T þ ¼ constant (3:12)
2
where

gR
cp ¼ (3:13)
g1
 CHAPTER 3 Measuring Airspeed 47

Combining Eqs. (3.12) and (3.13), and the equation of state in Eq. (1.5),

g p1 V12 g p2 V22 g p2
þ ¼ þ ¼ þ0
g  1 r1 2 g  1 r2 2 g  1 r2

g p2 p1 g 1 r1
1
2 V12 ¼  ¼ p2  p1
g  1 r2 r1 g  1 r1 r2
If the fluid is slowed down isentropically, the following relationship holds:
p
¼ constant (3:14)
rg
Therefore,
2 3
 1
1 V2 ¼ g 1 4 p1 g
2 1 p2  p1 5
g  1 r1 p2

Condition 1 represents the free stream conditions (static air pressure p,

air density r, and true airspeed V ); condition 2 is the stagnation point
(zero airspeed) where the measured air pressure is the total air pressure
(pT or p0 ). The free stream true airspeed (V or TAS) is finally,
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 pT g
V ¼t  15 (3:15)
g1r p

The impact pressure can be extracted from Eq. (3.15) by combining it

with Eq. (3.11). This gives
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 pT  p g
V ¼t þ1  15
g1r p
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 qc g
¼t þ1  15 (3:16)
g1r p

Then,
2 3
  g
g  1 r g  1
qc ¼ p 4 1 þ V2  15 (3:17)
2 gp

Within Eq. (3.17) we can recognize the square of the speed of sound
[see Eq. (3.1)], dividing the square of the true airspeed (i.e., the square of
the Mach number). Then the impact pressure can be written in terms of
48 Operational Aircraft Performance and Flight Test Practices

Mach number as follows:

2 3
  g
g1 2 g 1
qc ¼ p4 1 þ M  15 (3:18)
2

Dynamic pressure q is defined as one half the air density times the air-
speed squared.
1
qW r V2 (3:19)
2
It represents the kinetic energy of the airflow. It can also be written in
terms of Mach number as follows:
1
q¼ g p M2 (3:20)
2
It is interesting to compare the dynamic pressure to the impact pressure.
Dividing Eq. (3.20) by Eq. (3.18), we get
1
q gM 2
2
¼  2 3 (3:21)
qc  g
4 1 þ g  1 M2
g 1
 15
2

Figure 3.8 shows the results of this equation. It can be seen that the
dynamic pressure underestimates the pressure rise due to airspeed. It can
also be noticed that below a Mach number of approximately 0.3, the differ-
ence between the dynamic pressure and the impact pressure is of the order
of 2.3% or less.

Total Pressure
How do you measure true airspeed? By investigation of Eq. (3.16), it can
be seen that one must measure the total pressure, the static pressure, and the
static air temperature. [The air density is computed from measured static air
pressure and air temperature using the perfect gas; see Eq. (1.5), Chapter 1.]
The static air pressure is measured from a static pressure port, as
described in Chapter 2, and it is also used for the pressure altitude compu-
tation. This tube is aligned with the flow in such a way that the pressure
ports are perpendicular to the flow.
To measure the total pressure, one must slow down the airflow to zero
airspeed at the point of measurement of the pressure. The instrument that
is widely used on airplanes to measure the total pressure is the pitot tube,
named after Henri Pitot who first used it to measure the flow velocity of
the Seine River in Paris in 1732. A pitot tube is basically a tube aligned
with the airflow, outside the airplane boundary layer, with the open-ended
 CHAPTER 3 Measuring Airspeed 49

0.95

0.9
q/qc

0.85

0.8

0.75
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Fig. 3.8 Ratio of dynamic to impact pressure.

side facing the airflow and the other side connected to a pressure-sensing
device such that there is normally no mass flow inside the system (i.e., the
speed of the air is zero inside the system); see Fig. 3.9.

Fig. 3.9 Pitot-static tubes. Source: Goodrich Sensors—Rosemount Aerospace.

50 Operational Aircraft Performance and Flight Test Practices

The impact pressure is extracted from the total pressure by removing the
static pressure, written as
qc ¼ pT  p (3:22)
Now, to be able to compute a true airspeed, we not only need to measure
total and static pressures, we also need the static air temperature.

Total Temperature and Mach Number

The static air temperature is usually not measured directly for an
airplane in flight; rather, a sensor is usually inserted into the moving
flow, away from the boundary layer. The air around this sensor will
usually be slowed down adiabatically to an almost complete stop relative
to the airplane. Some air movement is required to ensure a proper
change of air mass inside the probe to prevent heat accumulation and
false temperature reading. The temperature then measured is the stagnation
air temperature or total air temperature (TAT). Figure 3.10 shows typical
TAT probes.
During an adiabatic slowdown, the air temperature rises from the
ambient static air temperature (SAT) T to the total air temperature (TAT)
TT . From Eqs. (3.12) and (3.13), one can derive the following relationship
between TAT and SAT:
g1 2
TT ¼ T 1 þ M (3:23)
2
The SAT can then be determined from the measured TAT if the
Mach number is known. From Eq. (3.3), the Mach number is equal to
the airspeed divided by the local speed of sound. Combining Eqs. (3.1) and

Fig. 3.10 TAT probe. Source: Goodrich Sensors—Rosemount Aerospace.

 CHAPTER 3 Measuring Airspeed 51

(3.15) into (3.3), one gets

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 pT g
t  15 v u
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
g1r p u  g1
u 2 4 pT g
M¼ rffiffiffiffiffiffiffi ¼t  15
gp g1 p
r
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3ffi
u  g1
u
u 2 4 pT  p g
¼t þ1  15 (3:24)
g1 p

Therefore, to measure the Mach number, a

static pressure port and a pitot tube are required, Note: One must know the
but temperature is not. static air temperature to
determine the true airspeed
The resulting equation to compute SAT is then
from air data, but one does
(see also Fig. 3.11) not need to know the
temperature to determine
TT the Mach number. The Mach
T¼ (3:25)
g1 2 number can be computed
1þ M from the pressures obtained
2 by a pitot-static probe.

300

280
M = 0.0
M = 0.1
M = 0.2
260
M = 0.3
M = 0.4
SAT (K)

240

M = 0.5
220 M = 0.6
M = 0.7
M = 0.8
200 M = 0.9

180
220 230 240 250 260 270 280 290 300

TAT (K)

Fig. 3.11 SAT as a function of TAT and Mach.

52 Operational Aircraft Performance and Flight Test Practices

50

45
M = 0.9
40

35 M = 0.8
Temperature rise (K)

30

25 M = 0.7

20 M = 0.6
15
M = 0.5
10
M = 0.4
5 M = 0.3
M = 0.2
0 M = 0.1
220 230 240 250 260 270 280 290 300
SAT (K)

Fig. 3.12 Temperature rise due to airspeed vs SAT.

So the measured temperature, TAT, will always

be larger than the static air temperature for a flying Note: Use only absolute
airplane. For very low speeds, the difference will temperatures [K or 8R] when
be small. converting from TAT to
SAT.
Of interest, from Eq. (3.23), the temperature rise
DT (total temperature minus static temperature) due
to the airspeed is equal to (see Fig. 3.12)
g1
DT ¼ T M2 (3:26)
2
The temperature rise for airspeeds of less than 0.1 Mach are less than 18C,
whereas the temperature rise for typical airliner cruising speeds (Mach 0.75
to 0.85) is of the order of 258C to 358C.
All variables are now available to compute the TAS using Eqs. (3.15)
or (3.16).

Equivalent Airspeed
To remove the need to account for temperature, an equivalent airspeed
(EAS or VE ) is defined by setting the air density value in Eq. (3.15) or Eq.
(3.16) to standard sea level air density. Using Eq. (3.16), the EAS takes on
 CHAPTER 3 Measuring Airspeed 53

the following form:

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 pT  p g
VE ¼ t þ1  15
g  1 rSL p
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3ffi
u  g1
u
u 2g p 4 qc g
¼t þ1  15 (3:27)
g  1 rSL p

With this form, it can be noted that only the static pressure p and total
pressure pT remain as variables; therefore, only a pitot-static probe or the
equivalent pitot probe and static port combination are required to
compute the equivalent airspeed. If the EAS is expressed in terms of knots,
it is often labeled KEAS for knots equivalent airspeed.

TAS vs EAS
Comparing Eq. (3.16) to Eq. (3.27), one will note that the only difference
is with using a fixed value of air density for Eq. (3.27). The ratio of EAS to
TAS gives
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g p 4 pT  p g
t þ1  15
g  1 rSL p rffiffiffiffiffiffiffi
VE r pffiffiffi
¼ vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3 ¼ ¼ s (3:28)
V u  g1 rSL
u
u 2g p 4 pT  p g
t þ1  15
g1r p

or
pffiffiffi
VE ¼ sV (3:29)

From this relationship, one can see that flying at

a constant TAS will result in a reduction in EAS as Note: The variation of EAS
the altitude increases in a standard atmosphere. vs TAS with altitude is not an
exact relation, but rather
Graphically, this can be show as in Fig. 3.13.
depends on the variation of
the density ratio with
altitude. In a standard
Calibrated Airspeed atmosphere, the relation is as
shown in Fig. 3.13.
Let’s further simplify the airspeed equation by
setting the static pressure in Eq. (3.27) to the
54 Operational Aircraft Performance and Flight Test Practices

70,000
KTAS
100 150 200 250 300 350

60,000
Standard atmosphere
50,000
Pressure altitude (ft)

40,000

30,000

20,000

10,000

0
0 50 100 150 200 250 300 350
KEAS

Fig. 3.13 KTAS vs KEAS variation as a function of altitude.

Example 3.3
Compare the variation of KEAS with altitude from sea level to 20,000 ft
pressure altitude for an airplane flying at a constant 150 KTAS for the follow-
ing two cases:
1. Standard atmosphere.
2. The temperature at sea level is standard, but remains constant to the sea
level value until 4000 ft pressure altitude. From that point on, the
temperature lapse rate is standard up to 20,000 ft.
Solution: First we must establish the temperature profile as a function of
pressure altitude. From the problem statement, the speed profile to be estab-
lished is a function of pressure altitude; therefore, the variation of pressure vs
altitude is established and is equal to the standard atmosphere variation. We
can use Eq. (1.11) of Chapter 1.
Next, we know that for both cases, the temperature at sea level is standard,
so its value is 288.15 K. For the first question, we have a standard temperature
lapse rate of approximately 28C per 1000 ft pressure altitude, and the tempera-
ture ratio varies per Eq. (1.9). This is plotted as the lower line on the left chart

(Continued)
 CHAPTER 3 Measuring Airspeed 55

Example 3.3 (Continued)

20,000
Te
18,000 m
pe
ra
16,000 tu
re
pr
14,000 ofi
le
Pressure altitude (ft)

pe
St rE
12,000 an xa
da m
rd pl
10,000 at e
m 3.
os 3,
8000 ph par
er t2
e
6000

4000

2000

0
0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.00 1.02
Temperature ratio
20,000

18,000
KTAS = 150
16,000
Te
m St
14,000 pe an
Pressure altitude (ft)

ra da
tu rd
re at
12,000 pr
ofi m
le os
10,000 pe ph
rE er
e
xa
8000 m
pl
e3
6000 .3,
pa
rt
4000 2

2000

0
100 105 110 115 120 125 130 135 140 145 150
KEAS

Fig. 3.14 Temperature ratio and KEAS variation as a function of altitude, Example 3.3.

of Fig. 3.14. For the second question, the temperature remains constant from
sea level to 4000 ft pressure altitude and then varies from that value following
the standard atmosphere temperature lapse rate. An equation that represents

(Continued)
56 Operational Aircraft Performance and Flight Test Practices

Example 3.3 (Continued)

this variation takes on the form of
8
< u ¼ 1:0 for Hp  4000 ft
aLR  
:u ¼ 1 þ Hp  4000 for Hp . 4000 ft
TSL

This is represented by the upper line on the left chart of Fig. 3.14.
Now, with the temperature and pressure profiles vs altitude established, one
can compute the density ratio as a function of altitude using the following
relation:
d
s¼
u
Finally, the equivalent airspeed can be computed from Eq. (3.29). The
chart on the right of Fig. 3.14 shows the variation of KEAS vs altitude for
the two test conditions. Note that the second case, having a larger than stan-
dard temperature at all altitudes above sea level, has a resulting lower KEAS
for a given altitude (and fixed KTAS) as compared to the standard atmosphere
condition due to the lower air density than standard.
Of further interest in the charts of Fig. 3.14, note how the isothermal
section between sea level and 4000 ft (part 2 of the problem) results in a
faster decrease in KEAS vs altitude than under standard atmospheric con-
ditions. This is due to the air density, which is decreasing faster than the stan-
dard atmospheric condition because the temperature was warmer than
standard. (Isothermal, in this case, is actually equivalent to an increase in
the DISA for a given pressure altitude.) You can note a similar impact in
Fig. 3.13 when the temperature lapse rate of the troposphere gives way to
the isothermal conditions in the stratosphere at 36,089 ft; notice the distinct
change in KEAS slope with altitude.
Consider the following: the definition of the dynamic pressure is, from
Eq. (3.19)
1
q ¼ rSL s V 2
2
Combining this with Eq. (3.29) for the relationship between VE and
V, we get

 2
1 V 1
q ¼ rSL s pEffiffiffi ¼ rSL VE2 (3:30)
2 s 2

Final note on KEAS: KEAS is equal to KTAS only when the airplane is
flying under sea level density altitude.
 CHAPTER 3 Measuring Airspeed 57

standard sea level value so that the only variable left

is the impact pressure qc . Doing this, one gets Note: Flying at a constant
equivalent airspeed is the
same as flying at a constant
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3 dynamic pressure.
u  g1
u
u 2g pSL 4 qc g
VC ¼ t þ1  15 (3:31)
g  1 rSL pSL

This is called the calibrated airspeed (CAS) or knots calibrated airspeed

(KCAS) VC . Now, observing Eq. (3.31), note that
the only variable left in the equation is the impact Note: Flying at a constant
pressure qc . calibrated airspeed is the
To measure the calibrated airspeed, one only same as flying at a constant
impact pressure.
needs to measure total and static pressure to get
the impact pressure.

CAS vs Mach
In this section, we compare the calibrated airspeed to the Mach number.
To do so, we examine the impact pressure as defined by Eqs. (3.17) and (3.18).
In Eq. (3.17), we replace the true airspeed V with the calibrated airspeed Vc ,
and set the values of pressure and air density to their standard sea level
values. We then get
2 3 2 3
  g   g
g  1 r g 1 g  1 g 1
pSL 4 1 þ SL
V2  15 ¼ qc ¼ pSL d4 1 þ M2  15
2 gpSL C 2

(3:32)
or
2 3 2 3
  g   g
14 g  1 rSL 2 g1 g  1 2 g1
1þ V  15 ¼ 4 1 þ M  15 (3:33)
d 2 gpSL C 2

From this, we can see that for a given calibrated

airspeed, the Mach number will increase as the Note: The relationship
altitude increases, as noted by the ratio (1/d). between Mach and KCAS vs
Graphically, the relationship between CAS and altitude is exact; for example,
100 KCAS at 30,000 ft
Mach, as a function of pressure altitude, would be
pressure altitude will always
as shown in Fig. 3.15. be equal to Mach 0.2756.

CAS vs EAS
This section compares the calibrated airspeed, Eq. (3.31), to the
equivalent airspeed, Eq. (3.27). Looking at both equations side by side,
notice that the CAS is a function of qc only and the EAS is a function of qc
58 Operational Aircraft Performance and Flight Test Practices

65,000
60,000
55,000
50,000
45,000 KCAS
100
Pressure altitude (ft)

40,000
150
35,000
200
30,000 250
25,000 300
350
20,000
15,000
10,000
5000
0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Mach

Fig. 3.15 Mach vs altitude for fixed KCAS.

and p (i.e., a function of q, altitude effects).

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u  g1
u
u 2g pSL 4 qc g
VC ¼ t þ1  15
g  1 rSL pSL
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3ffi
u  g1
u
u 2g p 4 qc g
VE ¼ t þ1  15
g  1 rSL p

Under standard sea level conditions, both airspeeds are equal. As the
altitude increases above sea level standard with the resulting decrease in
static pressure, there will be an increasing difference between the two
speeds. This difference will also be amplified by increasing values of airspeed.
The ratio F of EAS to CAS gives
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 3
u
u   ðg1Þ
u p 6 pT p g 7
u 4 1þ 15
u r p
u SL
u EAS
F¼u 2 3¼ (3:34)
u    ðg 1Þ CAS
u p p
u SL 6 1 þ T p g
15
7
t 4 pSL
rSL
 CHAPTER 3 Measuring Airspeed 59

1.00 0
0.2

0.4 500
0.98 0 ft

10
,00
0.96 0f
0.6 t
15
F = KEAS/KCAS

,00
0f
20 t
0.94 ,00
Mach 0f
25 t
0.8 ,00
0f
0.92 t

30
,00
35
Pressure altitude

0f
,00

t
40

0f
,00
0.90
45,

t
50, 0 ft

0
000

1.0
55, 00 ft

ft
000
60,0
00

ft
ft

0.88
0 100 200 300 400 500 600 700
KCAS

Fig. 3.16 Ratio F as a function of KCAS.

This ratio can be plotted as a function of KCAS with lines of equal Mach
and altitude to give the results shown in Fig. 3.16.
From this, one can see that for pressure altitudes larger than sea
level, the equivalent airspeed will be smaller than the calibrated altitude,
which is as expected because flying at a constant EAS is the same as
flying at a constant dynamic pressure q, and flying at a constant CAS
is the same as flying at a constant impact pressure qc , and q is always
smaller than qc .
As well, the difference between EAS and CAS is less than 0.5%
(F . 0.995) if the Mach number is less than 0.2 for altitudes varying from
sea level to 60,000 ft.
Another way to look at the difference between Tip: When examining flight
EAS and CAS is just that, a difference. data for slow flight
conditions such as stall
(Mach typically less than
DVC ¼ EAS  CAS (3:35) 0.20), using KEAS instead of
KCAS often simplifies the
In a chart form, we get what’s shown in Fig. 3.17. analysis; the error introduced
The difference between the two speeds for by this assumption is well
within the expected scatter of
Mach 0.2 or lower is less than 0.25 kt. Note again the data collected.
that EAS is smaller than CAS, as shown by the
DVc that is always negative, that is, for pressure alti-
tudes higher than standard sea level.
60 Operational Aircraft Performance and Flight Test Practices

0 0
0.4
–5 0.6

–10 500
0.8 0 ft
Mach
–15
⌬Vc

10
,00
–20 1.0 0f
60,000 t
55,000

15
–25

,00
50,000
ft

0f
45,00

20

t
ft

–30 40,00

,00
25,
35,
30,
ft
0 ft

0f
000
000
000

t
0
Pressure altitude
ft

ft
–35
ft
0 100 200 300 400 ft 500 600 700
KCAS

Fig. 3.17 Difference (DVc) between EAS and CAS.

CAS vs TAS
Finally, we now compare the variation of TAS as a function of CAS.
To do so, we use a combination of EAS to CAS conversion and EAS to
TAS conversion. The resulting equation takes on the following form:
F
TAS ¼ pffiffiffi CAS (3:36)
s
where F is defined by Eq. (3.34) and sigma is the air density ratio. In general,
for a constant CAS, TAS will increase with altitude. There will be some
impact due to the temperature profile as the altitude changes. This can be
seen form the chart in Fig. 3.18.

Ground and Vertical Speeds

The speed of the airplane relative to the ground is the ground speed (GS),
which is a two-dimensional value used for navigational purposes. It differs
from the true airspeed, which is a three-dimensional value (i.e., along the
flight path of the airplane), depending on the climb (or descent) angle (g,
not to be confused with the ratio of specific heats) as well as on the wind
speed and direction (wind velocity VW ), as seen in Fig. 3.19.
RC ¼ V sinðgÞ
(3:37)
VH ¼ V cosðgÞ
 CHAPTER 3 Measuring Airspeed 61

2.3
2.2
2.1
2
ISA – 10C
1.9
1.8 ISA
1.7
1/␴ 0.5

1.6
1.5 ISA + 10C
1.4
1.3
1.2
1.1
1
0.9
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000
Pressure altitude (ft)

Fig. 3.18 KTAS vs KCAS is a function of the air density ratio.

Wind velocity can help increase the range of the airplane if it is blowing
from the rear quadrants, with the maximum effects for tail winds aligned with
the horizontal speed vector. On the other hand, it can also decrease the range
sharply if the wind comes from any of the two front quadrants. (Range is dis-
cussed further in Chapters 10 and 11.) The ground speed will equal the TAS
for level flight and zero wind conditions only.

Instruments: Airspeed Indicators

In this section, we want to provide a few words on airspeed measuring
devices. All airspeed indicators (ASIs) require at least a source of static
pressure and a source of total pressure. Modern airliners use air data compu-
ters to compute the calibrated and true airspeed as well as the Mach number
from the equations developed earlier.

VW

VH
TAS
VV GS
␥

VH

Fig. 3.19 Ground and vertical speeds.

62 Operational Aircraft Performance and Flight Test Practices

Measuring KCAS
The simplest ASI is designed to measure KCAS, because it requires only
that the impact pressure be known. Figure 3.20 shows how the total pressure
is fed to a sealed disk and the static pressure is fed to an airtight casing around
the disk. As the total pressure increases due to airspeed, the walls of the disk
will move proportional to the difference between the total pressure and the
static pressure (i.e., qc ). This wall movement is picked up by a gearing
system that moves a needle on the ASI face. The gearing for this ASI is cali-
brated using Eq. (3.31).
As with any engineering discipline, assumptions are used to develop
mathematical models. On early airplanes that flew low and slow, one could
essentially assume that the air was incompressible; that is, there was constant
air density. Under such conditions, one can use Bernoulli’s equation as a
mathematical model. This model provides a variation of pressure vs airspeed
along a streamline.

1
p þ r V 2 ¼ constant (3:38)
2

Now, with a further simplification of assuming standard sea level air

density and that the constant is the total air pressure measured by the
pitot, one gets this version for CAS (see Fig. 3.21):

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2ð pT  pÞ
CAS ¼ (3:39)
rSL

It can be seen that the assumption of incompressible flow results in an

overprediction of the KCAS. Below 200 kt, the error is of the order of 2 kt
or less; at 150 kt, the error is 1 kt. So for lower speed airplanes, this type of
assumption to calibrate an ASI provides sufficient accuracy.

V pT ASI
Gear

pT

V p p

Fig. 3.20 Airspeed indicator (ASI).

 CHAPTER 3 Measuring Airspeed 63

500
Overprediction of airspeed
450

400

350

300
KCAS

250

200

150

100

50

0
0 50 100 150 200 250 300 350 400 450 500 550 600
KCAS (incompressible assumption)

Fig. 3.21 KCAS assuming incompressible flow.

Measuring Mach
A mechanical Machmeter must be calibrated against Eq. (3.24). The
gearing of a Machmeter is somewhat more complicated in that the total
pressure and static pressure must be combined with gears similar to a
KCAS ASI, but then corrected for altitude effects. A Machmeter is a combi-
nation of an altimeter and an airspeed indicator (Fig. 3.22).

Measuring KTAS
An ASI that measures KTAS must be calibrated against Eq. (3.15). Such
ASI adds an additional level of complexity over the Machmeter in that we
must now integrate the temperature reading (see Fig. 3.23).

Aneroid capsules
V for altitude correction
pT
Gear

pT

V p
p

Fig. 3.22 Machmeter.

64 Operational Aircraft Performance and Flight Test Practices

V
TAT

Aneroid capsules
V ASI for altitude correction
pT
Gears

pT

V p
p

Fig. 3.23 KTAS ASI.

No matter which airspeed is needed, they all rely on measuring the proper
total pressure and static pressure. Certification requirements for Part 25,
Airworthiness Standards: Transport Category Airplanes, dictate:
§25.1323 Airspeed Indicating System

(a) Each airspeed indicating instrument must be approved and must be

calibrated to indicate true airspeed (at sea level with a standard
atmosphere) with a minimum practicable instrument calibration error
when the corresponding pitot and static pressures are applied.
...
(h) Each system must be arranged, so far as practicable, to prevent
malfunction or serious error due to the entry of moisture, dirt, or
other substances.
(i) Each system must have a heated pitot tube or an equivalent means of
preventing malfunction due to icing.
(j) Where duplicate airspeed indicators are required, their respective pitot
tubes must be far enough apart to avoid damage to both tubes in a
collision with a bird.

Consider the impact of losing one pressure source with the following
example.

Example 3.4
Consider the following NTSB case, number NYC75AN070, 1 Dec. 1974. A
Boeing 727 –251 is cleared to climb to a cruise altitude of FL310. During
the climb, the crew encounters turbulent rainy weather. Passing through
16,000 ft pressure altitude at an airspeed of 305 KCAS and a rate of climb
(ROC) of 2500 fpm, all indications to the crew were normal and as expected
from Boeing’s flight manual for the airplane’s weight. From this point on,

(Continued)
 CHAPTER 3 Measuring Airspeed 65

Example 3.4 (Continued)

Altitude
Airspeed Vertical
speed
Pitot Airspeed indicator
heater switch indicator (VSI) Altimeter
Static port

Pitot Drain
tube opening
Pressure
chamber
Attitude
Vertical speed

Fig. 3.24 B727 pilot’s display and pitot-static system.

with no change in power setting, the airplane started accelerating. The crew
reacted by applying back pressure to the elevator to increase the pitch attitude
and maintain speed. By FL230, the ROC was more than 6500 fpm, and the
indicated airspeed was 405, KCAS, triggering the overspeed warning. This
was soon followed by the stall warning and an undercarriage warning indicat-
ing one of the crew had reduced the thrust to idle. Then the airplane stalled
and the crew lost control, and the plane eventually crashed. At the time of
the stall, the indicated KCAS was 420 kt and the indicated altitude 24,800 ft.
What happened? Elaborate.
Solution: The B727–251 is equipped with an independent pilot and copilot
air data system. Each system has a pitot probe on the nose and a static port
along the fuselage. These pressure sources are connected to mechanical indi-
cators in the cockpit (see Fig. 3.24).
So what do we know about the pressure sources? We know the static port
feeds the altimeter, the vertical speed indicator (VSI), and the airspeed indi-
cator (ASI). We also know the pitot tube feeds only the ASI. We know the alti-
tude continued to change with time during this event, indicating the static
ports were still recording some pressure change. This static pressure change
would most likely affect all three indicating systems equally. Let’s look at
the evolution of the static pressure and total pressure through the recorded
altitude and airspeed values (Table 3.1).

Table 3.1 Recorded Altitude and Airspeed Values

Altitude (ft) p (psf) Airspeed (KCAS) qc (psf) pt (psf)

16,000 1147.0 305 332.0 1479.0
23,000 856.4 405 609.3 1465.7
24,800 792.2 420 659.8 1452.1

(Continued)
66 Operational Aircraft Performance and Flight Test Practices

Example 3.4 (Continued)

It is interesting to note that the total pressure during the climb remained
essentially constant. This pressure is measured by the pitot tube and may indi-
cate that the tube became blocked during the climb. This finding was partially
substantiated by the NTSB inquiry, which found two of the five pitot heads on
the airplane contained water or ice at the time of the investigation; the others
were too badly damaged to provide useful information.

Exercises
1. You are flying at a constant Mach number of 0.80 at FL370. Along the way,
you encounter a front that results in a temperature increase of 58C. If you
maintain the Mach number at 0.80 and there is no change in the wind
component, will it take more or less time (or no change) to reach your
destination? Explain.
2. You are flying at 33,000 ft at a calibrated airspeed of 280 KCAS under ISA
–28C conditions and a headwind of 15 kt. A Boeing 737–800 is flying
directly over you at FL370 at Mach 0.76 under ISA þ18C and tailwind of
5 kt. Which airplane has the greatest ground speed?
1
3. The dynamic pressure is defined as q W r V 2 . What are the steps
2
required to transform this equation to be a function of Mach number?

Reference
[1] Asselin, M., An Introduction to Aircraft Performance, AIAA Education Series, AIAA,
Reston, VA, 1997.
Chapter 4 Altimetry Error

Chapter Objective
This chapter introduces the reader to the first dedicated testing that will define
the airplane’s performance models. We start with some very basic questions:
Are the airspeed and altitude recorded true? If not, what is an acceptable
error? What is the accuracy of the data recording system? How will this
error impact the creation of performance models? How will the error
impact the safety of the flight? How do you present such information to the
flight crew? Being able to answer these questions will help define a better per-
formance model for the airplane.

Indicated Airspeed/Altitude

T
he altitude provided to the flight crew is based on the measured
pressure altitude H properly corrected to the barometric condition
for the airspace. The airspeed is based on the calibrated
airspeed (CAS).
The main advantage of CAS over true airspeed (TAS), as will become
clearer as this book progresses, is that the calibrated airspeed pilots need
to follow for takeoff, landing, or any other flight conditions is independent
of atmospheric conditions and often of altitude as well, whereas TAS is
dependent.
Unfortunately, the airplane is not equipped with a perfect air data system,
so some degree of error will exist when measuring the static pressure, total
pressure, and total temperature. The indicated airspeed (IAS) is defined as
the calibrated airspeed minus pressure reading errors [also called position
errors (PEs)] and instrument errors.
IAS ¼ CAS  DVPE  DVi ð4:1Þ
The indicated altitude Hi is defined as the pressure altitude minus the
pressure reading errors (position errors) and instrument errors.
Hi ¼ H  DHPE  DHi ð4:2Þ

67
68 Operational Aircraft Performance and Flight Test Practices

Certification Regulations
The accuracy required out of an air data system depends on its intended
use, which will then dictate the level of design and calibration testing
required. One of the main reasons for the accuracy of the air data system
is to ensure separation between planes, where true airspeed and ground
speed (wind effect) are of great importance. If we take a look at the early
version of what is today’s Federal Aviation Regulations (FARs), we see that
in CAR 4b (1950s), the requirements of §4b.612 are as follows:
Flight and navigational instruments? (a) Air-speed indicating systems.
(1) Air-speed indicating instruments shall be of an approved type and
shall be calibrated to indicate true air speed at sea level in the standard
atmosphere with a minimum practicable instrument calibration error
when the corresponding pitot and static pressures are applied to the
instrument.

This was rolled into paragraph 1323 when Part 23 and Part 25 were created.
Paragraph 1323 of FAA 14 CFR/EASA CS/ TCCA CAR Parts 23 and 25
requires the following (see Fig. 4.1):

(a) Each airspeed indicating instrument must be approved and must be

calibrated to indicate true airspeed (at sea level with a standard
atmosphere) with a minimum practicable instrument calibration error
when the corresponding Pitot and static pressures are applied.
...
(c) The airspeed error of the installation, excluding the airspeed indicator
instrument calibration error, may not exceed three percent or five
knots, whichever is greater, throughout the speed range, from -
(1) VMO to 1.23 VSR1 , with flaps retracted; and
(2) 1.23 VSR0 to VFE with flaps in the landing position.

15

10
Unacceptable error
Airspeed error (kt)

5
Acceptable error boundary
0 for 14 CFR/EASA/CAR
certification
–5
Unacceptable error
–10

–15
0 50 100 150 200 250 300 350 400
Airspeed (kt)

Fig. 4.1 Maximum indicated airspeed error.

 CHAPTER 4 Altimetry Error 69

120
100
Unacceptable error
80
60
Altitude error (ft)

40
20 Acceptable error boundary
0 for 14 CFR/EASA/CAR
–20 certification
–40
–60
–80
Unacceptable error
–100
–120
0 50 100 150 200 250 300 350 400
Airspeed (kt)

Fig. 4.2 Maximum indicated altitude error.

For the static pressure system (feeding the altimeter and airspeed indicator),
the maximum allowed error in the CAR 4b requirements, as stated in
§4b.612(b), was
§4b.612(b) Static air vent and pressure altimeter systems
(1) All instruments provided with static air case connections shall be vented
to the outside atmosphere through an appropriate piping system.
(2) The vent(s) shall be so located on the airplane that its orifices will be least
affected by air flow variation, moisture, or other foreign matter.
(3) The installation shall be such that the system will be air-tight, except for
the vent into the atmosphere.
(4) Pressure altimeters shall be of an approved type and shall be calibrated to
indicate pressure altitude in standard atmosphere with a minimum
practicable instrument calibration error when the corresponding static
pressures are applied to the instrument.
(5) The design and installation of the altimeter system shall be such that the
error in indicated pressure altitude at sea level in standard atmosphere,
excluding instrument calibration error, does not result in a reading more
than 20 feet high nor more than 50 feet low in the speed range between
1.3VsO (flaps extended) and 1.8Vs1 (flaps retracted).

This was rolled into paragraph 1325 of FAA 14 CFR/EASA CS/TCCA CAR
Parts 23 and 25, which requires the following (see Fig. 4.2):

(b) Each static port must be designed and located so that:

(1) The static pressure system performance is least affected by airflow
variation, or by moisture or other foreign matter; and
(2) The correlation between air pressure in the static pressure system
and true ambient atmospheric static pressure is not changed when
the airplane is exposed to the icing conditions defined in Appendix C
of this part, and the following icing conditions specified in Appendix
O of this part:
70 Operational Aircraft Performance and Flight Test Practices

(i) For airplanes certificated in accordance with §25.1420(a)(1), the

icing conditions that the airplane is certified to safely exit
following detection.
(ii) For airplanes certificated in accordance with §25.1420(a)(2), the
icing conditions that the airplane is certified to safely operate in
and the icing conditions that the airplane is certified to safely exit
following detection
(iii) For airplanes certificated in accordance with §25.1420(a)(3) and for
airplanes not subject to §25.1420, all icing conditions
(d) Each pressure altimeter must be approved and must be calibrated to
indicate pressure altitude in a standard atmosphere, with a minimum
practicable calibration error when the corresponding static pressures
are applied.
(e) Each system must be designed and installed so that the error in indicated
pressure altitude, at sea level, with a standard atmosphere, excluding
instrument calibration error, does not result in an error of more than
+30 feet per 100 knots speed for the appropriate configuration in the
speed range between 1.3 VSR0 with flaps extended and 1.7 VSR1 with
flaps retracted. However, the error need not be less than +30 feet.

These errors may seem large, especially for the speed when one considers the
allowed error on the indicated airspeed can be 5 kt or more on airspeeds for
takeoff and landing. In practice, the error is simply accounted for in the air-
speed numbers provided in the flight manual (takeoff and landing speeds
provided in indicated airspeed).
If we convert the altitude error of Fig. 4.2 to an equivalent absolute
pressure error, we note that the magnitude reflects what was considered
the “state of the art” capability of pressure measuring systems in the 1960s
(see Fig. 4.3). This minimum error is of the order of 0.016 psi.
The regulations also reflect that the location of the static pressure port could
be a challenge in being able to meet the maximum error requirements as speed
increases (see Fig. 2.6 in Chapter 2), with changing angles of attack of the local
flow. The maximum error at increasing airspeed now reflects more forgiving
errors when not flying close to takeoff and approach speeds. But if one converts
this error into a relative value (Dp/qc , see Fig. 4.4), then one sees that the
manufacturer may still be challenged by the increasing relative precision.
With these precisions in mind, we will now address how a manufacturer
may be able to calibrate the airspeed and altitude indicating systems to meet
basic certification requirements for a Part 25 transport category airplane.

Reduced Vertical Separation Minimum (RVSM)

Note that operational requirements may impose more severe restrictions
on the precision of the air data system than those specified under §25.1323
and §25.1325. The requirements of Part 91 for reduced vertical separation
minimum (RVSM), covered in Chapter 9, is a major driver for this need
for increased precision. Again, the more restrictive requirements come
from the need to ensure safe airplane separation based on a standard that
is used by all equally.
 CHAPTER 4 Altimetry Error 71

140 0.0700

120 0.0600
p
100 0.0500
r
ro
er
Altitude error (ft)

e
ud
80 tit 0.0400

p (psi)
Al

60 0.0300

40 0.0200

20 0.0100

0 0.0000
0 50 100 150 200 250 300 350 400

Airspeed (KCAS)

Fig. 4.3 Maximum altitude error in terms of Dp in psi.

120 0.3

100 r 0.25
e rro
u de
tit
Al
80 0.2
Altitude error (ft)

p/qc

60 0.15

40 0.1

20 0.05
p/qc

0 0
0 50 100 150 200 250 300 350 400
Airspeed (KCAS)

Fig. 4.4 Maximum altitude error in terms of Dp/qc .

72 Operational Aircraft Performance and Flight Test Practices

Std Atm
Static pressure input baro setting
10.98 psi
(7850 ft std atm)

11.55 psi
Indicated airspeed = 150 KIAS (155 KCAS)
Indicated altitude = 7889 ft
Vi = IAS – CAS = –5 kt Total pressure input Hi = Hi – H = 39 ft

Fig. 4.5 Example of instrument error.

Instrument Error
Note that both §25.1323(a) and §25.1325(d) say the instrument must be
“approved,” so systems installed in Part 25 airplanes have been subject to
numerous tests approved by the FAA. The same paragraphs show that the
instruments must provide proper airspeed/altitude when subjected to the
correct pressures. The instrument errors, for either the airspeed DVi or the
altitude DHi, are errors in indication for a known (laboratory) pressure
input (see Fig. 4.5).
For this chapter, we will assume that these errors are small enough to be
neglected. We will focus our attention on the position errors (PEs) and how
to identify them through testing. We will have another look at instrument
calibration errors near the end of the chapter after we have had a chance
to review the PE testing methods.

Flight Testing for Air Data Errors

The requirements of §§25.1323 and 25.1325 have defined the maximum
errors that are allowable for airspeed and altitude indicating systems, so the
manufacturer of a new or modified airplane must now turn its attention to
defining what the actual errors of the installed systems are under various
flight conditions. In the early years of airplane design, empirical data were
used to define the expected error during the design stage; now, however,
manufacturers rely more on computational fluid dynamics to do the job.
There still remains a need to validate the design by flight testing; the
newer methods only improve the odds that the initial estimates were good,
thereby reducing the amount of flight testing required to calibrate the air
data system.
The main challenge for the manufacturer is to define the “true” source for
airspeed and altitude that will be compared to the airplane-indicated airspeed
and altitude and define the errors of those systems. Over the years, several
 CHAPTER 4 Altimetry Error 73

flight test methods were developed that provide various levels of precision.
Each also has its own execution difficulties, dedicated instrument needs,
and resulting data scatter, which the engineers assigned to the task must
sort out to define the errors. We will now look at a few methods that allow
such work to be done. We will also discuss some of the limitations of these
methods.

Source of Position Errors

Position errors (PEs), or instrument installation errors, are the differences
that exist between the indicated airspeed and the calibrated airspeed (or the
indicated altitude and pressure altitude) due to differences between the total
and static pressure measured at the air data probes and the atmospheric
values.
Sources of errors are multiple and include (but are not limited to):
• Probe misalignment
• Boundary layer interference
• Shock waves
The measured static pressure will be denoted ps , and the measured total
pressure will be denoted pT 0 (see Fig. 4.6).
Total pressure, measured by the use of a pitot tube, can usually be
measured very accurately as long as the pitot probe is located outside the air-
plane boundary layer and is aligned with the local flow. The typical sensitivity
of a pitot probe to an off-alignment angle is shown in Figs. 4.7 and 4.8, with
some probes providing more misalignment tolerance than others.
It should also be noted that the local angle of attack (AoA) on an airplane
may be significantly different (larger) than the free stream value. Loss of total
pressure due to a shock wave can be compensated for, in part, by the use of an
air data computer with numerical compensation.
We will tackle the subject of probe installation and production tolerance
later in this chapter. We will now focus on flight test techniques to identify
the errors.
ps pT

V ≈ V
ps
V V=0

p pT

Fig. 4.6 Measured pressure nomenclature.

74 Operational Aircraft Performance and Flight Test Practices

–0.02

–0.04

–0.06

–0.08
pT/qc

–0.1

–0.12

–0.14

–0.16

–0.18

–0.2
0 5 10 15 20 25 30
Local AoA (deg)

Fig. 4.7 Example of probe misalignment on total pressure error.

True Airspeed Calibration Methods

The true airspeed calibration method consists of comparing the true air-
speed measured by the air data system on board the airplane (computed
based on measured total pressure, static pressure, and temperature) to
another reference source. We present two methods that use measured
ground speed to compute a true airspeed with correction for winds.

Closed Course
The first method is the closed course method. It uses ground references at
known distances and measures the time it takes to cover the distance between
them to compute a ground speed. This ground speed is then compared to the
indicated airspeed (IAS) of the airplane converted to true airspeed.
From a test execution point of view, the ground references to be used
should:
• Be easily recognizable from the air.
• Have accurately known distances between them.
• Have an “end line” perpendicular to the course (heading) to be followed
that should be long enough to allow for the airplane’s drift due to winds.
This will allow the pilot to accurately determine the passage.
a) Small-bore cylindrical tube 15° conical-nose tube Cylindrical tube with slant profile

7.5° 1 in. .125 in.

.125 in.
Flow Flow Flow .01 in. 1 in.
1 in. 1 in.
.125 in. 10°
Tube A-1 Tube B-1 Tube A-6
0 0 0

pt –.2 pt –.2 pt –.2

qc qc qc
–.4 –.4 –.4

–40 –30 –20 –10 0 10 20 30 40 –40 –30 –20 –10 0 10 20 30 40 –40 –30 –20 –10 0 10 20 30 40
Angle of attack, deg Angle of attack, deg Angle of attack, deg

b) Thin-wall cylindrical tube Ogival-nose tube Cylindrical tube with 30º conical entry
.125 in. .188 in.
.91 in. .125 in.
.29 in.
Flow .01 in. 1 in. Flow Flow
1 in. 15º 1 in.
1 in. Tube A-2 Tube E-1 Tube A-9

CHAPTER 4
0 0

pt –.2 pt –.2 pt –.2

qc qc qc
–.4 –.4 –.4

–40 –30 –20 –10 0 10 20 30 40 –40 –30 –20 –10 0 10 20 30 40 –40 –30 –20 –10 0 10 20 30 40

Altimetry Error
Angle of attack, deg Angle of attack, deg Angle of attack, deg

Fig. 4.8 Examples of pitot probe off-angle tolerance. Source: [1].

75
76 Operational Aircraft Performance and Flight Test Practices

End lines could be roads, power lines, or any other easily recognizable
ground references (see Fig. 4.9).
The testing relies on the flight crew to measure the time between the
starting and ending points. This requires little more than a simple stopwatch.
This reliance on a human in the loop introduces our first source of scatter for
the test: How precisely did the test conductor start and stop the timer during
the test? We will address test scatter a little later in the section when we go
over an example.
During the test, the airplane heading should be maintained constant and
parallel to the selected course (allow the airplane to drift due to the wind) so
as to cross the end line at 90 deg. Then, to eliminate the effects of the wind, a
reciprocal run (180 deg from the initial heading) should be made. The need
for a reciprocal test leads to using parallel lines as start and end conditions.

En
dl
ad
ing ine
He

ed
s pe
nd
ou
Gr

Starting point

Fig. 4.9 Ground course.

 CHAPTER 4 Altimetry Error 77

Fig. 4.10 Google map of the terrain near Wichita, Kansas, showing several north/south and
east/west roads with known spacing between them. The pilot would just need to find the right
combination from the air. Source: Google Maps.

Depending on the region where one wants to execute such a test, it could be
easy to find parallel lines (Fig. 4.10 shows an option near Wichita, Kansas) or
very difficult, which may require the tester to physically install such lines to
support the test.
During the test, the indicated altitude (based on measured static
pressure) should be held constant, because we are trying to quantify the air
data system pressure reading errors. The testing should be done as low as
possible to the ground. (Most operational regulations, such as FAA Part
91, require that the airplane be at a minimum of 1000 ft above the closest
obstacle unless a waiver is granted.) The low altitude will help the flight
crew to better determine when the start and stop ground lines are crossed,
reducing the test scatter. One way to help determine when the airplane
crosses the reference lines is to use the airplane shadow, as shown in Fig. 4.11.
To compute the TAS from a known ground course (known distance d
along the heading), the airplane must establish a constant (stable) indicated
airspeed and altitude prior to reaching the starting line; again, due to wind
drift, performing the testing between parallel lines will greatly help the test
execution. The timer is started as the airplane crosses the starting line.
During the test, the airplane will actually follow a ground track that is
defined by the geometric sum of the airplane true airspeed and heading
plus the wind speed and direction at the test altitude (see Fig. 4.12).
The ground speed (GS) along the heading for the first leg is then
 
d
GS1 ¼ ¼ TAS þ VWalong heading ð4:3Þ
t1
where t1 is the flight crew measured time to cover the distance between the
two reference lines. The wind magnitude along the heading is represented by
VW along heading . To correct for the wind effects, the course is flown immedi-
ately in the opposite direction (to minimize the chances of wind changes) and
at the same indicated airspeed and altitude.
78 Operational Aircraft Performance and Flight Test Practices

Fig. 4.11 Airplane shadow getting to reference line.

Finish
ing
ad

d
He

ck
TAS tra
nd
ou
Gr

Start

Fig. 4.12 Ground course heading vs ground track.

 CHAPTER 4 Altimetry Error 79

The ground speed along the heading for the second leg is
 
d
GS2 ¼ ¼ TAS  VWalong heading ð4:4Þ
t2

The TAS is then the average of the two legs

  
1 d 1 1
TAS ¼ ðGS1 þ GS2 Þ ¼ þ ð4:5Þ
2 2 t1 t2

The wind component perpendicular to the flight path is eliminated by

flying the reciprocal heading. During the test over a known distance d, the
flight crew must record the following information:
• Time for each leg
• Indicated airspeed
• Indicated altitude
• Indicated temperature
• Heading
The heading is not used in any computation; it is simply a reference to
remind the crew to perform a reciprocal pass to the first one. The preferred
altitude to use for this should be the pressure altitude, so the barometric
setting should be set to standard conditions (29.92 inHg). If the barometric
setting is not set to standard conditions, then the barometric setting used
for the test should be recorded. As for the temperature, in most cases, the
temperature provided to the crew is the total air temperature.
A typical test card to support this type of testing could look like Table 4.1.
The TAS must now be compared to the IAS (CAS with errors) to deter-
mine the pressure reading error of the air data system. The steps required to
convert a TAS into a CAS were detailed in Chapters 2 and 3. The broad lines
will be presented here to help with the discussion.
1. For each airspeed condition, average out the indicated altitude and
airspeed for one direction and its reciprocal.
2. Convert the averaged indicated altitude into a static pressure.
a) If the baro was set to standard condition, you can use Eq. (2.1) of
Chapter 2.

Table 4.1 Typical Ground Course Test Card

Baro
Test Distance d KIAS setting Altitude Heading Temp. Time
Conditions (n miles) (kt) (inHg) (ft) (deg) (88 C) (s)
80 Operational Aircraft Performance and Flight Test Practices

b) If the baro was not set to standard, then first use Eq. (2.5) of Chapter 2
to convert the baro altitude into a pressure altitude, and then use
Eq. (2.1) to convert the pressure altitude into a static pressure.
3. Use the recorded temperature (average again) to compute the air density.
4. Use the computed true airspeed from the ground course to compute an
impact pressure qc using Eq. (3.17) of Chapter 3.
5. Use the computed impact pressure to compute a CAS using Eq. (3.31) of
Chapter 3.
6. Compare this CAS to the recorded averaged IAS. The difference becomes
the airspeed error due to the measured pressure.
IAS ¼ CAS  DVPE  DVi ð4:6Þ
7. Before this test is executed, one must know the instrument error DVi . For this
exercise, assume that value is zero. The airspeed error then becomes simply
DVPE ¼ CAS  IAS ð4:7Þ

Example 4.1
You are flying a Rockwell Commander lightpiston-prop airplane.Test speed is 125
KIAS, and test altitude is 1200 ft, indicated with a barometric setting of 29.47 inHg.
You have selected two parallel roads that are 3.62 n miles apart. The first pass, at a
heading of 45 deg, takes 112.0 s to cover. You then turn around and execute a
pass at that same indicated airspeed and altitude with a heading of 225 deg.
This takes 93.4 s to cover. For both passes, the measured total temperature is
138C. What is the airspeed error (difference between indicated and calibrated)
Solution: As you perform the test, you collect the data in the flight log
(Table 4.2).
Based on the data in Table 4.2, the ground speed for test condition 1 would be
3:62 n miles
GS1 ¼ ¼ 116:3 kt
112:0 s
The ground speed for test condition 2 would be
3:62 n miles
GS2 ¼ ¼ 139:5 kt
93:4 s
The TAS for the test is then the average of the two ground speeds, 127.9 kt.
Table 4.2 Flight Log Data

Baro
Test Distance d KIAS setting Altitude Heading Temp. Time
Conditions (n miles) (kt) (inHg) (ft) (deg) (88 C) (s)
1 3.62 125 29.5 1200 45 13 112.0
2 3.62 125 29.5 1200 225 13 93.4

(Continued)
 CHAPTER 4 Altimetry Error 81

Example 4.1 (Continued)

The next step is to compute static pressure, which is measured based on the
indicated altitude. The indicated altitude was 1200 ft with a barometric setting
of 29.47 inHg; this converts to a pressure altitude of 1612 ft. From this
pressure altitude, we can extract the following indicated static pressure of
1995.7 lb/ft2 . We then convert the indicated airspeed into an indicated
impact pressure qci by using Eq. (3.31), and we get 53.37 lb/ft2 . The indicated
total pressure pT 0 is simply the sum of the static and impact pressure, or
2049.1 lb/ft2 .
We now need to convert the total temperature into a static air temperature
so that the first estimate of the true airspeed computed by the airplane system
can be made. The conversion can be made using Eq. (3.25), which requires a
Mach number. The estimate of the Mach number is made by measuring
the static and impact pressure using Eq. (3.24); this is equal to 0.195, which
leads to a static air temperature of 10.858C. With these values, we can now
compute a first estimate of the true airspeed based on measured parameters
using Eq. (3.16); we get 127.75 KTAS. We therefore have an error between the
truth source, the ground course, and the measured airspeed of 20.15 KTAS.
To compute the error of the probe in calibrated airspeed, we would then
convert the KTAS error to a KCAS, to get 125.126 KCAS. We would then get
an error (KCAS2KIAS) of 20.146 kt; the airspeed indicator is indicating low.
The error found in the previous example is low (though well within certifi-
cation tolerances), so one would most likely stop at this point. However, the
airspeed error needs to be assigned to one or several of the air data sensors
(the static pressure source, total pressure source, and temperature). From
experience, we know that if a pitot head is relatively well aligned with the
local flow and is well above the local boundary layer, it will read a precise
value. The same can be said for the total temperature sensor. So manufacturers
using this method will typically assign the error to the static pressure source.
If this is done in the previous example, the delta impact pressure equal to the
dVpe computed (20.146 kt equal to 0.1258 psf) would convert into an altitude
difference of a little less than 2 ft; the calibrated pressure altitude would
be higher. (A lower static pressure with the same total pressure read from
the pitot head would give a higher airspeed.). With a larger error, one would
use this new static pressure information and reiterate to get to a stable answer.
This example assumed that all measured data were exact, but there are
always some errors in all readings. What if the instruments used for the air-
speed, altitude, and temperature were as shown in Fig. 4.13—how precisely
would you expect the flight crew to read the value of interest? At first inspec-
tion, one could assume that the precision of the airspeed reading would be
within about 1 mph, the altitude within 5 ft, the barometric setting within
0.02 inHg, and the temperature within 18F. All those imprecisions will add
to the scatter of the testing.
What about the impact of the time recording? If we assume, for example,
that the pilot is using the shadow of the plane crossing the start and finish

(Continued)
82 Operational Aircraft Performance and Flight Test Practices

Example 4.1 (Continued)

Fig. 4.13 Example of instrument reading accuracy.

lines to start and stop the stopwatch, we could reasonably expect a 1/4- to
1/2-second error in the pilot action that gets recorded. Would such an
error be significant? If we use the distance from the previous example of
3.62 n miles and the computed true airspeed of 127.9 kt, under a no
wind condition, the time to cover the distance would be 101.86 s. Now, let’s
say that the total error on the time recording is an additional 1/2 second, or
102.36 s; the new computed airspeed would have been 127.3 kt, a 1/2-knot
error. Is this a lot? Probably not if the airspeed indicator was as shown in
Fig. 4.13, but the airspeed “error” here is not all attributed to the position
error effects on the pressure reading; it is in part coming from the test
execution. Figure 4.14 shows the impact of that +0.5 s on the calculated
ground speed as a function of ground course distance. It can be seen that for
a given course length, the faster the test airspeed, the greater the error. In
essence, the same 0.5-s error is applied to a smaller measured test time for a
given course length.
Another observation from Fig. 4.14 is that the course distance to use for a
1-kt test measurement error (for the 0.5-s error) would be of the order of 1.5 n
miles for a 100-KTAS test speed. The course that would give the same test
measurement error for a test speed of 300 KTAS would be about 13 n miles.
One could easily conclude that a much longer course would provide fewer
errors, but one must now look at other sources of error while using this
method. One can expect that the pilot will now need to hold constant indi-
cated altitude and airspeed for a longer time, that over the longer course
one could expect shifts in wind speed and direction, and that even barometric
pressure could change. The course needs to be completed in both direction,
and the chances of having the exact same air mass over both courses diminish
with increasing course length. Therefore, the course distance used should be
compatible with the airspeed being tested; the expected precision of this
method may not be much better than about 1 kt.

(Continued)
 CHAPTER 4 Altimetry Error 83

Example 4.1 (Continued)

+0.5-s timing error on ground course

0
–1.0 100 KTAS

–2.0
300 KTAS
Calculated ground speed error (kt)

–3.0
–4.0
–5.0
–6.0
–7.0
–8.0
–9.0
–10.0
–11.0
–12.0
0 2 4 6 8 10 12 14 16 18 20
Course length (n miles)

Fig. 4.14 Impact of measured time error on calculated ground speed.

GPS Method
The ground course method, as we have seen, consists of using a ground
speed to compute an airspeed that can be compared to the airplane indicated
airspeed. This method was a useful method for evaluating the precision of air
data system for small general aviation (GA) airplanes up to the mid to late
1990s, because it required no special instrumentation other than a stopwatch
and could easily be executed by a single pilot in the plane recording the infor-
mation on a flight log.
In the late 1970s, the U.S. military started sending Global Positioning
System (GPS) satellites into orbit. The satellites provide three-dimensional
positions and precise times. The GPS constellation was fully operational
by 1995. Although the GPS receivers were somewhat expensive initially,
they have now evolved to the point that everyone carries a smartphone
with GPS integrated and there are multiple apps available to provide a
GPS-computed speed.
We can take advantage of this newer source of ground speed data and
relook at the previously defined ground speed course method. The ground
course uses a two-leg approach to compute a true airspeed from measured
84 Operational Aircraft Performance and Flight Test Practices

Finish

in g
VW

ad
d

He
TAS

ck
TAS d tra
n
ou
Gr
GS

Start

Fig. 4.15 GPS ground track speed component.

ground speeds by “eliminating” the wind component (within a stable air

mass). Unlike the ground course method, where the ground speed along
the heading is computed by dividing the distance between two parallel
lines by the time it took to cover the distance, most GPS units will provide
the ground speed along the ground track (see Fig. 4.15).
Therefore, performing only a two-leg course and using the ground speed
provided by the GPS unit will result in an error in the computation of the
TAS. This error will be proportional to the wind speed and angle with
respect to the heading being flown. Figure 4.16 illustrates the magnitude of
the error in true airspeed if only a two-leg GPS run is performed. As can
be seen, a small wind fraction (wind speed over test true airspeed) will
result in a small error at all wind angles.
On the other hand, if the wind fraction is high (such as with low-speed
flights), the error can be minimized by flying with little crosswind. The prin-
cipal problem with using a two-leg GPS course is that the wind direction
must be known. There are ways to determine wind direction in flight, but
this may prove to be time consuming.
We need to define the wind components (speed and direction) as well as
the true airspeed. A methodology using a three-leg approach with the GPS
will allow us to define a mathematical model to find the three unknowns.
Although several methods have been defined over the years (the first
evidence of this approach being used dates back to 1918 in Germany), this
author prefers the horseshoe course, which consists of flying three con-
secutive constant-heading and constant-altitude legs at 90 deg to one
another (see Fig. 4.17). The illustration shows a three-leg course with the
first leg being straight north.
 CHAPTER 4 Altimetry Error 85

Wind fraction
45%
1.0
40%
0.9
35%

30%
GPS TAS error (%)

25% 0.7
20%

15%
0.5
10%

5% 0.3
0.1
0%
0 10 20 30 40 50 60 70 80 90
Wind angle relative to heading (deg)

Fig. 4.16 GPS method–induced error caused by using the two-leg method.

VWE is the easterly component of the wind, and VWN is the northerly
component. Using this approach, we now have three known ground speeds
(GS1 , GS2 , GS3 ) and three unknowns (TAS, VWE , VWN ). The speed
equations for the three legs are

GS21 ¼ ðTAS  VWN Þ2 þVWE

2

GS22 ¼ ðTAS þ VWE Þ2 þVWN

2 ð4:8Þ
2
GS23 ¼ ðTAS þ VWN Þ 2
þVWE

VWE
TAS
VWE VWN
GS2
VWN

GS3
TAS
TAS
GS1
VWN

VWE

Fig. 4.17 GPS horseshoe course method.

86 Operational Aircraft Performance and Flight Test Practices

Solving for the three unknowns, we get

2 GS22  GS21  GS23
VWE ¼ ð4:9Þ
4 TAS
GS23  GS21
VWN ¼ ð4:10Þ
4 TAS
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
b + b2  4c
TAS ¼ ð4:11Þ
2
 2 
 GS1 þ GS23
b¼ ð4:12Þ
2
GS41 þ 2 GS42 þ GS43  2 GS21 GS22  2 GS22 GS23
c¼ ð4:13Þ
8
Heading 1 does not necessarily need to be true north; only three headings
at 90 deg to each other are required for this analysis, with heading 3 being the
reciprocal of heading 1. The “north” for the purpose of using these equations
can be the initial heading. Because several airspeed points will most likely be
taken for airspeed calibration, and in order to minimize the risk of the erro-
neous data from one GPS leg throwing a speed completely out of the data set,
a GPS course can often be done as a four-leg course from which one can get
four three-leg courses (see Fig. 4.18).

GPS course 1

GS2

GS1

GS1 GPS course 2 GS2 GS3

Race track at constant

KIAS and Hpi

GS3

Fig. 4.18 Four-leg GPS course.

 CHAPTER 4 Altimetry Error 87

Following the four-leg course, the collection of speeds essentially starts at

the same point in space; the radiuses of the turns are dictated only by the air-
plane speed (discussed more in Chapter 14). During the execution of this
GPS course, one would need to collect:
• KIAS of interest (ship’s systems, nose boom)
• Hpi (indicated pressure altitude from systems of interest)
• Static air temperature (SAT; from most reliable source)
• Ground speed (from GPS)
• Heading (as reference)
For each segment, one would need to collect the ground speed for about
10 s on each leg once KIAS and indicated altitude are stabilized. (GPS units
typically stabilize very rapidly.) Data collection (repeatability) is improved in
calmed, smooth air.

Example 4.2
You are flying a small business jet and want to validate the indicated airspeed
at 140 kt. You perform a four-leg GPS course where you record the following
information:

Leg KIAS (kt) Hpi (ft) SAT (deg C) GS (kt) Heading (deg)
1 140 10,100 8.2 175 0
2 141 10,050 8 165 90
3 140 10,110 8.1 153 180
4 140 10,200 8.2 164 270

You note that although you did your best to achieve constant altitude, you
were not able to achieve exactly the same value during each leg of the GPS
course; this may add scatter to your test results. Also note that you collected
static air temperature (SAT). What is the calibrated airspeed of the airplane
for this 140-KIAS nominal test condition?
Solution: You set about to determine the position error. Because this was a
four-leg course, you will pick two three-leg courses, leg 1-2-3 and leg 2-3-4, for
data reduction. For the first leg, you compute the following:
Average KIAS ¼ 140:3 Average Hpi ¼ 10,087 ft
b ¼ 27,017 c ¼ 3,265,232
KTAS ¼ 164:0 VWN ¼ 11:0 kt VWE ¼ 0:6 kt
Now using the indicated pressure altitude, the recorded static air tempera-
ture, and the computed true airspeed, one can compute the expected

(Continued)
88 Operational Aircraft Performance and Flight Test Practices

Example 4.2 (Continued)

calibrated airspeed.
KCAS1 ¼ 137:7 kt
Compared to the measured average indicated airspeed, it represents an
airspeed error of 22.6 kt [Eq. (4.6) with zero instrument error]. We then do
the same exercise with the second three-leg course, and we get
Average KIAS ¼ 140:3 Average Hpi ¼ 10,120 ft
b ¼ 27,060:5 c ¼ 3,340,128
KTAS ¼ 164:1 VWN ¼ 0:5 kt VWE ¼ 11:1 kt
The true airspeed computed closely matches the one computed for the
first three-leg course, an indication that the air mass was reasonably stable.
Note the wind magnitude is also similar, with the north and east components
now reversed because the second three-leg course is 90 deg to the first one, so
the “north” shifted 90 deg from an equation point of view. We compute the
calibrated airspeed for this course and get
KCAS2 ¼ 137:7 kt
This is also an error of 22.6 kt, another indication of a stable air mass.
This 2.6-kt error is an impact pressure difference of 2.5 psf for a fixed static
pressure. As per the ground course, the GPS course does not identify if the
reading error is from the static pressure source or the total pressure source.
As stated before, most total pressure probes will read the correct value, so
we naturally assign this 2.5-psf error to the static system. This represents an
altitude error of approximately 245 ft. Applying this altitude error to the
static system and recomputing the calibrated airspeed with the previous
steps would lead to 137.8 KCAS, a small adjustment. Note that this relatively
large 45-ft altitude error at 10,000-ft altitude, which represents a static
pressure error of 0.0173 psi, is still an acceptable certification error per
Fig. 4.3 once the pressure error is compared. Note the words of 25.1325(e)
“at sea level“—the pressure error is equivalent to an altitude error at sea level.

The main advantages of the GPS course over the speed course are:
• The GPS course can be done at any altitude (not required to be close to the
ground per the speed course method, thus improving safety) and
essentially any speed/Mach number.
• The airplane only needs to stabilize on speed and heading (a few seconds)
before changing heading. There is no need to fly a given ground
distance.
• The data can be continuously recorded so the most stable ground speed
can be selected for any leg.
• The GPS course can be done with any three-leg course. (Only the
mathematics would change from the methodology described with the
horseshoe method.)
 CHAPTER 4 Altimetry Error 89

The main disadvantage is that it assumes the winds are constant for the
three legs, but this is the same limitation as a normal speed course. Another
disadvantage is the relatively laborious computation steps, just like the
ground course, although the introduction of tools like spreadsheets greatly
helps reduce the workload. This will lead us to consider the next method-
ology category, the free stream static-pressure calibration method.
For determining speed, a regular GPS system is sufficient; no differential
GPS (DGPS) required. The inherent speed error of the GPS signal is of the
order of 0.06 kt, an error that is not necessarily improved by going to
DGPS. Also, this error should easily be hidden by the normal scatter encoun-
tered in any testing. Prior to 2 May 2000, the U.S. Department of Defense
deliberately degraded the GPS signal for all users (except the military, of
course) with selective availability (SA). SA was simply a degradation of the
GPS satellites’ atomic clocks that results in a position error of approximately
100 m and a speed error of about 1 kt.

Free Stream Static-Pressure Calibration Method

Unlike the previous category of calibration methods, for which one must
make the assumption that all or most of the air data error is assigned on a
given system (static pressure), in this category, one can now make measure-
ments that will give better insight into the static-pressure system error and
then also possibly derive total-pressure measurement error.

Pacer Airplane
This method consists of two airplanes flying at the same altitude (see
Fig. 4.19); one is the test airplane with the noncalibrated pitot and static
systems, and one is the pacer airplane with calibrated pitot and static

Fig. 4.19 CRJ700 (Pacer) following a CRJ900 on its first flight. Source: Bombardier.
90 Operational Aircraft Performance and Flight Test Practices

Fig. 4.20 View from the cockpit of chase plane following an Airbus A350 on first flight, 14 June
2013. Source: Airbus.

systems. Flying at the same altitude now provides a direct comparison of the
test airplane static-pressure system to a truth source.
For the truth source to be of any value, the two airplanes must be close
enough to each other to allow the pace pilot to zero out the relative airspeed
and altitude between the two airplanes, but far enough so that they are not
affected by each other’s pressure field, and of course for safety reasons.
The relative altitude error between the two planes now relies on the pilot’s
ability to stabilize the chase plane and visually gauge the relative altitude
between the two planes (see Fig. 4.20).
The relative airspeed can be more easily zeroed out than the relative alti-
tude while flying the two airplanes relatively close to one another; however,
the pacer airplane also should be relatively small (Fig. 4.21). (Two 747s, for
example, would need to fly too far apart for the pilot to judge relative airspeed
to within a few knots and altitude within a few feet.) A large relative size
difference between the test and pacer plane (as shown in Fig. 4.22) introduces

Fig. 4.21 Cessna Citation  (pacer) following a Cessna Citation 10 on its first flight.
Source: Cessna.
 Pacer/chase pitot-static probe

CHAPTER 4
Test airplane pitot probes Static ports

Fig. 4.22 Relative size of pacer plane compared to test airplane adds test complexity.

Altimetry Error
91
92 Operational Aircraft Performance and Flight Test Practices

test logistics for air data sensor alignment. The following flight test traces give
an idea of the relative precision of the method (see Fig. 4.23).
One way to help the pilot of the pacer plane to zero out relative
airspeed and altitude is to use a design feature of the airplane, such as the
wing leading edge sweep. Once the pacer airplane has essentially zeroed
out the relative speed and altitude, there is then a need to establish a
clear communication strategy to accurately read both airspeed and altitude
simultaneously (see Fig. 4.24). As a note, it must also be remembered that
the airspeed being zeroed out is really the true airspeed between the two
planes.
As well as the test-related errors, one must consider that any error in the
pacer airplane air data system will also be transferred to the test airplane. It is
understood that these errors are small and known from the calibration exer-
cise of the pacer plane and that they are much less than the test-related
errors.

Chase Plane
Most airplane manufacturers have an airplane (or two) accompany the
first prototype of a new model on its first flight (or the first few flights) to
inspect the airplane in flight and report back to the crew flying the prototype
(Figs. 4.25 and 4.26). This chase airplane is not necessarily equipped with
an experimental calibrated air data system; it relies on its production
system for airspeed and altitude (known airspeed and altitude errors, as cer-
tified). It can be used to determine if the air data system of the test airplane
provides a reasonable reading, especially in the low speed region for takeoff
and landing.
After one or two flights with part of the envelope expanded and the air-
speed and altitude reading having been judged representative, the use of the
chase airplane can be discontinued. Other PE testing methods can then be
used to complete the air data system calibration.

Tower Fly-By
The most accurate altimetry method to determine static pressure error is
the tower fly-by method, which allows the test engineer to resolve a geo-
metric height above a reference (ground or tower height) into a pressure
height (see Fig. 4.27).
dp ¼ rSL sgSL dH ¼ rSL sg dh ð4:14Þ
dh ¼ x tanðuÞ ð4:15Þ
This method consists of the airplane making a low-level, constant-
altitude and constant-airspeed pass along a reference line (such as a
runway) at a known distance x from a tower or reference ground point.
The method is referred to as a tower fly-by simply because one way to
reduce the height difference between the airplane flying low above the
 Test airplane Pacer airplane
155.0 158.0
Indicated airspeed (kt)

154.5 156.5 Zeroing out

154.0 relative airspeed
155.0
153.5 153.5
153.0 152.0
152.5 150.5
152.0 149.0
12,100 12,050

12,050 12,000

12,000 11,950

11,950 11,900

11,900 11,850

CHAPTER 4
Reading

Altimetry Error
Fig. 4.23 Flight test data of pacer vs test airplane.

93
94 Operational Aircraft Performance and Flight Test Practices

Fig. 4.24 Means to help zero out relative airspeed.

ground and the reference pressure source [minimizing dh in Eq. (4.14)] is to

install such a source in a control tower. The method does not need a tower,
however, simply a means to resolve the difference in geometric height
between a reference (truth) pressure source and that of the airplane.
The tower must contain a way to track the airplane true geometric alti-
tude above a reference altitude (the altitude at which a calibrated weather
station is located). One way to do this is by taking photographs of the airplane
to compare its height against known distances (see Fig. 4.28). From these

Fig. 4.25 Learjet 45 (chase) following a Learjet 40 on its first flight. Source: Bombardier.
 CHAPTER 4 Altimetry Error 95

Fig. 4.26 Bombardier CS100 (test) with a Global 5000 (chase).

Calibrated
tower


Reference
x fly-by line

Fig. 4.27 Geometry of a tower fly-by test.

Fig. 4.28 Photographs of airplanes performing tower fly-bys with geometric references.

photographs, the geometric height of the airplane (its static pressure sensor
actually) can be resolved by trigonometry.
This height is compared to the known height of the reference pressure
sensor (truth source) to get a differential height dh. The static pressure
96 Operational Aircraft Performance and Flight Test Practices

measured by the airplane’s air data system is compared to the static pressure
measured by the calibrated weather station at the time the picture was taken,
corrected by the use of the hydrostatic equation of Eq. (4.14) for the differ-
ence in height between the airplane and the reference altitude. It is then
very important to properly record the time of each event and reconcile this
to the picture taken. An airplane with continuously recorded data and a
weather station with the same capability, both time synchronized, will
greatly help with the data reduction; however, this method still requires
laborious data reduction because each picture must be analyzed carefully
and a geometric height derived. Of the three methods shown in Fig. 4.28,
the one on the left uses a grid in the tower as a means to resolve the geometric
height. The one in the center has cones located at known distances (the two
arrows) to provide a known measure against the picture. Finally, the picture
on the right uses a pole with reference white and black marks of known
height to resolve the airplane height some distance behind [see Fig. 4.29
and Eq. (4.16)].
h1
h2 ¼ ðx1 þ x2 Þ ð4:16Þ
x1
Other sources of height that can be used to resolve the difference in height
between the test airplane and the truth source weather station include
phototheodolite, a radar tracking antenna, or even the on-board radar
altimeter if the information is continuously recorded and resolved against
the airplane location over the runway (time and location; see Fig. 4.30).
From a test execution point of view, there are inherent risks in performing
tower fly-bys. First, the airplane’s altitude over the ground needs to be low;
thus, the airplane will be exposed to:
• Flying towards the ground; the pilot must be able to stabilize at the
correct altitude
• Turbulence close to the ground
• Higher terrain near the test site (good visibility required)
• Birds

h2

h1
Camera

x1 x2

Fig. 4.29 Trigonometric height.

 CHAPTER 4 Altimetry Error 97

255 300

254 250
KIAS

253 200
Airspeed (KIAS)

Hpi Crew events

Altitude (ft)
252 150

251 100
RADALT

250 50

240 0
0 5 10 15 20 25 30 35 40 45
Ref time (sec)

14.66 250

14.64
Cone Ps 200
Static pressure (psi)

14.62
150

Altitude (ft)
Ships Ps
Crew events
14.60

100
RADALT
14.58

50
14.56

14.54 0
0 5 10 15 20 25 30 35 40 45
Ref time (sec)

Fig. 4.30 Copilot event in data stream (on location) vs air data and radar altitude (RADALT).

Then, the airplane airspeed also needs to be considered. For the very low
speed conditions:
• There is the possibility of stall near the ground, especially in
turbulent conditions.
• There is a lack of climb capability following the loss of power in one engine.
Even the high-speed test includes increased test risks such as reduced
pilot reaction time on the vertical plane and extended test distance. For
98 Operational Aircraft Performance and Flight Test Practices

this reason, tower fly-bys are most often performed over a runway. Then,
coordination with air traffic control (ATC) and surrounding air traffic
becomes very important, because most of the speeds evaluated are much
faster than the approach speeds of other airplanes in the pattern.

Alternative Tower Fly-By

An alternative to the tower altimetry method requires using DGPS on
board the airplane to record the geometric altitude (DGPS Altitude 2) on
a continuous basis during a fly-by over a calibrated weather station at a
known geometric altitude (DGPS Altitude 1). The weather station geo-
metric altitude should also be recorded on a continuous basis to reduce
the geometric altitude error even more (see Fig. 4.31). DGPS time sync
allows the operator to compare the measured pressure altitude (static
pressure) at the time the airplane flies over the weather station (the
DGPS providing latitude and longitude as well as time and height) to the
latter recorded pressure.
If the geometric altitude difference between the calibrated weather
station and the airplane at the time of the recording is kept to less than
500 ft (most traditional tower fly-by testing is done at 50–100 ft), a good cor-
relation can be obtained by using the standard temperature gradient derived
from the hydrostatic equation of Eq. (4.14). This method dispenses with the
requirement to have a dedicated tower and provides for a higher above-
ground clearance for increased test safety. The height of the airplane and
the weather station are typically continuously recorded for these tests, so
this also simplifies the data reduction activities.

Trailing Cone
As we have seen earlier, using a secondary airplane (pacer or chase) as a
means to provide a source of air data error (via difference in indicated alti-
tude) is a practical way to isolate the error on the static system and pitot
system. That method does, however, involve the cost of flying a second

DGPS altitude 2

Calibrated
DGPS altitude 1 weather
station

Fig. 4.31 Alternative tower fly-by.

 CHAPTER 4 Altimetry Error 99

Static pressure ports

Fig. 4.32 Trailing static cone.

airplane and adds the inaccuracy of altitude gauging by the pacer plane to the
overall equation.
One way to measure the free stream static pressure with minimal to zero
error is to trail a long tube containing static pressure ports behind the air-
plane. This was first presented as a flight test tool by the Douglas Aircraft
Corporation in 1964 [3] from work the company had started back in 1959.
The basic concept came from studies that indicated the static pressure
field around an airplane would revert to atmospheric pressure within a
reasonable distance behind the airplane and that orifices in an “infinite-
length” cylindrical tube at zero angle of attack record static pressure
accurately.
The basic concept that Douglas went on to develop was to install the
perforated cylindrical tube near the end of a reasonably long tube (to
carry the measured pressure to the sensor). The cylindrical tube was stabil-
ized with a perforated cone to create drag and keep the tube as straight and
stable as possible, aligned with the flow, and therefore independent of the
airplane angle of attack (see Fig. 4.32). The measured static pressure pro-
pagates inside the tube, all the way to the air data system. Douglas
showed that the results were very accurate to high Mach numbers (see
Fig. 4.33).
The trailing cone effectively became a perfect pacer airplane equivalent
with a much-reduced cost of operation. The free stream static pressure is
channeled to the test airplane data recording system and can be compared
directly to the production system static pressure via a much more precise
differential pressure sensor, thus continuously recording the difference
between the two (Fig. 4.34). Another advantage of using this approach is
that the pressure is now compared at the exact same altitude; the hydro-
static pressure in the tubes automatically corrects the difference in
height between the production pressure ports and the cone static
pressure ports.
If an airplane manufacturer elects to use a trailing cone as a truth source,
the next decision will be how long it should be and where to locate it (see
Fig. 4.35). Locating the trailing cone at the back of the fuselage usually sim-
plifies the installation, but this subjects the cone more to the wake of the
100 Operational Aircraft Performance and Flight Test Practices

Symbol Aircraft type Tail to orifice distance

EA-38 43–92 ft. Notes:
Aero commander 57 1. Tower pass method
A-4C 41 2. Power approach – open symbols
A-4E 40 3. P = Pcone–Pambient
4. Qc = Impact pressure
.10
.08 Equivalent to
.06 10 ft error band Max. Mach No. = 0.825
+
.04
Cone static error

.02
P 0
Qc
.02
.04
–
.06
.08
.10
0 100 200 300 400 500 600
Vm – Knots

Fig. 4.33 Douglas experimental investigation of the trailing cone accuracy [3].

fuselage and landing gear, downwash of the flaps (especially landing flaps at
low speed), and possibly engine thrust setting. A longer cone is then required
to move away from those influences, but longer trailing cones are often less
stable at airspeed extremes (low or high).
Locating the cone at the top of the vertical tail is usually seen as a good
solution with a slightly more complex installation. Next comes the decision
of whether to have a fixed trailing cone or a retractable one. Fixed cones
are simpler to design, but require the tester to contend with the possibility
of having the cone drag on the runway for takeoff and landing, which leads
to some damage. A retractable cone solves this issue, but the airplane must
have room in the fuselage to contain the retraction device and the extra
length of tubing. This also leads to longer tubing overall, thus increasing
pressure sensing lag.
Although the trailing cone is expected to provide a good reference from
the start, most manufacturers will still perform a tower fly-by with the

p sensor

Cone static
pressure ports

Production static
pressure ports

Fig. 4.34 Trailing cone truth source compared to production static pressure.
 CHAPTER 4 Altimetry Error 101

– Flaps up – 76 KTS
– Flaps up – 150 KTS
– Flaps dn – 75 KTS
– Flaps 1/2 – 122 KTS .06
Location and length
Horizontal tail root chord = 4.8 ft .04
P = Ps – Ps .02
cone trailing bomb
0
–.02
–.04
–.06
0 4 8 12 16 20 24 28 32 36 40 44 48
Length – Feet

Fig. 4.35 Selecting trailing cone location and length. Source: Data from [3].

trailing cone extended to validate the reading (Figs. 4.30 and 4.36). This also
ensures that, should the cone not be totally stable behind the airplane, the
tester can have another precise reference (tower fly-by) to validate
the signal.

Lag in the Experimental System

Depending on the size of the airplane, the length of tubing (both
outside and inside the airplane) between the static ports and the air data
system can be significant. This can cause significant propagation time, or
lag, from the instant a pressure is registered at the static ports to the
time it reaches the air data system. For this reason, calibration using trail-
ing cones is usually done at constant altitude and airspeed to ensure that
the differential pressure sensed between the trailing cone and the pro-
duction system is due only to installation/location of the production
probe and not to a lag in the experimental system. Figure 4.37 shows
the data from a test flight with a trailing cone. During the stable part of
the maneuver (between the dark lines, fairly constant airspeed and very

Fig. 4.36 USAF Learjet C-21, trailing cone extended, tower fly-by testing.
102 Operational Aircraft Performance and Flight Test Practices

38550 EOT
Prod statc P (psi) Indicated alt (ft)
38500
38450
38400
38350
2.940
2.935
2.930
2.925
2.920
2.930
Trailing cone (psi)

2.925
2.920
2.915
2.910
15:50:20 15:50:30 15:50:40 15:50:50 15:51:00 15:51:10 15:51:20 15:51:30

EOT
238.0
237.0
KIAS

236.0
235.0
234.0
0.765
Indicated mach

0.760

0.755

0.750

0
Delta P (psi)

–0.005

–0.010 Trailing cone – Production static pressure

–0.015
15:50:20 15:50:30 15:50:40 15:50:50 15:51:00 15:51:10 15:51:20 15:51:30

Fig. 4.37 Test data showing trailing cone lag.

small climb), the delta pressure between the production system and the
trailing cone (bottom right trace) is stable and shows the measured
error. The traces also show that as soon as the test is done [end of test
(EOT)] and the airplane rapidly changes altitude, the delta pressure
between the production system and the trailing cone rapidly changes, yet
the airspeed and altitude are fairly the same; this is an indication of
 CHAPTER 4 Altimetry Error 103

pressure lag in the trailing cone. Had the trace been extended a little more,
the conditions would have stabilized again.
Therefore, for a steady airspeed and altitude, the pressure read by the
pressure sensor is the “same” (hydrostatic correction for height difference
expected) as the pressure at the pressure port, and there is no air mass
flow into or out of the system. When there is a dynamic change in pressure
at the pressure port (such as during a climb or an acceleration), the equili-
brium between the pressure in the system and at the source is disturbed.
Air will then move into or out of the system, depending on the rate of
change in airspeed/altitude. It will take some time for the pressure disturb-
ance to propagate to the pressure sensor; this time is proportional to the
length and volume in the system. Experimental systems also are usually
more impacted by these changes due to their longer tubing requirements
(measuring pressure further away from the airplane). This can be seen in
Fig. 4.37 where the production static pressure (midstrip left side) reacts
much faster to the change in pressure than the trailing cone (bottom left
strip) with its longer tubing.
The two major air data lags in a system are the acoustic lag and the pneu-
matic lag (see Fig. 4.38). The acoustic lag is the time it takes for a pressure
wave to travel a certain distance; in this instance, it is the time from the
pressure port to the pressure sensor. This time will be proportional to the
length of tubing from the pressure port to the instrument divided by the
speed of sound in the system (i.e., t ¼ L/a). The speed of sound is pro-
portional to the square root of the temperature of the air in the system
and should fall roughly between 950 and 1150 ft/s. For production
systems, where the tubing length can be very small, this lag is often negligible.
For an experimental system such as a trailing cone, where the total length of
tubing between the pressure port and the pressure sensor can reach hundreds
of feet, this value can become significant and should be accounted for by
the tester.

Pressure Acoustic lag

Pneumatic lag Ps

Pressure at
gauge

Pressure at source
Sensor
Time

Fig. 4.38 Air data lag.

104 Operational Aircraft Performance and Flight Test Practices

Example 4.3
A production system has 5 ft of tubing between the static pressure port and
the pressure sensor. The experimental trailing cone installed on the airplane
to calibrate the production system has 250 ft of tubing between the pressure
port and sensor. Assuming that the air temperature in the respective systems
(production and experimental) are the same and result in a speed of sound of
1000 ft/s, compute the acoustic lag of each system while the airplane is in a
3000 ft/min descent.
Solution: For the production system, with 5 ft of tubing, the acoustic lag
would be
5 ft
t¼ ¼ 0:005 s
1000 ft=s
The lag in the experimental system would be
250 ft
t¼ ¼ 0:25 s
1000 ft=s
Thus, in a 3000 ft/min descent, the production system would react to the
pressure change within 0.25 ft, but it would take the experimental system
about 12.5 ft to react to the pressure wave.
Once the pressure disturbance has propagated to the instrument, the
pressure at the instrument will start changing to catch up with the pressure
at the source; this is the pneumatic lag. This change comes from the air
flowing into or out of the tubing. Mass flow in a tube encounters resistance
that will slow it down. The pressure variation over time, or pneumatic lag,
for a given pressure disturbance can be modeled as a first-order system
with a time constant. The pressure drop is then related to the lag as follows:

dp
Dp ¼ l ð4:17Þ
dt

This time constant is best seen and modeled when a sudden pressure drop
happens (such as when the tube is being calibrated on the ground). In flight,
the change in pressure is more gradual and the lag is not as easily quantified.
Some authors have offered simplified equations based on a set of assumptions
to try to quantify this lag. Gracey [3] offers Eq. (4.18), in which the lag is pro-
portional to the tube length L, the diameter of the tube d, its volume C, the
viscosity of the air in the system m, and the pressure p in the system. He
also makes the assumption the tube is straight and laminar flow exists in
the tube.
128 mL C
l¼ ð4:18Þ
pd 4 p

(Continued)
 CHAPTER 4 Altimetry Error 105

Example 4.3 (Continued)

The equation is useful to compare the relative lag of a given system (pro-
duction vs experimental) once the dimensions of each are known.
Of interest for the air data lag, some newer “smart” air data systems
have integrated the air data probe directly into the pressure sensor and air
data computer; they are a single integrated unit. This has reduced the total
length of tubing from the pressure port to the pressure sensor to only a few
inches. The lags, both acoustic and pneumatic, are essentially zero with
those systems.

Display Lag
Electronic flight information displays providing digital information have
largely replaced the mechanical analog displays on airplanes. The trend
started in the late 1970s. Although small movements of the airspeed
and attitude gauges on analog systems were acceptable, they also provided
the crew with an intuitive direction of the change. When the airspeed
started to be displayed in digital form, oscillation of that value during critical
tasks like takeoff would increase the crew workload. Electronic displays
with their digital values could be too jittery to read if the values displayed
were not filtered, so avionics manufacturers had to add filtering to smooth
out the reading. This filtering of the signal introduces lag into the displayed
value.
FAA Advisory Circular AC 25-11B provides the following guidance for
electronic displays (underline by author):
5.9 Dynamic (Graphic) Information Elements on a Display. The following
paragraphs cover the motion of graphic information elements on a display,
such as the indices on a tape display. Graphic objects that translate
or rotate should do so smoothly without distracting or objectionable jitter,
jerkiness, or ratcheting effects. Data update rates for information elements
used in direct airplane or powerplant manual control tasks (such as
attitude, engine parameters, etc.) equal to or greater than 15 Hertz have
been found to be acceptable. Any lag introduced by the display system
should be consistent with the airplane control task associated with that
parameter. In particular, display system lag (including the sensor) for
attitude that does not exceed a first order equivalent time constant of 100
milliseconds for airplanes with conventional control system response is
generally acceptable.
5.9.1 Movement of display information elements should not blur, shimmer,
or produce unintended dynamic effects such that the image becomes dis-
tracting or difficult to interpret. Filtering or coasting of data intended to
smooth the motion of display elements should not introduce significant
positioning errors or create system lag that makes it difficult to perform
the intended task.
...
106 Operational Aircraft Performance and Flight Test Practices

5.11.5.3 Dynamic images should meet the guidance in paragraph 5.9 of this
chapter. The overall system lag time of a dynamic image relative to real time
should not cause flightcrew misinterpretation or lead to a potentially hazar-
dous condition. Image failure, freezing, coasting, or color changes should not
be misleading and should be considered during the safety analysis.

One such dynamic condition is the takeoff phase, during which the airplane is
subjected to large acceleration and the takeoff distance is based on the crew
executing precise maneuvers (rotate at VR , pitch to target, maintain V2 or
V2 +10, etc.). The airplane manufacturer needs to reconcile this extra lag in
the system with the air data errors and lag of the pitot-static system to
provide accurate displayed airspeed (DAS) to the crew. We point out
the difference between displayed airspeed (DAS) and indicated airspeed
(IAS) in this section to emphasize that the former is display filtering induced
whereas the latter depends strongly on the installation of the air data probes
but also includes acoustic lag, pneumatic lag, and ADC computation lag.
To better understand the dynamics of this new lag, we look at an example
of air data system (from probe to display) lag testing (see Fig. 4.39). For this
test, the acceleration was of the order of 2.5 kt/s, which is representative of a
one-engine-inoperative takeoff for a multiengine transport-category airplane.
The value of the airspeed was recorded at the analog input source (total
pressure and static pressure; note the smoothness of the data), then at the
output of the ADC, and finally at the “output” of the display.
The lag of the ADC (acoustic-pneumatic lag plus computation time to
output) is within 1 kt of rotation speed (say 110 kt, read left scale to input
value line, then read down for fixed time); however, because of filtering,
the displayed value (the one the pilot really uses) is approximately 3 kt
slower than the input value. This speed difference must be accounted for
during the preparation of the flight manual. Also note how the displayed

120 4

Input 115 3.5

(truth source)
110 3
Input, ADC, display

Acceleration (kt/s)

Pneumatic 105 2.5

lines Display filtering and 100 2
refresh rate
Air data computer
(ADC) 95 1.5

90 1

Electronic signal 85 0.5

80 0
Time
ADC Input Display Accel

Fig. 4.39 Elements of display lag.

 CHAPTER 4 Altimetry Error 107

airspeed (and the ADC output) are more digital in nature than the smooth
input value.

Nose Boom
Having the truth sensor away from the pressure field of the airplane
greatly improves the sensor’s ability to provide a good reference reading
with little to no correction. This improves the data comparison between
the reference system and the ship systems. Locating the reference sensors
ahead of the plane requires shorter separation from the airplane than when
they are located behind, such as a trailing cone. One means to carry the
sensors is through the installation of a nose boom.
For subsonic Mach numbers, some NASA work showed that the angle of
attack at the nose boom at a distance approximately 1.5 times the effective
diameter of the fuselage was essentially that of free stream (see Fig. 4.40).
This also applies very well for static and total pressure reading, thereby
making the nose boom a good test tool to carry truth sensors. Other
sensors that can be found on nose booms include total temperature
probes, angle of attack, and angle of sideslip vanes.
Although the sensors on the nose boom may be sufficiently far enough
from the airplane’s pressure field to provide reliable data under static con-
ditions, the tester still needs to verify that they do so under the conditions
of interest. For example, a pitot probe located at the tip of the boom will
have no influence from the airplane, but depending on the model used, it
may still suffer from medium to high flow AoA conditions (see Fig. 4.8
earlier in the chapter). The same can be said for the static pressure ports.
Any probe on the nose boom will also be impacted in some form by the
boom vibration, such as may happen at high angles of attack. The boom
tip is far from the center of gravity (point of rotation), so any vanes on the
boom tip will be influenced by pitch rate and yaw rate (change in local

20
x 1.5
16 CN
0.6
1, deg

Location of vane D = Maximum effective 12

fuselage diameter NACA TN 4351

8 0.4

0.2
4

0
0 0.4 0.8 1.2 1.6 2.0
a) Sensor location on airplane b) Variation of 1 with x/D.

Fig. 4.40 Locating reference sensors away from airplane. Source: Data from NACA TN-4351 [4].
108 Operational Aircraft Performance and Flight Test Practices

Photo: USAF

Fig. 4.41 Lockheed C-5M nose boom.

angle of the flow due to the angular velocity of the plane) and possibly even
roll rate at high angles of attack (roll around the velocity vector). The tester
still needs to review the data from the boom and apply appropriate correc-
tions and calibration.
Because of the length requirements, and thus the strength and rigidity
requirements, the use of a nose boom is usually limited to relatively small air-
planes. As the airplane size increases, the boom length must also increase to
keep the truth sensors out of the pressure field of the airplane; however, there
are exceptions. When the U.S. Air Force decided to proceed with the
improvement of the Lockheed C-5 to bring it up to the C-5M standard,
they elected to install a very large nose boom (see Fig. 4.41) with the stiffening
required to make the boom useable to high Mach numbers.
Figure 4.40 shows that a nose boom of reasonable length provides a good
environment for the truth sensors; however, there are some drawbacks
associated with the long boom including increased risk of hangar rash, poss-
ible need to use a longer tow bar, and possibly even needing a bigger hangar.
In recent years, we have seen some test airplanes use short booms (see
Fig. 4.42). These booms are easier to install, are much stiffer (not subject
to boom vibration), do not require ground handling procedures much differ-
ent than for a normal plane, and do not need much more hangar space than
the baseline plane. They are much smaller than the ideal boom length shown
in Fig. 4.40, so they do need to have special in-flight calibration procedures to

Mini-boom

Fig. 4.42 Bombardier CSeries CS100 experimental mini-boom.

 CHAPTER 4 Altimetry Error 109

M Sensor
1.6 0.35 A
0.60
0.35 B
1.2 0.60

Sensor A 90° 0.5m

p
Orifices .8
q

Sensor B 60°
.4
Orifices

0 10 20
, deg

Fig. 4.43 Differential pressure sensor to measure angle of attack, on short boom.

ensure that the data recorded are as expected. One such procedure could be
flight path reconstruction in which other systems (like an inertial reference
system) are used to get reference angles of attack and sideslip.
This type of boom uses differential pressure sensors to compute the
angles (see Fig. 4.43). These are similar to the air data probes use on airplanes
like the Grumman F-14 fighter or the SR-71.

System Leak
Another contributor to the altitude/airspeed error is air data system leak.
A well-designed, perfectly built air data system will have one opening to the
air where the pressure needs to be measured, and there will be no other
source of pressure between that opening and the pressure sensor. Typical
systems have a series of tubes connected via a series of connectors; these con-
nectors are typically a source of air pressure leak. Leaks result in the contami-
nation of the pressure being read by the instrument from pressure other than
that at the source.
If part of the system (such as the pressure reading instrument) is in a
pressurized compartment and there is a leak, some of the pressure within
the compartment will enter the system and alter the pressure at the instru-
ment, even in a steady airspeed and altitude condition, resulting in a different
value being displayed to the flight crew.
Certification requirements for Part 25 specify the level of acceptable
leakage in an air data system under §25.1325:
§25.1325(c) The design and installation of the static pressure system must be
such that—
(1) Positive drainage of moisture is provided; chafing of the tubing and
excessive distortion or restriction at bends in the tubing is avoided; and
the materials used are durable, suitable for the purpose intended, and
protected against corrosion; and
(2) It is airtight except for the port into the atmosphere. A proof test must be
conducted to demonstrate the integrity of the static pressure system in
the following manner:
110 Operational Aircraft Performance and Flight Test Practices

(i) Unpressurized airplanes. Evacuate the static pressure system to a

pressure differential of approximately 1 inch of mercury or to a
reading on the altimeter, 1,000 feet above the airplane elevation at
the time of the test. Without additional pumping for a period of 1
minute, the loss of indicated altitude must not exceed 100 feet on
the altimeter.
(ii) Pressurized airplanes. Evacuate the static pressure system until a
pressure differential equivalent to the maximum cabin pressure
differential for which the airplane is type certificated is achieved.
Without additional pumping for a period of 1 minute, the loss of
indicated altitude must not exceed 2 percent of the equivalent
altitude of the maximum cabin differential pressure or 100 feet,
whichever is greater.

Regular system checks can then easily be done to verify that the air data
systems meet these requirements.

Total Temperature Probe Calibration

The total temperature equation in Eq. (3.23) was created assuming we
had a perfect adiabatic compression of the air and all the kinetic energy of
the airflow was converted into a temperature rise. From a measured total
temperature, we could then extract the static pressure using Eq. (3.25),
rewritten here for clarity as Eq. (4.19):
TT
T¼   ð4:19Þ
g1 2
1 þ M
2
As with all measurement systems, one should expect that errors exist in
the measurement of the temperature. Experience has shown that some of the
energy in the airflow is lost through kinetic heating (not recorded at the
probe) and some is added to the airflow inside the probe by the anti-ice
system. To account for this nonideal condition, we introduce a recovery
factor K in Eq. (4.19) and rewrite the equation as follows:
g1
TT ¼ T þ K T M2 ð4:20Þ
2
Using Eq. (4.20), one can now plot the total temperature recorded by the
sensor vs Mach squared (see Fig. 4.44). We can then extract from the linear
curve fit of the data both a static air temperature (at a zero Mach number) of
211.61 Kelvin and, from the slope (m ¼ ðg  1Þ T K =2Þ, a recovery factor
K ¼ 0.858. This, of course, assumes a constant air mass temperature,
which may not be the case for extended accelerations (both time and dis-
tance). The validity of a given test condition can be reflown to get a new
static air temperature (T, at M ¼ 0).
This approach also assumes that the recovery factor is constant through-
out the flight envelope, which is not always the case. A more complete
 CHAPTER 4 Altimetry Error 111

Pressure altitude (ft) Mach

45 0.94

0.92
44
Pressure altitude

0.90
(ft) (103)

43

Mach
0.88

0.86
42
0.84
41
0.82

40 0.80

Total air temperature (°C) Static air temperature (°C)

–20
Static air temperature (°C)
Total air temperature (°C)

–30

–40

–50

–60

–70
100 200 300 400 500
Time (s)

242

241

240
Total temperature (K)

239

238

Y = 36.337x + 211.61

237

236

235
0.6 0.62 0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8
M2

Fig. 4.44 Total temperature probe calibration.

112 Operational Aircraft Performance and Flight Test Practices

relationship can be expressed by a recovery correction defined as

TT  Tr
h¼ ð4:21Þ
TT
where Tr is the recovery temperature. The recovery temperature is the adiaba-
tic temperature of the air due to an incomplete recovery of the kinetic energy.
The relationship between the recovery correction and the recovery factor is
 
2
K ¼1h 1 þ ð4:22Þ
ðg  1ÞM2
Most total temperature probes properly located on an airplane and
aligned with the local flow while flying at a steady airspeed and altitude in
the normal operating range will measure a good temperature value Tm
that relates very closely to the total temperature of the air. That same
probe, however, can show a different value than the total air temperature
for several reasons. Figure 4.45 shows the relationship among the total air
temperature, the recovery temperature (typically lower than the total temp-
erature), and the measured temperature.
Factors that can impact the measured temperature Tm include the wake
of objects located in front of the temperature probe such as other air data
probes or even a landing gear (the entire flow field from near stall to
maximum operating speeds should be reviewed for local angle of attack),
as well as high off-angle flow. (The local angle of attack on a fuselage can
differ significantly from that of the airplane angle of attack.) The measured
temperature also can be influenced by thermal conduction (e.g., the tempera-
ture difference between the fuselage and the temperature sensor), thermal
radiation, and anti-icing heat. (Total temperature probes generally are anti-
iced to prevent icing in flight.) In addition, temperature sensors typically
have a delayed response to temperature changes (a lag) due to the heat
capacity of the sensor and its surrounding structure. As an example of all
the possible interactions, consider Fig. 4.46. It compares the total air

TT
Tr
Tm
Temperature

Mach

Fig. 4.45 Relationship among TT , Tr , Tm , and T.

 CHAPTER 4 Altimetry Error 113

Mach
0.36

0.34

0.32

0.30
Mach

0.28

0.26

0.24

0.22

0.20

Ship TAT (K) Exp TAT (K)

270

268
Ship TAT (K)
Exp TAT (K)

266

264

262

260
50 100 150
Time (s)

Ship TAT (K) Exp TAT (K)

270

268
Ship TAT (K)
Exp TAT (K)

266

264

262

260
0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13
Machˆ2

Fig. 4.46 Variation of TAT during a stall test.

114 Operational Aircraft Performance and Flight Test Practices

1
0.5

Ship KTAS – Exp KTAS 0

–0.5
–1
–1.5
–2
–2.5
–3
0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36
Mach

Fig. 4.47 KTAS difference as computed from the ship TAT and experimental TAT of
Fig. 4.45.

temperature (TAT) measured by the airplane ship system against the TAT
measured by an experimental probe. The experimental probe is not subject
to anti-ice heat, has a smaller lag constant, and is located on a nose boom
away from local flow disturbance.
Although differences between the two sensors seem large on this scale
and highly nonlinear, if the temperature of each probe is converted to a
static temperature probe using a constant recovery factor K for each probe,
the resulting error on KTAS, if the same reference Mach is used, is less
than approximately 1 kt (see Fig. 4.47). Given that the total air temperature
is used to derive a true airspeed and that the latter is simply used for naviga-
tion, this error is more than acceptable. If one were to use the true airspeed
derived from the ship probe to define lift or drag coefficients to define the
performance of a test airplane, then more scatter could be introduced into
the model.
As a final note, this information was derived for a subsonic airflow.
Supersonic airflows with the presence of shock waves introduce additional
corrections not covered in this book.

Probe Location: Manufacturing and Maneuver-Based Errors

The early transport category airliners with their more limited airspeed
and altitude range could afford more flexible installation of the air data
probes, but as the flight envelope grew with the advent of turbojets, more
care needed to be applied to such installations (see Fig. 4.48). This section
highlights only a few items of interest to show the impact of installation
and production tolerances on the baseline airplane. The reader should
expect variability from one plane to the next on the same type of design
 CHAPTER 4 Altimetry Error 115

based on this quick review. The manufacturers do strive to minimize these

deviations during the design stage (probe location selection) and by using
good manufacturing techniques and inspections. This is by no means an
extensive coverage of the subject, just enough to show production variability
on the reading of air data. The reader is encouraged to consult more detailed
references on the subject such as NASA Reference Publication 1046,
“Measurement of Aircraft Speed and Altitude” [1].
We will focus our discussion on the static port because small deviations in
the measured value have large impacts on airspeed and altitude. This author
likes to refer to Fig. 4.49 to show how the design of a static port is not as easy
as a simple tube open at 90 deg to the flow. The shape of the opening and how
it was installed on the airplane (repeatability of installation) will create small
deviations in the expected reading (quantified in pressure deviation as a func-
tion of airspeed, Dp/q). Note how small deviations (scale of x-axis) can create
significant pressure changes. To put numbers to this chart, consider the con-
dition for the protuberance at the rear (solid thick line). A step of 0.02 in. (the
thickness of a little more than five sheets of paper stacked) at 200-kt cali-
brated airspeed (q ¼ 135.2 psf) would produce an error of 0.02 Dp/q, or
Dp ¼ 2.7 psf (0.0187 psi). This is equivalent to about 35 ft error in altitude

Fig. 4.48 Evolution of flush fuselage-mounted static port.

116 Operational Aircraft Performance and Flight Test Practices

.06
Flow
.04 y
Protuberance at rear
.02

0 Protuberance at front
p
q –.02
Protuberance all around
–.04

–.06 Edge curved

–.08
Edges burred
–.10
0 .02 .04
.23 in. diam
y, in.

Fig. 4.49 Static pressure port edge impact to static pressure measurement [1].

at sea level, more than half of the allowable certification error (see Fig. 4.3
earlier in the chapter).
Another condition of interest is the deviation of the fuselage contour
from a nominal value. The manufacturer will use a test airplane to calibrate
the air data system. The location of the probe as well as the quality of build of
the fuselage where the probe is located now become the reference for the rest
of the production line. Two airplanes are never built exactly the same, so one
should expect small deviation of the contour from tail number to tail number.
Figure 4.50 gives an idea of the impact of such a deviation from the baseline
contour used for the initial calibration. The deviation is expressed in terms
of a wave. One notes that the closer the wave is to the static pressure port,
the larger the impact. One can then refer back to Fig. 4.47 to see how
some manufacturers have elected to address such manufacturing tolerance.
With this in mind, the airplane manufacturer can use simple guidelines
to select the best available aerodynamic location for the air data probes.
The manufacturer would consider the expected flow direction throughout
the flight envelope of the airplane; computational fluid dynamics (CFD)
has greatly helped in more recent times to provide such information at the
design stage, even prior to the first wind tunnel test. Air data probes,
especially the static port, must be located so as to be least influenced by
the following (also see Fig. 4.51):
• Doors, panels, radome, and so forth that are normally fixed in flight, but
that could be opened during airplane ground operation or servicing
W These will introduce service unknowns on the quality of the static

pressure reading.
 CHAPTER 4 Altimetry Error 117

11 in.
Vent x

.23 in. diam .09 in.

1.8 in. Long wave
Vent x

.04 in.
Short wave
Short
0
wave Long wave
1 wave, forward or aft
–.02
p 1 wave, forward or vent
q 2 waves, 1 forward, 1 aft
–.04
1 wave, aft of vent
2 waves, 1 forward, 1 aft
–.06
0 2 4 6
x, in. x = Distance from orifice to center of wave

Fig. 4.50 Contour deviation impact on static pressure measurement [1].

• Items that can be operated in flight such as landing gear and landing
gear doors
• The wake of other probes

Manufacturers may be faced with difficult choices of available locations

for air data probes at the design stage, especially for smaller planes where the
probes available to be used are not necessarily able to be scaled down to the
size of the plane. Probes need to be built to a minimum size to ensure they
can sustain the rigors of flight and maintenance. As this author likes to
remind his students, an airplane is a collection of systems fitted within an
aerodynamically and structurally efficient airframe. There may not always
be room to fit the probe at the ideal aerodynamic location. One should
strive to find an available location with the least pressure gradient (rate of

Probe with
static port

Radar
radome
seam

Avionics door Gear wake boundary

at low speed

Fig. 4.51 Locating the air data probes using tools like wind tunnel and CFD [5].
118 Operational Aircraft Performance and Flight Test Practices

Smooth contour of
engineering loft Joint at radome and
results in smooth variation fuselage causes pressure
in pressure disturbances in air

Probe
Probe location
location

Engineering loft contour pressure map Laser tracked surface of production

aircraft pressure map

Alternate
probe
Ridge build-up
location
to simulate worst-
case radome joint

Original probe AOA

location sensor

Fig. 4.52 In-service issue with probe location [2].

change of pressure vs distance along the flow line) for the entire flight envel-
ope. This means that the ideal aerodynamic (Dp/q ¼ 0) location may not
always be the best, even if available at the design stage. Consider the real-life
example shown in Fig. 4.52. The manufacturer had located the pitot-static
probe on the nose of this small business jet and had calibrated the probe
air data with its flight test airplane and certified the design per Part 23
requirements. The plane was then put into production and deliveries
began. As production planes came off the line and were checked under
RVSM conditions, a good number of planes were not meeting the strict alti-
metry requirements and had to be reworked. This cost the company enough
that it performed a further engineering investigation, which indicated that
the contour deviation at the junction of the fuselage and removable nose
cone could not be controlled enough in a production environment without
 CHAPTER 4 Altimetry Error 119

impacting production rate and cost. The company considered the cost of
relocating the probe and the requirement certification flight test against
the ongoing production cost and elected to proceed with the relocation of
the probe to a region where the fuselage contour could be more easily con-
trolled on the production line.
The location of the pitot probe can influence the reading of the total
pressure during a maneuver. For example, on a single-propeller-driven air-
plane, it would not be appropriate to locate the pitot probe on the fuselage,
behind the propeller disk; it would be subjected to the propeller wash and
would provide varying pressure information as a combination of airspeed
and engine power setting. An acceptable location for a pitot probe could
be the wing tip of the plane (see Fig. 4.53). Under normal flight conditions,
this location would allow the pitot probe to be well aligned with the
airflow, the bottom of the wing would help to align the airflow, and the
probe would be away from any object that could influence the pressure
reading; however, the probe is also now subjected to the nonsymmetric
maneuvers of the plane. For example, there will be an induced airspeed
delta whenever a sideslip rate is generated, with that delta speed being pro-
portional to the distance of the probe from centerline times the sideslip
rate. In a slow airspeed turn, where the radius of turn is relatively small,
the airspeed read by the pitot on the left wing would be smaller for a left
turn than for a right turn. (For the same turn rate, the inside wing flies
slower than the outside wing.)
The same goes for the location of the static pressure ports. Typically
located on the fuselage for a more predictable reading, on the same small
single-propeller-driven airplane (see Fig. 4.54), those pressure ports may
still be impacted by the airplane’s maneuvering (sideslip), speed (angle of
attack), and engine power setting. In this case, the airplane is taxiing
to takeoff position and then holds (up to time 1452); one can see the com-
puted altitude (measure of the static pressure). When the power is applied,
there is a noticeable increase in measured altitude, yet the airplane is still

Airspeed (V) Airspeed (V)

Sideslip () = 0

Airspeed (V)
˙
≈17 ft V – ˙ × (17 ft)

Fig. 4.53 Location of pitot’s impact on airspeed reading.

120 Operational Aircraft Performance and Flight Test Practices

static (as seen by the ground speed parameter). Following brake release, as
the airplane accelerates and the airspeed increases, the altitude reading
gradually returns to runway altitude (time 1472) just before start of rotation
and liftoff.

Prop wash

Static button Static button

Cirrus SR20 measured airspeed and altitude during takeoff – Example

100 100
Power
Pitch angle (deg), airspeed true (KTAS),

Taxi increase 50
80
0
ground speed (KTGS)

60
Pressure altitude (ft)

–50

40 –100
Brake
release Rotation
–150
20
–200
0
–250
High power low airspeed
air data error
–20 –300
1440 1450 1460 1470 1480 1490 1500
Time (s)
Airspeed true Ground speed Pitch Altitude press

Fig. 4.54 Example of impact of propeller flow on static pressure reading.

 CHAPTER 4 Altimetry Error 121

Relative
General thickness
Moon-shaped flow view
aerodynamic direction
compensation

Fig. 4.55 Aerodynamic compensation moon-shaped step.

Aerodynamic Compensation
It may not always be feasible to have a good air data probe location (pitot
and/or static) for all flight conditions; one should expect some level of inac-
curacies. It may sometimes be possible to adjust the pressure field near the air
data port, especially for the static pressure source, by adding or removing
aerodynamic shapes or even by contouring the probe itself. Figure 4.55
shows one such aerodynamic compensation where the airplane manufacturer
determined that, in this case, the probe generally read a little lower static
pressure than the ambient air pressure at a given flight condition. The moon-
shaped excrescence added to the pressure port slows down the flow near the
opening, thus increasing the static pressure felt. This is a similar approach to
the concept illustrated by Fig. 4.48.

Static Source Error Correction (SSEC)

The advent of air data computers (ADCs) provided airplane manufac-
turers with an additional means to tackle the issue of air data accuracy.
The manufacturer must still strive to best locate the static pressure ports,
pitot tubes, and other sensors, and it still needs to perform an appropriate
flight test campaign to identify the installation errors. What the ADC
offers is the possibility of applying a correction to the value read (static
pressure, total pressure, total temperature) before a value is computed
from these measured data. The most common corrections are typically
done to the static pressure; the expression used for these corrections is
static source error correction (SSEC).
To illustrate the benefit of SSEC, we use the data from Fig. 4.56 as a refer-
ence. The data collected throughout a given flight test program are plotted
against a Mach number for this example. Out of this, we identify a best fit
correction of Dp/q vs indicated Mach number. This correction, the SSEC,
is included in the ADC software such that for a given Mach number
measured (function of total pressure and static pressure), a correction to
the static pressure (the SSEC) is applied before the final calibrated values
are displayed to the flight crew. This calibrated value is now very close to
122 Operational Aircraft Performance and Flight Test Practices

0.0203

0.0202
SSEC
0.0201
p/q

0.02

0.0199

0.0198

0.0197
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9
Mach

0.00008

0.00006

0.00004

0.00002
Residual p/q

–0.00002

–0.00004

–0.00006

–0.00008
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9
Mach

Fig. 4.56 Example of residual error following the application of the SSEC.

the proper value that should be displayed; however, one must remember that
there are still some residual errors (difference between the measured and cor-
rected values) that are part of the value provided to the flight crew. The SSEC
still provides a means to greatly reduce the residual errors compared to the
baseline air data recording.

Sensor Accuracy, Precision, Resolution, and Calibration

Now for a quick word on the terminology used to describe sensor output.
Let’s start with the word accuracy, because we have used it a few times in this
 CHAPTER 4 Altimetry Error 123

chapter. Accuracy is defined as the amount of uncertainty in a measurement

with respect to an absolute standard. The precision of a sensor represents the
reproducibility of a measurement. An ideal sensor subjected to a given input
parameter (e.g., pressure, temperature) several times would always output
exactly the same value (indication, voltage, etc.). Real-life sensors, however,
will typically indicate close to but not exactly the same value for a given
input parameter. The tightness of the grouping of the output values
defines the sensor’s precision. Then, the difference between the average of
that grouping and the true value is the accuracy.
A sensor’s signal resolution is the smallest change in the input parameter
that can be detected and transformed into an output signal. This is particu-
larly true for analog-to-digital converters. The resolution is a relative quan-
tity. A sensor’s sensitivity is an absolute quantity and represents the
smallest amount of change that can be detected by measurement.
An ideal transducer will follow the change in the input parameter regard-
less of the direction of change of the value, whereas a real sensor will react
slightly different. This difference is the sensor’s hysteresis.
With all that said and the expected variation of the output signal, we now
need to convert the output signal into engineering units related to the param-
eter measured. An analog signal could require mechanical linkages to display
a value of airspeed or altitude (see Fig. 3.19 in Chapter 3), for example, whereas
a digital signal (typically voltage) is converted to a value like a pressure or an
altitude. One must perform a calibration of the signal to a usable engineering
unit. We reproduce part of §§25.1323 and 25.1325 here:
§25.1323 Airspeed Indicating System

(a) Each airspeed indicating instrument must be approved and must be cali-
brated to indicate true airspeed (at sea level with a standard atmosphere)
with a minimum practicable instrument calibration error when the
corresponding Pitot and static pressures are applied.

§25.1325 Static Pressure Systems

(d) Each pressure altimeter must be approved and must be calibrated

to indicate pressure altitude in a standard atmosphere, with a
minimum practicable calibration error when the corresponding static
pressures are applied.

To calibrate a sensor, such as one that records pressure, we must apply a

truth source to the sensor and record the output. To account for hysteresis,
the recording can be done with increasing and decreasing pressure. We
provide an example in Fig. 4.57 to help with the discussion. The input
(from the truth source) is in psi; the output, for this example, is in counts.
The dataset is then fitted with a calibration curve that best represents the
input signal (Fig. 4.57a), and here the fit appears very reasonable. This
curve is then a representation of the measured pressure converted to engin-
eering units of psi. This sensor is then put into use, and when a pressure is
124 Operational Aircraft Performance and Flight Test Practices

16

14

12
Pressure (psi)
10

8
y = 8.613E-05x3 – 1.2100E-02x2 + 7.3515E-01x – 2.4371E+00
R2 = 9.9971E-01
6

0
0 10 20 30 40 50 60
Counts

Calibration residual error

1
function – calibration reading) (psi)

0.8
Pressure error (calibration

0.6
0.4
0.2
0
–0.2
–0.4
–0.6
–0.8
–1
0 10 20 30 40 50 60
Counts

Fig. 4.57 Example of calibration error for a pressure sensor.

applied, a signal is generated (counts) that is converted into psi. The user of
the sensor should expect a calibration error, per Fig. 4.57b, in the signal con-
verted to pressure. In this case, the error is of the order of 0.1 psi, highest
at the lower count range, thus lower input pressure or equivalent to higher
altitude (above 25,000 ft) if this pressure is used to compute altitude.
This was presented to show some of the limits of accuracy one can expect
from the calibration exercise of a given sensor. One must understand the
limits of the accuracy of the sensor that is now used to collect data for our
performance models and thus how it impacts the value measured and how
it will introduce scatter in the creation of such models.
 CHAPTER 4 Altimetry Error 125

Flight Manual: Presenting the Information to the Flight Crew

Now that the airspeed and altitude errors have been determined, appro-
priate corrections applied to the pressures and temperature by various
means, and residual errors identified, one must now present the information
to the flight crew. The data must be provided for:
• Proper airspeed for takeoff/landing to ensure safe operational speed
• Navigation, understanding impact on time and range (proper KTAS and
SAT)
• Proper separation of altitude levels (proper Hp )

Part 25 certification requirements, per §25.1587, specify the following:

25.1587 Performance Information
(b) Each Airplane Flight Manual must contain the performance information
computed under the applicable provisions of this part. . . within the oper-
ational limits of the airplane, and must contain the following:

(6) Corrections to indicated values of airspeed, altitude, and outside air

temperature.

The simplest way to comply with the requirement could be a statement in the
airplane flight manual (AFM) saying, “The indicated airspeed on the display is
equal to calibrated airspeed within 1/2 kt.”
This works well when the air data and applied corrections lead to
an accurate airspeed (and altitude as required) indication. If slightly
larger corrections are needed, one could include a chart with a direct com-
parison of KIAS to KCAS for a given airplane configuration, as shown in
Fig. 4.58.
When the corrections to the air data system do not remove the pressure
errors as well as one would like, a slightly more complicated chart may have
to be created. One such format could take on the shape of Fig. 4.59.
Observe the notes provided in the chart of Fig. 4.59. The airplane con-
figuration was defined as well as the system to which the correction was to
be applied. It also clarified how to compute the calibrated speed and altitude
from the indicated values. Another acceptable means to present the infor-
mation to the crew is in table format, such as the two examples provided
in Fig. 4.60.
These formats are more than adequate for most of the flight envelope
of the airplane, but takeoff and landing airspeeds cannot require the crew
to first find the appropriate calibrated speed in one part of the AFM and
then convert that speed to an indicated speed to use for takeoff and
landing. This is especially true for takeoff where lag in the displayed airspeed
could lead to a significant difference between the calibrated airspeed of the
airplane and that used by the pilot, thus impacting the actual takeoff distance.
For this reason, manufacturers will provide those speeds directly in the
126 Operational Aircraft Performance and Flight Test Practices

350

300
Calibrated airspeed – knots

250

200 2

150

100 1
100 150 200 250 300 350
Indicated airspeed – knots

Fig. 4.58 Airspeed error chart.

indicated values in the AFM, accounting for residual position errors and
display lag.

Unreliable Airspeed (URA)

Despite having multiple calibrated airspeed indicators, there are still
some flight conditions and events that may lead to the airspeed indicators
providing erroneous values. When this occurs to only one system, the use
of three airspeed systems in the airplane allows the pilots to compare and
use the two systems that provide the same value (voting one system out).
When all systems provide unreliable airspeed, it not only could lead to
pilot loss of situational awareness, but also could affect other systems
relying on these airspeed sources. Such was the case for the Air France
AF447 accident [6]. What may make these events even more difficult to
diagnose is that they can be transient in nature, which may then trigger
a series of warnings and failures on the airplane that compound the
problem.
The airplane manufacturer should provide clear ways to help the crew
identify such events and provide procedures on how to deal with them
safely. Once an event is identified, the crew should, using procedures
 CHAPTER 4 Altimetry Error 127

defined by the airplane manufacturer, focus on managing the airplane’s

energy state within acceptable bounds. Having lost the airspeed indication,
other systems can be used that do not rely on the Pitot (the loss of reliable
total pressure, blocked Pitot, is often the source of URA) and/or static
systems.
The best alternative to airspeed is angle of attack, but this source is often
not available to the crew. The advantage of angle of attack indication is that it
provides a clue to the slow speed limit not to exceed and the margin to this
limit. It is not necessarily a good indication of the maximum airspeed not to
exceed. Alternative sources of guidance can include:
• Pitch attitude
• Engine setting
• Ground speed (GPS or inertial system)
• Altitude and altitude rate (pressure-altitude driven, but also GPS and
inertial system)

Press
correction, Vp - knots

3 Airspeed u re alt
i tude
Airspeed position

15 - 100
0 ft.
2 20
25
30
1 35
40
0 Notes:
1. Flaps UP gear UP, primary flight display
2. Calibrated airspeed = indicated airspeed + Vp
Altitude
3. True pressure altitude = indicated altitude + hp
100
Pre
ssure
90 alti
tud
15 e-
100
80 20 0 ft
correction, hp - FT.

.
Altitude position

25
70 30
35
60
40
50

40

30

120 160 200 240 280 320 360 400

Airspeed – knots IAS

Fig. 4.59 Airspeed and altitude correction chart example.

128 Operational Aircraft Performance and Flight Test Practices

• Airplane weight (best source, which can be a fuel quantity and known zero
fuel weight)
• Airplane configuration (flap position)

Concluding Remarks
In Chapter 3, you were introduced to the relationship among TAS, EAS,
and CAS (see Fig. 4.61). This chapter presented the possible sources of errors
in the measurement of the various parameters needed to compute these
three speeds. We also introduced two new speeds. Indicated airspeed (IAS)
may include instrument errors and probe position errors affecting its

Airspeed indicators system – Flaps up, gear up

KIAS KCAS KIAS KCAS
55 60 130 129
60 65 140 139
70 74 150 148
80 83 160 157
90 92 170 166
100 102 180 176
110 111 190 185
120 120 194 (VNE) 189

Airspeed position error correction – v

Indicated Flaps/Gear up or down
airspeed
(KIAS) 10° 20° 30°
100 0.5 1.0 1.0

120 0.5 1.0 1.0

140 0.5 1.0 1.0

160 0.5 0.3 0.1

180 0.3 0.0 –0.3

200 0.2 0.0

210 0.2 0.0

v = Calibrated airspeed – indicated airspeed (knots)
a positive value indicates that the indicator reads low

Fig. 4.60 Examples of table format for air data correction.

 CHAPTER 4 Altimetry Error 129

TAS
EAS
=
TAS

EAS
EAS
F=
CAS

CAS
Position errors (steady state) IAS = CAS – VPE – Vi
acoustic & pneumatic lags
leaks IAS

Display processing lag

DAS

Fig. 4.61 Transformation from true airspeed to displayed airspeed.

relationship with CAS. We also discussed the impact of electronic filtering on

the displayed airspeed (DAS).
The aim of this chapter was to introduce you to the accuracy limit of the
information provided to the crew. The same limits must be understood by
the engineers using the data to derive the performance models for the air-
plane. The residual errors carried by the airplane’s production air data, or
even the experimental system, will be transferred to the performance
models. Variability from plane to plane for the same type of design will
also complicate the comparison of performance between two otherwise
identical airplanes.
One must remember that the truth source could have errors, the test
execution induces errors, and even calibration can have residual errors—
calibrated does not mean it is perfect. The engineers should always be pre-
pared to question the validity of an air data source and know how to
troubleshoot and identify the source of the error: in search of the truth!

Exercise
1. A flight crew is requested to perform an altitude calibration of the ship
systems (two primary air data systems and one standby system) using a
trailing cone. The condition of interest is 200 KIAS and 15,000 Hpi . The
following data were recorded (Fig. 4.62). Pick a stable condition and
compute the static pressure error in terms of Dp/qci [delta pressure
production system (experimental system) over production indicated
impact pressure]. Assume the trailing cone reading is the best available
source. Hp1 and KIAS1 are production primary system 1, Hp2 is primary
system 2, Hp StdBy is the standby system. The reference speed (for qci ) is
production 1.
130 Operational Aircraft Performance and Flight Test Practices

KIAS1 (kts)
202

201
KIAS1 (kts)

200

199

Hp1 (ft) Hp2 (ft) Hp StdBy (ft) Trailing cone Hp (ft)

Hp1 (ft), HP2 (ft), Hp StdBy (ft)
trailing cone Hp (ft) (103)

15.04
15.02
15.00
14.98
14.96
14.94
10 20 30 40
Time (s)

Fig. 4.62 Data for exercise 1.

References
[1] Gracey, W., “Measurement of Aircraft Speed and Altitude,” NASA Reference
Publication 1046, May 1980.
[2] Downen, C. M. “Reduced Vertical Separation Minimum Certification for the Premier
I Business Jet,” AIAA 2003-6783, 2003.
[3] Ihtiari, P. A. and Marth, V. G., “Trailing Cone Static Pressure Measurement Device,”
Journal of Aircraft Engineering Note, Vol. 1, No. 2, 1964.
[4] Gracey, W., “Summary of Methods of Measuring Angle of Attack on Airplane,”
NACA TN-4351, 1958.
[5] Fujino, M., “Design and Development of the HondaJet,” Journal of Aircraft, Vol. 42,
No. 3, May-June 2005.
[6] Bureau d’Enquêtes et d’Analyses pour la sécurité de l’aviation civile, “Final Report on
the accident on 1st June 2009 to the Airbus A330-203 registered F-GZCP operated by
Air France Flight AF 447 Rio de Janeiro - Paris,” July 2012, https://www.bea.aero/
docspa/2009/f-cp090601.en/pdf/f-cp090601.en.pdf [retrieved 23 Sept. 2020].
Chapter 5 Weight and
Balance

Chapter Objective
It’s all about the weight! This chapter introduces students to the importance of
weight for airplane performance. This is a vast subject and is covered only in
terms of the information that would be used by an airplane performance engin-
eer. We will give examples, review accidents/incidents related to weight and
center of gravity, present some definitions, address basic weight and balance
theory, discuss weighing procedures, and describe the impact of a weight mis-
computation on performance models and operational safety. This chapter will
also address the subject of allowable weight and center of gravity envelope.

Introduction

W
eight may seem like a trivial subject; however, for an airplane,
weight is one of the most important parameters in aircraft per-
formance. It has an impact on all phases of flight:
• It dictates the amount of lift that must be generated.
• In turn, the induced drag (drag due to lift) will increase proportionally to
the square of the lift coefficient required to sustain the flight.
• Trim drag and aircraft handling (safety) depend on the location of the
center of gravity.
• The thrust required to perform a mission is directly proportional to the
drag of the aircraft.
Typically, an aircraft with a higher gross weight, as compared to a lighter
one, everything else being equal, will have:
• Increased takeoff speed and distance
• Increased stall speed
• Increased landing speed and distance
• Decreased rate of climb and service ceiling
• Increased fuel burn
• Decreased range

131
132 Operational Aircraft Performance and Flight Test Practices

Incorrect computation or measurement of the aircraft weight will have a

direct impact on the performance model developed and/or the safety of flight
of the aircraft. For example, consider the following accident, NTSB case
DCA03WA030.

NTSB DCA03WA030: On 12 March 2003, a Boeing 747-400 from Singapore

Airlines, tail number 9V-SMT, had a severe tail strike on takeoff while depart-
ing from Auckland International Airport, New Zealand, where the tail of the
airplane scraped the runway for nearly 1600 ft before the airplane became air-
borne. The aircraft circled for 20 min and then landed safely back at Auckland.
There was severe damage to the tail structure (see Fig. 5.1). A postflight review
of the takeoff incident revealed an error in the takeoff weight transcription,
where a weight of 247.4 tons was used instead of the actual airplane weight
of 347.4 tons. This transcription error led to takeoff airspeed miscomputation
with a rotation speed nearly 33 kt lower than the normal 163 kt for the actual
weight. It also resulted in the crew applying lower thrust than required. During
the accident, the airplane lifted off at nearly stall speed, and the stall warning
(stick shaker) activated for nearly 6 s during the takeoff.

Fig. 5.1 Results of tail strike accident, NTSB DCA03WA030.

 CHAPTER 5 Weight and Balance 133

Weight Breakdown and Definitions

Although it is the total weight of the airplane that affects the flight
condition being evaluated, one must expect that the airplane’s weight
will fall between a maximum and a minimum value. As the accident
just described shows, it is also necessary for a flight crew to get a good
understanding of the weight range they should expect when planning to
take off, or for any other flight regime, so as to validate the speed they
intend to use. The definitions used here are as provided in FAA Advisory
Circular (AC) 120-27E [1].

1) Empty Weight (WE ): the weight of the airframe, engines, propellers,

rotors, and fixed equipment. It excludes the weight of the crew and
payload but includes the weight of all fixed ballast, unusable fuel,
undrainable oil, and total quantity of hydraulic fluid.
a) Some aircraft manufacturers may elect to define the WE of the airplane
with full oil reservoir rather than undrainable oil.
b) There may also be a manufacturing empty weight (WME ) defined. This
allows a manufacturer to fly an airplane before completing the build of
its interior; you do not need passenger seats to flight test a business jet
for example, just crew seats.
2) Maximum Zero Fuel Weight (MZFW): the maximum permissible weight
of an aircraft with no disposable fuel and oil. This is a structural
design limit.
3) Maximum Landing Weight (MLW): the maximum weight at which the
aircraft may normally be landed.
a) Manufacturers will provide special procedures in the
airplane flight manual (AFM) to provide guidance to pilots when they
need to perform a landing with a weight larger than MLW (overweight
landing procedure). There may also be special instructions in the
airplane maintenance manual (AMM) for an inspection after such an
event (overweight landing inspection).
4) Maximum Take-off Weight (MTOW): the maximum allowable, total
loaded aircraft weight at the start of the take-off run, at brake release.
5) Maximum Ramp Weight (MRW): the maximum allowable, total loaded
aircraft weight for taxi.

Table 5.1 are some values of weight extracted from the Type Certificate Data
Sheets (TCDSs) of various airplanes.
Some certification regulations, such as FAA 14 CFR 25.1523, Mini-
mum Flightcrew, define the minimum number of pilots required to
operate an airplane, whereas operational rules, such as FAA 14 CFR
121.385, Composition of Flight Crew, and 121.391, Flight Attendants,
may define the minimum number of crew members required to operate
an airplane. It is therefore useful to provide another definition. The
operating weight empty (OWE) is the empty weight of the airplane plus
the weight of the crew and other minimum equipment required to
operate it.
Now, looking at it from a pilot’s and/or a load master’s point of view, they
start with an airplane empty weight WE . Then the operational items and the
134 Operational Aircraft Performance and Flight Test Practices

Table 5.1 Weights of Various Aircraft

OEM Airbus Boeing Bombardier Gulfstream Syberjet

Model A320-100 737-700 Global 6000 GV-SP SJ30-2
TCDS A28NM A16WE T00003NY A12EA A00001AC
MRW (lb) 150,820 155,000 99,750 91,400 14,050
MTOW (lb) 149,940 154,500 99,500 91,000 13,950
MLW (lb) 138,915 129,200 78,600 75,300 12,725
MZFW (lb) 130,100 121,700 56,000 54,500 10,500
Min weight (lb) 81,083 — 48,200 — 9000
Min crew 2 pilots 2 pilots 2 pilots 2 pilots 1 pilot
Cert basis Part 25 Part 25 Part 25 Part 25 Part 23
MRW ¼ maximum ramp weight; MTOW ¼ maximum takeoff weight; MLW ¼ maximum landing weight;
MZFW ¼ maximum zero fuel weight.
Note: Min weight is not the same as empty weight. It is the minimum weight for which the manufacturer will
define performance numbers and is not always provided in the TCDS.
 The SJ30-2 weight is the minimum takeoff weight.

crew are loaded into the weight and balance sheet to get the OWE. Then
comes the tricky part—the pilot and/or load master balance the payload (pas-
senger, bags, cargo, etc.) that is desired to be carried vs the fuel weight that
must be loaded to carry out the mission. The payload weight, once loaded,
must not bring the zero fuel weight (ZFW) of the airplane above the
MZFW authorized by the design.
The fuel load must include a reserve (one the pilot does not plan to use
for the flight) and the sum of the ZFW plus reserve fuel, which equals the
planned landing weight (LW), must not exceed the MLW of the airplane.
Then the pilot must determine the required amount of fuel to carry out
the mission based on the airplane configuration and payload. The weight
of the plane at takeoff (TOW), with the payload and fuel loaded to
perform the mission, cannot exceed the MTOW of the airplane. Finally,
the pilot must consider some fuel for engine start and taxi to the start of
the runway, and this ramp weight (RW) cannot exceed the MRW. This
fuel quantity must be sufficient to account for potential extra fuel burn
due to delays on the taxiway on the way to the runway and still have
enough fuel for the mission, yet it can’t be too much that the airplane gets
to the start of the takeoff run with a total weight exceeding MTOW.
Figure 5.2 shows the weight buildup.

Aircraft Weighing: Theory and Practice

As you will see throughout this book, the weight of the airplane has a
major impact on its performance. For two airplanes having similar aerody-
namic characteristics and engine performance, the one with the lowest
 CHAPTER 5 Weight and Balance 135

RW
Taxi out
TOW

Trip Fuel

LW
Reserve
ZFW
Cargo
Payload
Checked bags

Pax + Carry-on

Ops item

OWE
WE

Fig. 5.2 Weight buildup.

weight will generally have the best overall efficiency in terms of fuel burn
(economics). As well, an airplane must not exceed the design maximum
and minimum weights it was certificated to, because these have been
shown to be safe within a set of certification requirements.
So, how do you weigh an airplane?
One method involves portable roll-on scales. The scales are located in
front of each landing gear, and the airplane is rolled on, typically by a
tow truck. These scales have a very low profile with small ramps on
either side so the airplane can easily climb on them. This affords great flexi-
bility because they can be adapted to airplanes of different sizes and landing
gear layouts; however, one must ensure the scales used are appropriate for
the expected weight range of the airplane. For small planes, with one tire
per gear, three scales are typically required. For larger planes with more
than one tire per landing gear, if the scales are not large enough to cover
all tires on a given gear, then multiple scales per gear are used (see
136 Operational Aircraft Performance and Flight Test Practices

Fig. 5.3 Airplane weighing.

Fig. 5.3). The total airplane weight W is simply the sum of all the scales’
weight Wi .
P
W ¼ Wi (5:1)

Another method uses permanent in-ground scales (see Fig. 5.4). These
are laid out in the ground for the expected airplane size and landing gear con-
figuration; they are generally flush with the ground, making it easier to tow
the airplane onto them. These scales are not as flexible as portable scales
in that the geometric distribution of the scales is fixed, but these are typically
used by fleet operators and airplane manufacturers where the airplane geo-
metry and weight range are well defined.
The last method covered in this class uses load cells sandwiched between
the airplane and ground jacks (see Fig. 5.5). Typically, three jacks are used to

Fig. 5.4 Airplane roll-on scales. Source: Left: http://jandascales.com/aircraft;

right: http://www.gecscales.com/products/aircraft-weighing-scales
 CHAPTER 5 Weight and Balance 137

Fig. 5.5 Airplane weighing with load cells on jacks.

support the plane—two under the wings at a given fuselage station (distance
along the longitudinal axis) and one near the nose. The locations where the
jacks support the airplane are fixed, and the structure of the airplane at these
locations is designed to take on the concentrated load (often more than one
third the total airplane weight) distributed over 1 to 2 in.2 .
The FAA offers Handbook FAA-H-8083-1B [4] that provides good
general reference material on weighing procedures.

Center of Gravity: It’s a Balancing Act

The center of gravity (CG) of an airplane is the average location of the
weight where, if the airplane were supported exactly at that position, it
would be in equilibrium and not tend to rotate. For the majority of airplanes,
the longitudinal center of gravity is the most important, and often the only,
center of gravity that requires tracking; the vertical and lateral CGs are dis-
cussed in the following sections. When one refers to the center of gravity
of an airplane, is it almost always the longitudinal value; if not, the terms ver-
tical or lateral are generally used.
The basic principle to determine the location of the center of gravity of an
airplane is simple: For the airplane to be in equilibrium about all points on
which it rests (landing gear, jacking points, etc.), the sum of the moments
about a given reference position must be equal to zero; that is, the sum of
the product of a given weight Wi applied at a given position xi , such as a
138 Operational Aircraft Performance and Flight Test Practices

landing gear, a jack, or the like, about the center of gravity must be equal to
zero.
P
Wi xi jCG ¼ 0 (5:2)
If the reference position is moved to a datum line away from the center of
gravity, then the equation takes the following form:
P
W  xCG – Wi xi ¼ 0 (5:3)
where
W ¼ total weight of the airplane
xCG ¼ distance of the center of gravity from the datum line
Wi ¼ weight measured at a given position xi away from the datum line.

Example 5.1
You need to determine the longitudinal center of gravity of a business jet on
which your company just completed some maintenance and upgrade work.
The airplane flight manual (AFM) provides a weighing procedure for
roll-on scales. The airplane has two main landing gears with two wheels per
main, and one nose landing gear that also has two wheels. You have roll-on
scales that are large enough to accommodate the two wheels, so you will
need only three scales. The AFM specifies that the datum point should be
established 40 in. in front of the nose cone. The AFM goes on to say that,
once the airplane is on the roll-on scales, you must measure the distance
from the datum line to the nose gear and from the datum line to a line
joining the two main gears (see Figs. 5.6 and 5.7).

Datum

XNLG
XCG

XMLG

Fig. 5.6 Finding the CG of an airplane.

(Continued)
 CHAPTER 5 Weight and Balance 139

Example 5.1 (Continued)

Datum

MAC LEMAC

Fig. 5.7 Example of an airplane datum line location

Solution: You set the scales up and allow them to warm up properly, and
then you pull the airplane onto the scales per the manufacturer’s procedure.
You record the values and weights shown in Table 5.2.

Table 5.2 Measured Distances and Weights

Item Arm Weight

Nose landing gear XNLG ¼ 132.5 in. 1747 lb
Main landing gear right XMLG ¼ 461.6 in. 8124 lb
Main landing gear left XMLG ¼ 461.6 in. 8075 lb
Total — 17,946 lb

You have established the weight, 17,946 lb; now, where is the center of
gravity? You proceed to compute the moments from all measured points
(see Table 5.3).
140 Operational Aircraft Performance and Flight Test Practices

Table 5.3 Computing the Moments

Item Arm (in.) Weight (lb) Moment (in. . lb)

Nose landing gear 132.5 in. 1747 231,477.5
Main landing gear right 461.6 in. 8124 3,750,038.4
Main landing gear left 461.6 in. 8075 3,727,420.0
Total — 17,946 7,708,935.9

The location of the center of gravity is then simply the total moment
divided by the total weight.

XCG ¼ (Total Moment)=ðTotal WeightÞ

¼ (7,708, 935:9 in:  lb)=(17,946 lbs) ¼ 429:6 in:

So the CG is 429.6 in. behind the datum line.

Some notes:
• Typically the airplane needs to be level within +1 deg (flat floor with a
slope of less than 0.2 in./ft), recognizing that while on a jack, for example,
the airplane may not be in an attitude that provides a flat floor. But it must
be per the manufacturer’s weighing procedure.
• To establish a valid basic empty weight (BEW), the airplane should have
been emptied of all fuel (including fuel sumping).
• All other fluids on board should be per AFM weighing procedure.
The distance from the datum point along the length of the airplane (its
longitudinal axis) is typically referred to as the fuselage station (FS). From
a maintenance support point of view, FSs are easy to track than MAC and
provide a concrete way to find positions on the airplane. From an aerody-
namic point of view, a more convenient percent mean aerodynamic chord
(MAC) position is usually employed.
To convert fuselage stations to percent MAC, one must know where the
leading edge of the MAC is located in terms of FS. Then, the following
equation can be used:

ðbalanced arm  LEMACÞ

percent MAC ¼  100 (5:4)
MAC
For Example 5.1, the MAC is 7.5 ft long (90 in.). If the leading edge of the
MAC (LEMAC) is at FS 389, then converting the CG position (429.6 in.) to a
percent MAC value would translate in a CG position equal to

ð429:6  389Þ
%MAC ¼  100 ¼ 45:1%
90
 CHAPTER 5 Weight and Balance 141

Example 5.2
In the same airplane as Example 5.1, you load two pilots at 200 lb each and
sitting at FS 135 plus four passengers (two at FS 350 and two at FS 400), all
at 170 lb. What is the new CG of the plane?

Solution: Starting from the known basic empty weight of the airplane, one
must now add the new weights of the six occupants plus their respective
moments (weight times longitudinal moment arm). The results are shown
in Table 5.4.

Table 5.4 Moments When Including Passengers

Item Arm (in.) Weight (lb) Moment (in. . lb)

Basic empty weight 429.6 17,946 7,708,935.9
Pilots 135.0 2  200 ¼ 400 54,000
Passengers (forward) 350.0 2  170 ¼ 340 119,000
Passengers (aft) 400.0 2  170 ¼ 340 136,000
Total — 19,026 8,017,935.9

Therefore the new CG position is at

8,017,935:9 in  lb
¼ 421:4 in: or
19,026 lb
ð421:4  389Þ
%MAC ¼  100 ¼ 36:0%
90

Example 5.3
Assume you are now going to weigh the airplane from Example 5.1 on a jack
with load cells and gear off the ground but not retracted. What would be the
expected load on each jack if the nose jack (L1) is at FS 200, and the underwing
jacks are at FS 465 (L2, combine both jack loads into a single total load at the
FS location)?

Solution: Writing the moment balance equation from Fig. 5.8, and setting
the value equal to zero (equilibrium), one gets
CG  W  x1  L1  x2  L2 ¼ 0

(Continued)
142 Operational Aircraft Performance and Flight Test Practices

Example 5.3 (Continued)

Datum

L1
CG

L2

Fig. 5.8 Finding the CG of an airplane on jack.

We also know that the sum of the load on the jacks will equal the airplane
weight, therefore,
L1 þ L2 ¼ W
We have two equations and two unknowns, which we can solve as follows.
Let
L1 ¼ W  L2
Then,
W  ðCG  x1 Þ 17,946  ð429:6  200Þ
L2 ¼ ¼ ¼ 15,548:7 lb
ðx2  x1 Þ ð465  200Þ
The total load for the underwing jacks, and
L1 ¼ 17,946  15,548:7 ¼ 2397:3 lb
for the nose jack.

The impact of the airplane attitude on the

computed center of gravity can easily be seen in Note: To accurately measure
Fig. 5.9. The reference location (fuselage station) an airplane’s longitudinal
for the landing gears or jacking points are defined center of gravity, the airplane
must be properly leveled per
based on a longitudinal axis that typically runs paral-
the manufacturer’s
lel with the floor of the cabin or a reference indication instructions.
on the airplane. If the airplane is properly leveled,
 CHAPTER 5 Weight and Balance 143

Reference FS
attitude to
weight
W

L1 L2

FS

W FS
L1 W
L2 L2
L1
Nose high Nose low

Fig. 5.9 Impact of not leveling the airplane prior to weighing.

one will determine the proper CG because the procedure requires a weight
vector (remember, we measure weight, not mass) perpendicular to the refer-
ence longitudinal axis.
Figure 5.9 shows that changing the airplane attitude moves the center of
gravity with respect to the measuring points L1 and L2 . In the case of the
nose-high condition shown, we can even see that the weight vector moved
behind the main landing gear (MLG) L2 measurement position, which
would typically result in the airplane tipping over onto its tail.

Lateral CG
Note that the airplane weighed in Example 5.1 was not perfectly sym-
metric left to right; in this case, the right main landing gear was carrying
more weight than the left. Most airplanes are designed to be laterally sym-
metric, with items added to one side of the airplane that are generally
matched in moment with items on the other side of the airplane. This sim-
plifies the aerodynamic design of the airplane. A small imbalance can still
remain, as measured in the prior example. Because of this inherent lateral
symmetry from a weight distribution point of view, and because airplanes
are generally built aerodynamically symmetrical, the operator of an airplane
is generally not required to measure the lateral center of gravity of
the airplane.
Figure 5.10 presents an example of a possible lateral center of gravity
envelope. If we use the data from Example 5.1 and apply a lateral main
landing gear position of 88 in. from the center line, the resulting moment
imbalance for the airplane as weighed is (8124 lb –8075 lb)  88 in. ¼ 4312
in. . lb. This imbalance is plotted on Fig. 5.11.
144 Operational Aircraft Performance and Flight Test Practices

28,000

26,000

24,000
Airplane weight (lb)

22,000
Flight &
ground
20,000 operation

Ground
18,000 operation
only

16,000
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000
Lateral unbalance (in. • lb)

Fig. 5.10 Lateral center of gravity limit envelope.

The lateral CG can be computed, just as we did with the longitudinal CG,
by dividing the imbalance moment by the total airplane weight. For Example
5.1, the lateral CG would be at (4312 in. . lb)/(17,946 lb) ¼ 0.24 in. off center,
towards the right main landing gear.
With most airplane configurations, the bulk of the payload and fixed
weights are located in the fuselage, which lies in the plane of symmetry, so

WfR
WfL

YfR YfL

Fig. 5.11 Lateral fuel imbalance.

 CHAPTER 5 Weight and Balance 145

there is very limited impact of moving the payload around on the lateral
center of gravity. On the other hand, one thing that must still be tracked con-
tinuously in flight is the lateral distribution of fuel. The fuel weight typically
represents a large percentage of the airplane takeoff weight, and it is usually
carried by the wing. This means an imbalance of left to right will generate
larger rolling moments, pound for pound, than a payload movement in
the fuselage.
This rolling moment must be compensated for by aileron, and possibly
wing spoiler, deflections. These deflections will generate additional drag
over the baseline, balanced airplane configuration, which will have a direct
impact on the airplane’s performance. Some agencies, like Transport
Canada (TCCA AC525-013) offer special guidance material to address hand-
ling characteristics with lateral CG [3].

Vertical CG
Just like the lateral CG, the vertical CG of an airplane is usually not some-
thing the operator will track because the normal operation of the airplane will
result in a loading that will keep this value within a very small range, well
within the boundary limits of the airplane.
The vertical CG has an impact on the airplane flying characteristics; for
example, a high vertical CG for a given longitudinal CG location can result
in an apparent increase in the aft CG movement as the angle of attack
(AoA) of the airplane is increased. This is illustrated in Fig. 5.12.
An additional impact of the vertical CG location is on the airplane’s longi-
tudinal trim requirements. The further away from the CG the engines are, the
greater the impact on the trim requirements. A high thrust line as compared
to the vertical CG will result in an increasing trim requirement, nose up, with
increasing thrust (increasing thrust results in increased nose-down pitch
moment); the effect is reversed if the thrust is reduced. Increased nose-up
trim also leads to increased trim drag, which has a direct impact on the per-
formance of the airplane (see Fig. 5.13). A low thrust line with respect to the
vertical CG has the opposite effect.
To measure the vertical CG, an extra step must be added to the baseline,
level weighing of the airplane—we are, after all, measuring the weight, and

␣ AC ␣ AC

FS
W FS
W
Low CG High CG

Fig. 5.12 High vertical CG impact on apparent longitudinal CG at high AoA.

146 Operational Aircraft Performance and Flight Test Practices

therefore a force acting along the vertical, not the mass of the airplane. The
step consists of changing the airplane pitch attitude by a large enough value
(5–10 deg) while rotating about a fixed point (the MLG) so as to measure a
reasonable change in the nose landing gear load (reduction); see Fig. 5.14.
Using the MLG as the reference datum point, the vertical CG can then be
determined as follows:
ðL1  L1b Þ DFS
ZCG ¼ (5:5)
W tanðuÞ
Some of the risks involved in doing this exercise are that the airplane may
start rolling due to the fuselage attitude or it could tip over if the longitudinal
CG moves over the landing gear point.

Mass Moment of Inertia

The mass moment of inertia I is a measure of the extent of an object
resisting a rotational acceleration about a given axis. The units for the
mass moment of inertia are [mass  length2 ]. Although we do not discuss
much about mass moment of inertia in airplane performance, we must

Thrust

Thrust

Thrust

Fig. 5.13 Thrust line vs vertical CG.

 CHAPTER 5 Weight and Balance 147

W
L1 L2
⌬FS

W
L1b
L2b

Fig. 5.14 Two steps are required for vertical CG determination.

note that it has an impact on the resulting performance of the airplane in

dynamic conditions such as takeoff rotation or flare capability of the airplane.
Computing the mass moment of inertia of an airplane is one of the
responsibilities of the weights engineer.

Fuel Weight and CG

Fuel burn is usually the single most important weight change in flight,
representing typically 20– 40% of the plane’s takeoff weight. The locations
of the fuel tanks are therefore very important in terms of the airplane
center of gravity in flight while the fuel is burned. Figure 5.15 represents a
typical fuel tank distribution on a small business jet. Such a system can be
relatively simple, using mostly gravity to feed a fuselage tank to the main
wing with some booster pumps available to pick up as much of the fuel in
the tank as possible. In this example, the moment generated by a given
amount of fuel is referenced to the wing 25% MAC point, a negative
moment producing a forward movement of the CG and a nose-down pitch-
ing moment. The burning of fuel from the fuselage tank first reduces the
nose-up moment generated by the fuel weight. Then the fuel burned from
the wing fuel produces almost no movement of the center of gravity.
Larger, longer-range airplanes may have a more complex fuel system that
requires significant fuel management, whether manual or automatic, to
ensure that the center of gravity of the airplane remains within certified
148 Operational Aircraft Performance and Flight Test Practices

5000
Wing 25% MAC

4500
Reference

4000
3500
Fuel weight (lb)

3000
2500
Fuselage fuel
2000
Wing fuel
1500
1000
500
0
–200 –100 0 100 200 300 400 500
Moment/100 (in. • lb)

Fig. 5.15 Typical fuel tank locations on small business jet and fuel moment chart.

limits. Figure 5.16 shows such an example in which the airplane has several
wing and fuselage tanks as well as a horizontal tail trim tank.
Note that the maximum fuel quantity that can be carried by an air-
plane will be based either on the maximum allowable takeoff weight or,
if the airplane is carrying a light payload, on the maximum fuel volume.
When the airplane is limited by the fuel volume, another factor comes
into play—the density of the fuel varies according to its temperature.
Figure 5.17 shows typical variation in the specific weight of three different
fuels/gasolines with respect to their temperature. This represents the
average specific weight based on the fuel/gasoline specification. One can
expect that a given fuel type also has a small density variation at a
given temperature; that is if one uses the correct weight units for the
fuel volume measured (see Gimli glider accident box).
 CHAPTER 5 Weight and Balance 149

Center tank
120,000
Inner tank Inner tank Additional center tank

100,000 Combined trim tank

Outer tank and center tank
Outer tank
Vent
80,000 Vent tank
tank
Additional
Fuel weight

center tank
60,000 Inner tanks, part 3
Trim tank

40,000 Trim tank, part 1

Inner tanks, part 2

20,000
Outer tanks
Inner tanks, part 1
0
–25 –20 –15 –10 –5 0 5 10 15 20 25
Moment index

Fig. 5.16 Long-range airliner complex fuel system and fuel management
impact on longitudinal CG.

7.5

7.3

7.1

Jet A (JP-8) & Jet A-1

6.9
Specific weight (lb/U.S. Gal)

6.7

Jet B (JP-4)
6.5

6.3

6.1
AvGas

5.9

5.7

5.5
–40 –20 0 20 40 60
Fuel temperature (˚C)

Fig. 5.17 Average aviation fuel density.

150 Operational Aircraft Performance and Flight Test Practices

On 23 July 1983, a new Boeing 767 from Air Canada was on a scheduled flight
from Montreal to Edmonton with a scheduled stop in Ottawa. The onboard
fuel quantity indication system that converts the measured fuel volume and
fuel temperature into a fuel weight was inoperative. The flight crew asked
the ground crew to verify the fuel quantity using a procedure known as
dipping (inserting a dipstick into the fuel tank and computing an equivalent
volume). The ground crew did so and provided the value back to the flight
crew. The crew then used this value by using the factor 1.77, which they
had always used before. This factor is equivalent to 1.77 lb of fuel per liter.
The problem was that the brand new 767 was all metric, and the crew
should have used 0.8 kg per liter. The crew overestimated the fuel weight by
a factor of more than two. The error occurred again at their stop in Ottawa,
where the fuel volume measured was 11,430 l, which they converted to
20,400 kg of fuel; in reality, they had only 9144 kg, about half the fuel
needed to reach Edmonton. While cruising at 41,000 ft and 469 KTAS, the air-
plane ran out of fuel. The flight crew was able to “fly” the plane (gliding) down
to the remote airfield of Gimli, which had been partially transformed into a
race car strip. The incident has been coined the Gimli Glider Incident.

The following is another accident involving fuel, this time caused by a

longitudinal CG issue.

NTSB accident number CHI01MA006: On 10 Oct. 2000, at 1452 central

daylight time, a Canadair Challenger CL-600-2B16 (CL-604) was destroyed
on impact with terrain and postimpact fire during initial climb from runway
19R at Wichita Mid-Continent Airport (ICT), Wichita, Kansas. The flight
was operating under the provisions of 14 Code of Federal Regulations
 CHAPTER 5 Weight and Balance 151

(CFR) Part 91 as an experimental test flight. The pilot and flight test engineer
were killed. The copilot was seriously injured and died 36 days later. The
National Transportation Safety Board determined that the probable cause
of this accident was the pilot’s excessive takeoff rotation during an aft
center of gravity (CG) takeoff, a rearward migration of fuel during acceleration
and takeoff, and the consequent shift in the airplane’s aft CG to aft of the aft
CG limit, which caused the airplane to stall at an altitude too low for recovery.
Figure 5.18 shows the fuel tanks layout from the NTSB report.

Forward
Left main auxiliary
tank Center tank
auxiliary Collector
tank tanks

Aux tank system

Tail tank system
Main tank system

Tail
cone Right main tank
tank
Right Aft
Left saddle auxiliary
Fuel
saddle tank tank
dump
mast tank

Fuel tank system

Fig. 5.18 Challenger 604 fuel tanks layout, from NTSB report [5].

Not all fuel in the airplane’s fuel tanks can be used for a flight. Fuel, being
a liquid, tends to settle in the low points. Then it moves according to the
accelerations applied on the airplane and its various attitudes. The more
complex the fuel tank system is, the more chances of not getting all the
fuel out to be used by the airplane’s engines. The certification requirements
stipulate that the manufacturer establish the point where the remaining fuel
cannot be used (the unusable fuel).

§25.959 Unusable fuel supply.

The unusable fuel quantity for each fuel tank and its fuel system components
must be established at not less than the quantity at which the first evidence of
engine malfunction occurs under the most adverse fuel feed condition for all
intended operations and flight maneuvers involving fuel feeding from that
tank. Fuel system component failures need not be considered.
152 Operational Aircraft Performance and Flight Test Practices

That information must be provided in the flight manual.

§25.1585 Operating Procedures

(e) Information must be furnished that indicates that when the fuel quantity
indicator reads “zero” in level flight, any fuel remaining in the fuel tank
cannot be used safely in flight.

(f) Information on the total quantity of usable fuel for each fuel tank must
be furnished.

Any fuel below the zero level must be considered part of the airplane zero fuel
weight and cannot be used for mission planning. Therefore, a careful design
of the fuel tanks system is required to minimize this weight penalty.
As with any critical design value covered by FAA regulations Part 25, the
zero fuel level must be validated through testing. FAA AC 25-7C [2] provides
guidance on an acceptable means of compliance; some of it is provided here:

The unusable fuel quantity is the quantity of fuel that can be drained from
the fuel tank sump with the airplane in its normal level ground attitude
after a fuel tank unusable fuel test has been performed, plus the quantity
remaining in the fuel tank (undrainable fuel).
A fuel tank that is not designed to feed the engines under all flight conditions
need be tested only for the flight regime for which it is designed to do so (e.g.,
cruise conditions). Tanks that are not subject to aeroelastic effects of flight,
such as wing bending or tank flexing, may have their unusable fuel quantity
determined during a ground test.

The fuel system and tank geometry should be analyzed to determine the
critical conditions for the specific tanks being considered (i.e., main and
auxiliary or cruise tanks). The analysis should determine the amount of unu-
sable fuel as a function of airplane pitch and roll attitudes, including those
encountered when executing sideslips and dynamic maneuvers such as
go-around pitch-up and acceleration. The term “most adverse fuel feed con-
dition” is not intended to include radical or extreme conditions not likely to
be encountered in operation.

After the most adverse fuel feed condition and the critical flight attitude have
been determined for the specific fuel tanks being considered, the appropriate
flight tests should be conducted. The flight testing should investigate the
effects of the following:

(a) Steady state sideslips anticipated during operation with the airplane in
both the approach and landing configurations.

(b) For those airplanes capable of high roll and pitch rates, abrupt maneu-
vers should be considered.

(c) A go-around condition at maximum acceleration and maximum rotation

rate to the maximum pitch attitude should be considered.

(d) Effects of turbulence on unusable fuel quantity should be considered.

 CHAPTER 5 Weight and Balance 153

(e) If the airplane includes a low fuel quantity warning system, it should be
demonstrated that the airplane can complete a go-around, approach, and
return to landing, without fuel feed interruption, using the normal
go-around pitch attitude; this should include go-arounds accomplished
with the aid of automated flight guidance systems.

To get a better understanding of what this guidance and regulation means to

an airplane, let’s look at the design of a typical Part 25 jetliner. This type of
plane is designed to spend most of its time in a near-cruise pitch attitude
and has multiple fuel tanks typically feeding collector tanks that then, in
turn, feed the engines; see Fig. 5.18 for an example. Then the zero fuel quan-
tity of all the tanks, except the collector tanks, could be checked by ground
test. The collector tanks would need to be checked in flight. Based on the
layout of the fuel pickup point in the collector tanks, the worst flight con-
dition of the guidance material is selected for test, typically the go-around.
Then, the flight test airplane is modified to allow the crew to isolate the
test collector tank in flight. (Lateral balance must be maintained; one
cannot fly with one side of the plane with zero fuel and the other with a
large quantity, to ensure continued flight.) Once the test collector tank is iso-
lated and the fuel in it is near the zero point, the critical maneuver is executed
until the first sign of engine fuel starvation (fluctuation in engine rotation
speed, or burner temperature or pressure variation not normally associated
with normal operation). The engine being fed by this tank is then shut
down, and the airplane returns to base so the remaining fuel in the collector
tank can be measured; that quantity becomes the unusable fuel.

Passenger Weight
Have you ever been weighed before you got on board a commercial flight?
How does the pilot know how much passenger weight the airplane is carry-
ing? FAA AC 120-27 Rev E defined the average passenger weight as shown in
Table 5.5.
This weight has an allowance of 5 lb for summer clothing and 10 lb for
winter clothing. As well, it contains an allowance for 16 lb of personal
items carried in the cabin; it does not account for checked bags. The
average adult passenger weight is based on a 50% male/50% female load.
The AC goes on to say that the use of standard
average weights is limited to operators of multien- Note: In April 2013, Samoa
gine turbine-powered aircraft originally type- Air started pricing its first
international flights based on
certificated for five or more passenger seats that the weight of its passengers
hold a letter of authorization (LOA), OpSpecs, or and their bags. Depending on
MSpecs, as applicable, and were certificated under the flight, each kilogram (2.2
14 CFR Part 25 or 29, or Part 23 commuter cat- lb) costs $0.93 –$1.06.
egory, or the operator and manufacturer are able
to prove that the aircraft can meet the performance requirements prescribed
by Part 23 commuter category aircraft.
154 Operational Aircraft Performance and Flight Test Practices

Table 5.5 FAA Definitions for Average Passenger Weight

Weight Per
Standard Average Passenger Weight Passenger
Summer Weights
Average adult passenger weight 190 lb
Average adult male passenger weight 200 lb
Average adult female passenger weight 179 lb
Child weight (2 years to less than 13 years of age) 82 lb
Winter Weight
Average adult passenger weight 195 lb
Average adult male passenger weight 205 lb
Average adult female passenger weight 184 lb
Child weight (2 years to less than 13 years of age) 87 lb

For small airplanes (less than five seats), the operator must use actual pas-
senger and baggage weight.

Example 5.4
Consider a 75-ft-long passenger cabin of a 75,000-lb aircraft. A 200-lb passen-
ger decides to move from the first row to the last. What is the impact on the
longitudinal CG?

Solution: The resulting shift in CG of the airplane is simply the moment

change of the airplane (weight of passenger times the distance covered)
divided by the total airplane weight, or
DCG ¼ (200 lb)  ð75 ftÞ=(75,000 lb) ¼ 0:2 ft

If the MAC of the airplane is 10 ft, the resulting aft movement of the CG will
be 2% MAC. Considering that most airplanes have between 20% and 30%
MAC longitudinal CG travel allowance, the movement of several passengers
in flight can have an appreciable impact on the longitudinal CG of the air-
plane, especially if it were loaded near one extremity at the start of the flight.

NTSB report number NTSB/AAR-04/01: On 8 Jan. 2003, an Air Midwest

1900D crashed as it was taking off from Charlotte Douglas International
Airport in North Carolina, killing all 21 people onboard. The use of average
passenger weights (AC 120-27 Rev C in force at the time) that were too low
 CHAPTER 5 Weight and Balance 155

compared to the actual weight carried may have resulted in an overweight

takeoff with a CG more aft than allowed (see Fig. 5.19). This condition was
compounded by previous faulty maintenance that prevented full movement
of the elevators; the pilots were therefore not able to apply sufficient nose-
down elevator to stop the resulting nose-up pitching moment generated by
the aft loading.

18,000
Flight 5481
17,500 actual weight

17,000
Flight 5481
16,500 load mainfest
weight
16,000

15,500
Airplane weight (pounds)

15,000

14,500

14,000

13,500

13,000

12,500

12,000

11,500

11,000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
CG (percent MAC)

Fig. 5.19 Estimated weight and CG of Air Midwest 1900D on takeoff.

Cargo and Checked Baggage

FAA AC 120-27E requires an operator to use actual weight for company
materials, airplane parts, and freight carried aboard an airplane. Freight is
defined as cargo carried for hire in the cargo compartment that is not mail
or passenger bags. An operator should use the weights provided with mani-
fested mail shipments to account for the weight of the mail.
For checked baggage, FAA AC 120-27E allows the use of either:
• Standard, average baggage weights as allowed by regulations
• Standard, average baggage weights based on airline survey
• Actual baggage weight
An operator that chooses to use standard average weights for checked
bags should use a standard average weight of at least 30 lb. An operator
that requests approval to use a standard weight average of less than 30 lb
for checked bags should have current, valid survey data to support a lesser
156 Operational Aircraft Performance and Flight Test Practices

weight. Heavy bags are those that weight more than 50 lb but less than 100 lb.
Bags that are 100 lb or more are considered freight and must be weighed.

Typical Weight and Center of Gravity Envelopes

The weight and center of gravity limits are different for each certified
plane and may look very different from one model to the next, as shown in
Fig. 5.20. For all these planes, their respective defined envelope has been
proven to be safe by flight test and design.
These envelopes and many more are defined in the airplane’s respective
type certificate data sheet (TCDS). The FAA and EASA websites for these
TCDSs are:
https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgMakeModel.
nsf/Frameset?OpenPage
http://www.easa.europa.eu/ws_prod/c/c_tc_aircraft.php

Flight Testing and Flight Test Equipment and Ballast

How heavy is the experimental equipment on the first airplane of a new
product line? For most programs, that airplane is meant to explore the aero-
dynamics characteristics and performance of the new model and expand the
flight envelope for other prototypes that will join the program. Depending on

Lockheed L-1011
TCDS A23WE

Boeing 707-400
TCDS 4A26

22,000
Bombardier Challenger 300 21,000
TCDS T00005NY 20,000
19,000
Weight (lbs)

18,000
Learjet 45
17,000 TCDS T00008WI
16,000
15,000
14,000
13,000
12,000
0 5 10 15 20 25 30
CG (% MAC)

Fig. 5.20 Various weight and CG envelopes.

 CHAPTER 5 Weight and Balance 157

the plane and the exact test it is to cover, that equipment can be significant; as
an example:
The A350-900 “MSN1 carries a flight engineer test station, 8.4t of acqui-
sition racks and some 338km of harnesses, as well as water ballast tanks to
adjust the centre of gravity” (Source: flightglobal.com).
This first airplane is typically used to validate all corners of the weight and
CG envelope. The requirements for transport category airplanes are dictated
by the following:

§25.21 Proof of Compliance

(b) If the airplane flight characteristics or the required flight data are affected
by weight and/or center of gravity (c.g.), the compliance data must be
presented for the most critical weight and c.g. position per § 25.21(a).
Unless the applicant shows that the allowable c.g. travel in one or
more axes (e.g., lateral fuel imbalance) has a negligible effect on compli-
ance with the airworthiness requirements, the applicant must substanti-
ate compliance at the critical c.g.
(d) Parameters critical for the test being conducted, such as weight, loading
(center of gravity and inertia), airspeed, power, and wind, must be main-
tained within acceptable tolerances of the critical values during flight
testing.

From a flight test tolerances point of view, AC 25-7C provides the following
weight (see Table 5.6) and CG:

CG Limits. A test tolerance of þ7 percent of the total c.g. range is intended to

allow for inflight c.g. movement. This tolerance is only acceptable when the
test data scatter is on both sides of the limiting c.g. or when adjusting the data
from the test c.g. to the limit c.g. is acceptable. If compliance with a require-
ment is marginal at a test condition that is inside of the c.g. limits, the test
should be repeated at the c.g. limits.

To reach all these corners, the manufacturer must use a combination of fixed
and moveable ballast installed in the fuselage and sometimes in the wing;
allowing the airplane to be heavy weight forward CG for one flight and poss-
ibly light weight aft CG for the next. The containers for the moveable ballast
(Fig. 5.21), often referred to as water ballast, contain a liquid of known density
that resists well to colder temperatures.
Note that smaller airplanes or airplanes with no large, empty fuselage may
not have the ability to house the equipment to hold the movable ballast and
would rely on fixed ballast to complete the mission; then, the only variable
during the flight would be the fuel burn that affects both weight and CG in
a “fixed” noncontrollable way in flight.
Moving/removing/installing movable ballast can be done relatively easily
with pumps. These pumps can also be used to transfer the liquid from tank to
tank to move the CG in flight. On the other hand, moving fixed ballast into
and out of an airplane is a much more time-consuming activity. For example,
the weights that are moved in the airplane need to be securely attached to the
158 Operational Aircraft Performance and Flight Test Practices

Table 5.6 AC 25-7C Specifications

Weight
Flight Test Conditions Tolerance Limit
+5% +10%
Stall speeds X
Stall characteristics X
All other flight characteristics X
Climb performance X
Takeoff flight paths X
Landing braking distance X
Landing air distance X
Takeoff distance & speed X
Accelerate-stop distance X
Maximum energy RTOs X
Minimum unstick speed X
Minimum control speed X

plane to ensure safe operation. Consider how much time it would take to
move several hundred or even thousands of pounds of fixed ballast into or
out of an airplane.
While executing a test condition in flight test, the crew needs to know
prior to the start of the condition whether the airplane’s weight is within
acceptable tolerances. There are typically two ways to estimate the weight
at the time of the test:
1. Zero fuel weight plus measured fuel volume converted to weight
2. Weight at engine start minus fuel burned during the test
The first method is often used for testing not requiring precise weight
measurement/estimation (where a weight range is sufficient) because the

B777-200LR A380
water ballast system water ballast system

Fig. 5.21 Movable ballast inside test airplanes.

 CHAPTER 5 Weight and Balance 159

Takeoff
Weight (Volume) Weight (fuel flow)
ground
lb lb
Taxi roll Climb Cruise
61.1
61.9
[lb] (10ˆ3)
Weight

60.9
60.8
60.7
60.6
Calibrated airspeed knot Pressure altitude feet
250 3500
Calibrated airspped

Pressure altitude
Lift off
200 3000
150 2500
(kt)

(ft)
100 2000
50 1500
0 1000
Pitch angle deg
15
Pitch angle

10
(deg)

5
0
–5
0 50 100 150 200
Time (s)

Fig. 5.22 Time trace of a takeoff with weight estimate by two different means.

airplane typically has calibrated fuel sensors that convert the measured fuel
volume to a fuel weight, either by using standard fuel density or through a
fuel computer that corrects the fuel volume measured by also accounting
for the fuel temperature (also measured) (see Fig. 5.17). The fuel volume is
computed by a series of sensors that measure the local fuel height in the mul-
tiple fuel tanks on the airplane; those numbers are then summed up into a
volume. Those numbers are calibrated by ground testing at one or multiple
pitch angles. The main disadvantage of this method is that it is subject to fuel
movement in the fuel tanks due to the various accelerations and airplane atti-
tudes (pitch and roll).
The second method computes the fuel weight based on an initial weight
of fuel fed to a computer that is then adjusted down based on measured fuel
flow. The measured fuel flow is typically a fuel volume flow corrected to
weight by using a standard density adjusted for fuel temperature. This
method tends to produce a more stable signal with the weight continuously
decreasing from the time the first engine is started.
Figure 5.22 shows how the computed total airplane weight (zero fuel
weight plus estimated weight of fuel) based on fuel volume could be
impacted during a takeoff ground roll and initial climb compared to the
more stable signal provided by the fuel flow method. The drawbacks of
the fuel volume method are clear here from the point of view of getting
a reliable weight number. But it should also be remembered that the
fuel volume method is a self-contained system in that it does not need
an operator to enter the initial fuel weight; in addition, the fuel volume
method does provide the right order of magnitude of fuel weight under
stable conditions. In the example provided in Fig. 5.22, the maximum
error for the airplane weight due to the fuel movement is of the order
160 Operational Aircraft Performance and Flight Test Practices

of 0.3% compared to the fuel flow method, which for most testing is still a
good enough value.
The fuel flow method is not necessarily the best option all the time. It first
needs an operator to enter the initial fuel weight before it can compute a fuel
quantity based on fuel burn. Then, the fuel flow meters have their own tol-
erance on the computed weight with typical production fuel flow meters
for jet airplanes having accuracies of between less than 0.25% error and
about 2% error.
When a better accuracy is desired for flight test, the measured value of
fuel flow vs time can be corrected by postflight analysis. One method used
to do this is to perform a preflight and a postflight weighing with the airplane
in the exact same configuration both times. Then the delta weight measured
would correspond to the fuel burned. This can be compared to the computed
fuel burned using the fuel flow meter (fuel flow multiplied by time), and the
correction is applied to the fuel flow measured to have the airplane weight
vs time.

Airplane Weighing Procedures and Documentation

To ensure that the proper weight and CG are recorded at the time of
weighing, it is essential that the manufacturer’s weighing procedures are fol-
lowed because the resulting values are then compared to a weight and CG
envelope that has been proven to be safe for operation by design and flight test.
Although each airplane will have its own different weighing procedure,
they will all have the same general information; for example:
• Standard items and items required for operation: All items that are part of
the baseline configuration or that are required for safe operation (fire
extinguishers, first aid kit, etc.) need to be stored in their proper location.
All nonrequired items, such as tools and shop equipment, should be
removed from the airplane prior to performing the inventory check for the
airplane to be weighed.
• Flight controls, gear, doors, and moveable items: Flaps and spoilers
should be in their retracted configuration; landing gear should be
deployed, and the gear doors should be in the position defined by the
manufacturer. Flight control surfaces should be in their “normal” faired
position, and moveable stabilizers should be set to the angle specified in
the weighing procedure. Doors should be closed and access
panels installed.
• Fluid levels: Fluids, other than fuel, should be serviced full or drained
completely if the nonfull quantity is hard to determine. This should be
based on the airplane manufacturer’s directions. A typical list of liquids to
check include:
• Engines oil level
• APU oil level
 CHAPTER 5 Weight and Balance 161

• Landing gear oleo, serviced to proper pressure

• Potable water
• Toilet fluid
• Hydraulic fluids
• Fire extinguisher charge
• Oxygen system
• Nitrogen system
• Engine water injection system (on some older airplanes)
• Anti-ice fluid (TKS)
• Fuel: An airplane being weighed for a base weight requires empty fuel
tanks. The procedure for this also will be provided by the airplane
manufacturer. It should include:
• Normal draining procedure, which requires the airplane to be leveled to
very precise tolerance (within about 0.1 deg)
• Drainage of the sump fuel (see §25.971, at least 0.1% of the tank capacity)

Note: Weighing is not everything if you don’t know the configuration of

the weighed part/airplane. On several occasions, this author received
in-process build weight statements for airplanes’ major assemblies such as
a wing or fuselage but there was no accompanying build statement describing
exactly what was weighed, including application of sealant, tare values, and so
forth. The best way to describe the part weighed is typically to compare it
against a known design drawing (very detailed configuration), setting a
build status using known planning documents, and adding a list of missing
or extra parts (by part numbers).
In addition to this, one must consider the ever-increasing complexity and
preparation time required to weigh an airplane as the size and weight
increase. Figure 5.23 illustrates the problem.
Although certification regulations tend to change over the years, it is
always useful to write down the rule and have a discussion on what is

Fig. 5.23 As the airplane size increases (weight/dimensions), so does the weighing complexity.
162 Operational Aircraft Performance and Flight Test Practices

required. Currently, under 14 CFR Part 25, large transport category airplanes,
the following regulation is provided for weight and balance:
§ 25.1519 Weight, center of gravity, and weight distribution.

The airplane weight, center of gravity, and weight distribution limitations

determined under §§25.23 through 25.27 must be established as operating
limitations.

This paragraph requires that the weight and center of gravity limits be pro-
vided in the AFM. This is typically done by adding a dedicated section in the
AFM on weight and balance that allows the operator of an airplane to deter-
mine the airplane weight and center of gravity through an approved procedure.
§ 25.23 Load distribution limits.

(a) Ranges of weights and centers of gravity within which the airplane may be
safely operated must be established. If a weight and center of gravity
combination is allowable only within certain load distribution limits (such
as spanwise) that could be inadvertently exceeded, these limits and the
corresponding weight and center of gravity combinations must be
established.

(b) The load distribution limits may not exceed—

(1) The selected limits;

(2) The limits at which the structure is proven; or

(3) The limits at which compliance with each applicable flight require-
ment of this subpart is shown.

Section 25.23 specifies that the weight and balance that an operator can use
have been established by design and validated by test. Section 25.25, on the
other hand, specifies the maximum and minimum weights that an operator
can use while operating the airplane.
§ 25.25 Weight limits.
(a) Maximum weights. Maximum weights corresponding to the airplane
operating conditions (such as ramp, ground or water taxi, takeoff, en
route, and landing), environmental conditions (such as altitude and
temperature), and loading conditions (such as zero fuel weight, center
of gravity position and weight distribution) must be established so that
they are not more than—

(1) The highest weight selected by the applicant for the particular
conditions; or

(2) The highest weight at which compliance with each applicable struc-
tural loading and flight requirement is shown, except that for air-
planes equipped with standby power rocket engines the maximum
weight must not be more than the highest weight established in
accordance with appendix E of this part; or
 CHAPTER 5 Weight and Balance 163

(3) The highest weight at which compliance is shown with the certifica-
tion requirements of Part 36 of this chapter.

(b) Minimum weight. The minimum weight (the lowest weight at which
compliance with each applicable requirement of this part is shown)
must be established so that it is not less than—

(1) The lowest weight selected by the applicant;

(2) The design minimum weight (the lowest weight at which compliance
with each structural loading condition of this part is shown); or

(3) The lowest weight at which compliance with each applicable flight
requirement is shown.

Section 25.27 defines the forward and aft, as well as lateral center of gravity
limits (implicit in subparagraph c) within which an airplane can be operated.
These limits could be set by design and validated by flight testing, or they
could be more restrictive at the applicant’s choice.
§ 25.27 Center of gravity limits.
The extreme forward and the extreme aft center of gravity limitations must
be established for each practicably separable operating condition. No such
limit may lie beyond—

(a) The extremes selected by the applicant;

(b) The extremes within which the structure is proven; or

(c) The extremes within which compliance with each applicable flight
requirement is shown.

§ 25.29 Empty weight and corresponding center of gravity.

(a) The empty weight and corresponding center of gravity must be deter-
mined by weighing the airplane with—

(1) Fixed ballast;

(2) Unusable fuel determined under §25.959; and

(3) Full operating fluids, including—

(i) Oil;

(ii) Hydraulic fluid; and

(iii) Other fluids required for normal operation of airplane systems,

except potable water, lavatory precharge water, and fluids
intended for injection in the engine.

(b) The condition of the airplane at the time of determining empty weight
must be one that is well defined and can be easily repeated.
164 Operational Aircraft Performance and Flight Test Practices

Weight a Minute
Every new airplane program starts with a design basic empty weight
(BEW) target from which the airplane performance can be established.
Depending on the company maturity (established vs startup) and the pro-
gram’s complexity (new airframe/engine/avionics/electrical/systems vs
derivative/fuselage extension), some level of margin in the BEW value to
absorb the design experience/requirements changes/creep must be kept.
Too much margin and the marketing value of the plane is reduced (airplane
too heavy, lower published performance). Too little margin may lead to a dif-
ficult program decision when the airplane design evolves beyond the initial
BEW estimate by a large margin.
Every program is faced with design requirement creep where new
capabilities could be included. Adding such “capability” must be resisted
as must as feasible because it often leads to nonoptimal weight changes
(adding weight rather than modifying the original design to account
for the new capability). As well, during a certification program, when sche-
dule takes over, weight can rapidly creep up, almost increasing by the
minute.
Typically, the initial reaction of a new airplane program to the weight
increase is to raise the MTOW of the airplane so as to regain some of the
payload range lost due to the higher empty weight. But more weight is
more drag, and some of the cruise efficiency of the airplane is lost. Consider
the cruise performance triangle (see Fig. 5.24); when any one of the branches
is affected, it typically affects the entire triangle. We will touch more on this
in Chapter 10.
If the weight gets away too much from the target (typically 5% over
baseline), the program must make the hard and costly decision to redesign
parts of the airplane and remove weight. Figure 5.25 shows two airplanes
that were built at approximately the same time to very similar requirements
of payload and range, and both used the newly available Pratt & Whitney
JT9D large bypass turbofans. The Lockheed C-5A design went on to win
the military contract whereas the
Boeing 747-100 became a commercial
success. Weight
Some items that are hard to control
include primer thickness. This should
typically be 0.002 in. thick but can easily
reach 0.010 in. thick if the process is Cruise
not well controlled. Aviation primer performance
paint has a density of approximately
0.007 lb/ft2 /mil (0.001 in.) thickness. It Drag SFC
may not seem like much weight, but con-
sider this example: A Boeing 747-200
wing has a plan form area of 5500 ft2 . Fig. 5.24 Cruise performance triangle.
 CHAPTER 5 Weight and Balance 165

340

330
Empty weight - 1000 LB

Current weight
Guaranteed weight

320

Target weight
310
delivery
1st flight 9th A/C
300
1965 1966 1967 1968 1969 70 71 72
C-5A Manufacturer empty weight

350

345
MFG empty weight - (1000 LB)

340

335

330

325

320
Rollout
0
1966 1967 1968
747-100 Manufacturer empty weight

Fig. 5.25 Evolution of the empty weight of the Lockheed C-5A and Boeing 747-100.
166 Operational Aircraft Performance and Flight Test Practices

This plan form area translates to over 25,000 ft2 of structure covered by
primer (inside and out). A 0.010-in. primer thickness instead of 0.002-in.
primer thickness results in 1400 lb more weight for the same wing due to
the paint process not being followed.
Another item that can be hard to control is sealant. Drawings typically
specify a minimum quantity that must be applied in a given region, but no
maximum. Once the sealant is applied, it can be very hard to measure the
exact quantity applied except by weighing a part before and after application.
The typical density of sealant is 0.06 lb/in.3 .
It cannot be stressed enough that a manufacturer must have a good mass
properties department and that this department needs to be given some level
of program control over a program.

Tracking the BEW In-Service

Like people, airplanes have a tendency to gain weight over time, which
leads to a decrease in performance with passing years. This can be due to:
• Incorporation of service bulletins
• Repairs
• Dirt/trapped water
• Modifications of layout/equipment
• Unaccounted weight gain
It is therefore very important to track the airplane weight vs time. Oper-
ators should establish criteria for recording the weight of the airplane over
time. AC 120-27 recommends that, at a minimum, the following weight
change be recorded:
• Large cabin airplane (more than 70 passengers): 10-lb change
• Medium cabin airplane (30 to 70 passengers): 5-lb change
• Small cabin airplane (less than 30 passengers): 1-lb change
The AC also recommends a maximum time between base weighings of
36 months so as to reestablish a good empty weight value. It also rec-
ommends guidelines for fleet operators when one could be operating essen-
tially the same model airplane in large numbers. This is the case for
most airlines.

Impact of Weight on Performance: Examples

Let’s show some numbers to get a feel for the impact of excess weight on
an airplane’s performance. The first example will use the Boeing 747-100 and
the chart presented in Fig. 5.24 for the program. The initial empty weight at
the start of 1966 was projected to be 325,000 lb, which rapidly grew to
 CHAPTER 5 Weight and Balance 167

343,000 lb by mid-year. The impact on fuel burn of this extra weight can be
estimated as follows:
• Assume you are comparing the 325,000-lb empty-weight airplane to the
343,000-lb empty-weight airplane. They therefore have similar
aerodynamic characteristics and engines.
• Assume both airplanes carry 250 passengers at an average weight of 200 lb
per passenger with luggage, a total of 50,000 lb.
• Assume both airplanes take off with the same crew (2000 lb) and same
quantity of fuel (say 250,000 lb). By the time the airplane is in cruise, the
fuel left is around 200,000 lb. So the lighter weight airplane would weigh
approximately 577,000 lb, whereas the heavier version would weigh
595,000 lb.
1. As you will see in Chapter 10, the drag in cruise is equal to the airplane
weight divided by the lift-to-drag ratio in cruise.
W
D ¼
L=D
2. The fuel burn of the airplane for a given flight condition is proportional
to the drag times the engine efficiency as expressed by the specific fuel
consumption (SFC) in terms of pounds of fuel per hour per pound of
thrust produced.
Ẇ f ¼ SFC D

3. Therefore, the ratio of the fuel burn of the heavier airplane to the lighter
weight airplane can be approximated as

 W 
 SFC 
Ẇ f  L=D heavy
heavy
 ¼ 
 W 
Ẇ f  SFC
light L=Dlight

4. We have already mentioned that the airplanes are aerodynamically

similar, so we can cancel out the (L/D) under similar flight condition.
(You will see in Chapter 10 how the weight impacts the L/D.). As well,
under similar flight conditions, the engines’ SFC for both planes will be
similar and therefore cancel out. We are left with the approximate
impact of the heavier airplane on the fuel flow


Ẇ f  Wheavy 595,000 lb
heavy ¼ ¼ ¼ 1:031
 W jlight 577,000 lb
Ẇ 
f
light

5. The heavier airplane would have burned about 3% more fuel for a given
payload than the lighter weight airplane, which is directly proportional
to the cruise weight ratio of the two planes.
168 Operational Aircraft Performance and Flight Test Practices

After the weight reduction program, the airplane ended up weighing

approximately 4000 lb less than at the start of 1966, or 1.2% less with the cor-
responding fuel savings.
For the second example, consider two different projects, noting that all
numbers provided here are those of the author and do not necessarily
reflect the actual numbers from the manufacturers. (At the time of
writing this section, no numbers were available on the Internet from the
manufacturers; the numbers provided here are for use as an example
only. This section will be updated at a later date when data become avail-
able. As well, the Bombardier CS300 became part of the Airbus airliner
suite in October 2017 and officially was rebranded as the A220-300 in
2018). The Bombardier CS300 jetliner, an all-new design, competed
against the A319neo, a reengined version of the A319 with sharklets.
(Both airplanes are shown in Fig. 5.26.) The A319 is a derivative of the
A320, and as such, is not weight optimized for its size. The CS300 is
weight-optimized for its size and uses advanced materials and aerody-
namics. Both airplanes fly using fly-by-wire, and both use the geared turbo-
fan (GTF) from Pratt & Whitney. The CS300 operating weight empty is
approximately 80,000 lb, whereas the A319neo with the heavier (but
more efficient) GTF and heavier sharklets (improving L/D) has an approxi-
mate OWE of 95,000 lb.
If we use the same approach as presented in the first example, further
assuming that the L/D ratios are similar in cruise (not accounting for the
20 yr of aerodynamic evolution from the initial conception of the A319 to
the CS300) and that both airplanes are carrying 100 passengers (about
20,000 lb) and the same fuel quantity (use 25,000 lb), we get a ratio of fuel
burn in flight of the A319neo to the CS300 of


Ẇ f  W 140,000 lb
heavy ¼ heavy ¼ ¼ 1:12
 W jlight 125,000 lb
Ẇ f 
light

Fig. 5.26 Bombardier CS300 and Airbus A319neo.

 CHAPTER 5 Weight and Balance 169

where the weight-optimized CS300 burns 12% less fuel than the heavier
A319neo for the same payload and fuel quantity. The CS300 could therefore
cruise 12% more distance for the same quantity of fuel or carry approximately
12% (3000 lb) less fuel for the same range, the lower weight further increasing
its efficiency.

Exercises
1. You are preparing to weigh an airplane with fuel in preparation for a test
flight. You have three floor scales, each with a 10,000-lb capacity. Based on
your computation, you expect the airplane to weigh approximately 22,000
lb and have a longitudinal CG of 20% MAC. Can you use these scales for
the job? You know the following about the airplane geometry:
• LEMAC ¼ 414 in.
• MAC ¼ 87 in.
• Nose gear FS ¼ 165 in.
• Main landing gears FS ¼ 455 in.
2. For the weight and CG specified in question 1, what is the expected load
on each scale?

References
[1] Federal Aviation Administration, “Aircraft Weight and Balance Control,” FAA AC
120-27E, 10 June 2005.
[2] Federal Aviation Administration, “Flight Test Guide for Certification of Transport
Category Airplanes,” FAA AC 25-7C, 16 Oct. 2012.
[3] Transports Canada, “Flight Characteristics with Lateral Centre of Gravity,” TCCA AC
525-013, 1 Dec. 2004.
[4] Federal Aviation Administration, “Aircraft Weight and Balance Handbook,”
FAA-H-8083-1B, 2016.
[5] National Transportation Safety Board, “Aircraft Accident Brief,” 14 April 2004,
https://www.ntsb.gov/investigations/AccidentReports/Reports/AAB0401.pdf
[retrieved 20 Oct. 2020].
Chapter 6 Lift and Stall

Chapter Objective
One does not think much about the lift force until the airplane nears its limit,
but lift is central to the performance of the airplane. In the middle of the envel-
ope, lift is a “nonevent” (sufficient margin), but the resulting angle of attack at
which the lift is produced may impact the flight. This chapter provides a first
glimpse at lift generation, the impact of the angle of attack on the pilot field of
view from the cockpit at operational speeds, and its impact on cabin floor
angle in cruise. We talk about limiting lift conditions and how manufacturing
tolerance may impact baseline aerodynamics. Finally, we look at the impact of
wing and tail contamination on maximum lift.

Basic Airplane Geometry

B
efore we address the subject of lift generation, we will briefly review
the parameters of the airplane that have an impact on the aerody-
namics. This is a review in the context of airplane performance to
give the tools necessary to discuss the subject. Most of the lift is generated
by the wing, so we will focus the discussion on this part of the airplane.

Airfoil Shapes
A wing is made up of a series of airfoils (a two-dimensional shape, a cut of
the wing typically along the incoming flow) that are assembled to achieve a
given aerodynamic goal. An airfoil is described by a series of parameters
(see Fig. 6.1). Airfoils are typically defined by:
• Chord c, the length of the airfoil as measured from the leading edge to the
trailing edge.
• Thickness t, the dimension perpendicular to the chord. This is typically
represented as a relative value compared to the chord. The airfoil
maximum thickness to chord (t/c) is a key factor that defines low-speed vs
high-speed airfoils. The thickness distribution along the chord greatly
impacts its aerodynamic characteristics.

171
172 Operational Aircraft Performance and Flight Test Practices

Camber line
V Thickness
Leading Trailing
Airflow edge edge
V

Chord
line Chord length

Fig. 6.1 Airfoil geometry.

• Camber; the camber line falls midway between the lower surface and upper
surface of the airfoil at a given point along the chord. The maximum
camber (as a percentage of the chord) impacts the airfoil’s ability to provide
lift at a given angle relative to the flow.

Wing Plan Form

The wingspan of the airplane b is measured from one wing tip to the
other, as measured perpendicular to the incoming flow (Fig. 6.2). The root
chord cr of the wing lies in the plane of symmetry and is parallel to the incom-
ing flow. The tip chord ct is the length of the wing tip. The ratio of the tip
chord to the root chord is the taper ratio l.
ct
l¼ (6:1)
cr
The wing area S, for airplane performance analysis, is a reference dimen-
sion. This dimension represents a wing that is aerodynamically equivalent to
the airplane’s wing. Figure 6.3 shows various definitions of a wing area. The
exposed wing area is the part of the wing outside the fuselage that touches the

LE

yMAC cr
cMAC

ct

Fig. 6.2 General wing geometry.

 CHAPTER 6 Lift and Stall 173

Exposed Gross Aerodynamically

wing area wing area equivalent
wing area

Fig. 6.3 Wing area.

air. The gross wing area includes part of the wing inside the fuselage. The
aerodynamically equivalent wing represents the gross wing area with straight
leading and trailing edges.
The wind aspect ratio (AR) represents the finesse of the wing when
viewed from the top. Mathematically, the aspect ratio is defined as the
ratio of the wingspan squared to the wing area

b2
AR ¼ (6:2)
S
The wing sweep L represents an angle between the span line (perpen-
dicular to the flow) and a reference line on the wing. The typical reference
lines on the wing include the leading edge LLE , the trailing edge LTE , and
the quarter chord of the wing Lc=4 .
The mean geometry chord (MGC) of the airplane is the wing area divided
by the wingspan. The mean aerodynamic chord (MAC) is the theoretical
chord for a rectangular wing with the same aerodynamic forces as the
actual wing. It is computed by integrating the local chord of the wing
along the wingspan [Eq. (6.4)]. MAC and MCG are often the same value
or very close; MGC can be used as a first estimate.
S
MGC ¼ (6:3)
b
b=2
ð
2
MAC ¼ ½cð yÞ2 dy (6:4)
S
0
174 Operational Aircraft Performance and Flight Test Practices

Root Root
Tip
Tip
tip


Geometric twist Aerodynamic twist

Fig. 6.4 Wing twist.

A wing will incorporate some level of wing twist to adjust the lift distri-
bution along the span. The twist can be geometric (twist angle e ), where the
local chord line changes along the span; or it can be aerodynamic, where the
chord line is not changed but the aerodynamic shape of the airfoil changes
along the span; or it can be a combination of both. When one refers to the
wing twist, it is often a discussion of the geometric angle between the wing
tip and wing root (Fig. 6.4), rather than an aerodynamic twist discussion.
The wing’s dihedral is the mean line of the wing when seen from the front
as compared to a horizontal line. A positive dihedral is one where the wing tip
is above the wing root (Fig. 6.5). A negative dihedral (or an anhedral wing) is
one where the wing tip is lower than the wing root. The wing’s dihedral influ-
ences the airplane’s stability.
The wing’s incidence iw is the angle of the chord line of the wing root with
respect to the airplane’s reference line. This reference line is typically the
longitudinal axis; for commercial planes, this line is parallel to the fuselage
cabin floor (see Fig. 6.6).
The angle of attack a is the angle between the incoming airflow and the
reference line on the airplane. For the entire airplane, the reference line is
again the longitudinal axis. The local geometric angle of attack along the
wingspan will be a function of the airplane’s angle of attack, the wing inci-
dence, and the wing geometric twist (see Fig. 6.7).
Wings can be equipped with wing tip devices to improves their aerody-
namics from a lift and drag point of view. The most popular of these
devices are winglets (see Fig. 6.8). These devices may alter the geometric
wingspan in a different way than they alter the aerodynamic wingspan.

Dihedral Anhedral

Fig. 6.5 Dihedral.

 CHAPTER 6 Lift and Stall 175

Wing root chord

iw
Reference line

Longitudinal axis

Fig. 6.6 Wing incidence.

Wing
ro ot cho
rd
Reference
line iw

root

Fig. 6.7 Angle of attack.

Geometric wingspan

Aerodynamic wingspan

Fig. 6.8 Wing tipdevices and wingspan.

176 Operational Aircraft Performance and Flight Test Practices

Lift and Lift Equation

An object immersed in a moving air mass will be subjected to an aerody-
namic force F. That force is broken down into a component that is perpen-
dicular to the incoming airflow and is designated as lift L plus a component
that is parallel to the incoming airflow, the drag D (see Fig. 6.9).
The lift force is quantified in terms of dynamic pressure, (q ¼ 12rSL s V 2 ),
wing area S, and a lift coefficient CL [see Eq. (6.5)]. The lift coefficient pro-
vides a means to adjust the results to best match the actual behavior. That
coefficient allows for the same equation to be used throughout the flight
envelope and includes effects from Mach number change, variation in Rey-
nolds number, angle of attack, geometry and configuration, and so forth.

1
L ¼ rSL s V 2 S CL (6:5)
2
The airplane’s load factor n, the g loading, is defined as the ratio of the lift
produced to the airplane weight. In level flight, the lift generated by the air-
plane must equal the weight to maintain level flight; thus, the load factor is
equal to 1.0.

L
n¼ (6:6)
W
From the definition of the lift in Eq. (6.5), the lift coefficient of the air-
plane, accounting for load factor, is then

nW
CL ¼ (6:7)
1
r s V2 S
2 SL
In a perfect, steady level flight, the load factor will be 1.0; however, actual
flights are rarely perfectly steady (see Fig. 6.10). Notice, on the right side of
the flight traces, the variation in load factor for an essentially constant alti-
tude and airspeed condition. This small variation is part of expected flight
test data scatter, and some engineering judgement is required when data

L
D

V

Fig. 6.9 Aerodynamic forces on an airplane in flight.

 CHAPTER 6 Lift and Stall 177

Altitude (fT)
41,100

41,050

41,000

40,950

40,900
Indicated airspeed
240

238

236

234

232

230
 Load factor from 1g
0.04

0.02

0.00

–0.02

–0.04
Lift coefficient
0.60

0.55

0.50

0.45

0.40

Fig. 6.10 Example of flight test computed lift coefficient.

are collected. Consider that a variation of 0.01 in load factor (small) actually
corresponds to a force equivalent to 1% of the airplane’s weight.

Lift Curve and Lift Distribution

An airfoil and a wing will have expected lift coefficient behavior with
angle of attack. We first review the airfoil behavior. Figure 6.11 shows the
expected change in lift coefficient of a given airfoil. First there is the angle
 178
Operational Aircraft Performance and Flight Test Practices
a) b) c)

Cl
Clmax

dcl
d

0 max 

Fig. 6.11 Airfoil lift curve characteristics and NASA data for NACA 0012 and 23012. Source: [4].
 CHAPTER 6 Lift and Stall 179

of attack for zero lift coefficient a0 . From this point, the lift coefficient
increases almost linearly with a given lift curve slope (dcl /da), and then
tapers and drops off beyond a maximum value (clmax ). We will discuss the
loss of lift near and beyond that point later in the chapter.
A symmetrical airfoil like the NACA 0012 of Fig. 6.11b (symmetry with
respect to the chord line) will have a a0 equal to zero, whereas one with a
positive camber (camber line above the chord line, like the NACA 23012
in Fig 6.11c) will have a negative a0 and will have a lift coefficient greater
than zero at a zero angle of attack. An airfoil (two-dimensional) will have a
lift curve slope of 2p per radian (or about 0.1 cl per degree of angle of
attack) in the linear part of the curve.
A wing is built from airfoils that the manufacturer distributes as required
to achieve the proper aerodynamic for the intended mission of the airplane.
Starting with a wing loading [W/S, Eq. (6.7)], the airfoils selected will typi-
cally vary in shape and geometric angle of attack (wing twist). To visualize
the impact, we offer an example in Fig. 6.12. Here, the wing loading selected
was W/S ¼ 45.6 psf. The airspeed was 140 KCAS, resulting in a required lift
coefficient of 0.687 under steady level flight conditions (load factor, n ¼ 1.0).
The wing has zero sweep at the quarter chord location. The wing has an
aspect ratio (AR) of 11.5. Three different plan forms are presented, defined
by taper ratios of 1.0 (reference wing), 0.5, and 0.2. For the last two taper
ratios, two wing twist conditions are presented, 0 deg (same geometric
angle of attack along the span) and –3 deg (geometric angle of attack of
wing tip is 3 deg lower than at the wing root). For all wing cases, a single
airfoil is used to simplify the discussion. For each wing, the lift coefficient
at a given span location is presented and compared to the wing lift coefficient
of 0.687 (dotted line). We observe:

• For a wing of finite span, the lift coefficient at the wing tip is zero. (The
pressure difference between the bottom and top of the wing must be equal
at the tip.)
• Reducing the taper ratio from 1.0 to 0.2 moves the maximum local lift
coefficient from inboard to outboard.
• Reducing the geometric angle of attack from inboard to outboard moves
the peak lift coefficient back to the inboard of the wing.
• The wing angle of attack for a given wing lift coefficient will be impacted by
the taper and twist.

For lift to be created by a wing of finite span, there is a necessary transfer

of momentum to the air; this transfer of momentum (proportional to a mass
times the change of airspeed perpendicular to the incoming air; that is, the
induced speed dVv ) is distributed along the span of the wing (see
Fig. 6.13). For a wing of infinite span, this induced speed tends towards
zero (i.e., dVv  0). This induced speed is the wing’s downwash v.
 180
Operational Aircraft Performance and Flight Test Practices
Half span wing shape Half span wing shape

Chord length (m)

Chord length (m)

12 12
10 10
8 8
6 6
4 4
Half span wing shape 2
0
2
0
12
Chord length (m)

10 0.80 0.80

8 0.70 0.70
6 0.60 0.60
4
0.50 0.50
2 Twist = 0 deg
0 Twist = 0 deg

CL
0.40 0.40

CL
0.30
Taper = 0.5 0.30 Taper = 0.2
0.20  = 2.37 deg 0.20  = 2.37 deg
0.90
0.10 0.10
0.80 0.00 0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
0.70 Span location (m) Span location (m)

0.60 Half span wing shape Half span wing shape

12

Chord length (m)

12
W/S = 45.6 lb/ft2

Chord length (m)

0.50 10 10
CL

8 8
0.40 Twist = 0 deg 6
4
6
4
2
0.30
Taper = 1.0 2
0 0

c/4 = 0 deg 0.80 0.80

0.20
AR = 11.5 0.70 0.70
0.10  = 2.57 deg 0.60 0.60
0.50 0.50
0.00
Twist = –3 deg Twist = –3 deg

CL
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00

CL
0.40 0.40

Span location (m) 0.30 Taper = 0.5 0.30 Taper = 0.2

0.20  = 3.67 deg 0.20  = 3.60 deg
0.10 0.10

0.00 0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Span location (m) Span location (m)

Fig. 6.12 Example of spanwise lift distribution vs wing plan form, twist and taper.
 b

L L F
L

V
dVv  i Di
V

Front view

Half span wing shape Half span wing shape

12
Chord length (m)

12

Chord length (m)

Chord length (m)

12
10 10 10
8 8 8
6 6 6
4 4 4
2 2 2
0 0 0

0.80 0.80 0.80

0.70 0.70 0.70

0.60 0.60 0.60

0.50 0.50 0.50

2
0.40 ARW/S = 45.6 lb/ft
= 11.5 0.40 AR = 7.0 0.40 AR = 5.0
CL

CL

CL
Twist = –3 deg Twist = –3 deg Twist = –3 deg

CHAPTER 6
0.30 0.30 0.30
0.20 Taper = 0.5 0.20 Taper = 0.5 0.20 Taper = 0.5
0.10
 = 3.67 deg 0.10
 = 4.61 deg 0.10
 = 5.63 deg
0.00 0.00 0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Span location (m) Span location (m) Span location (m)

Lift and Stall

Fig. 6.13 Aspect ratio impact on lift and AoA, same conditions as Fig. 6.12.

181
182 Operational Aircraft Performance and Flight Test Practices

If we define a stream tube of air proportional to the wing span b, we note

that for a given lift force L generated by this wing of finite span, the down-
wash v produced will need to increase as the span is reduced; this increase
will be proportional to the span squared (i.e., the strength of the downwash
must increase as fast as the reduction in the span squared; see Fig. 6.13). The
average induced angle ai is half the downwash angle (e  v/V for small
angles). This induced angle tilts the aerodynamic force vector, creating two
components (when broken down into the wind axis). This tilting of the
force vector reduces the effective angle of attack of the airflow approaching
the wing, which means the wing will need a greater geometric angle of
attack to produce the same lift force. Figure 6.13 shows three examples of
aspect ratio and the required angle of attack to create the same lift coefficient
of 0.687 (same example as that of Fig. 6.12).
The impact of the aspect ratio on the wing lift curve can be modeled with
the following equation:

dCL a1
¼ a1  (6:8)
da 1þ
pAR

where a1 is the lift slope of the airfoil (two-dimensional). The impact of the
aspect ratio on the lift curve and lift coefficient at a given angle of attack is
illustrated in Fig. 6.14. It can be seen that the expected angle of attack for
a required flight lift coefficient is greatly impacted by the lift curve slope.
This will be a critical parameter to use for defining the airplane pitch attitude
in flight, pitch attitude being equal to angle of attack plus flight path angle
[see Eq. (12.1) in Chapter 12].

7 1.4
a
6 1.2 A
7
1.0
dCL/d (CL per rad)

5 6 5
4 3
Lift coefficient, CL

0.8 2
4 1
0.6
3
0.4
2
0.2
1
0
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 –0.2

AR –0.4
–10° 0° 10° 20°
Angle of attack,  (degrees)

Fig. 6.14 (a) Impact of aspect ratio on the lift curve slope; (b) example on lift curve. Source: [4].
 CHAPTER 6 Lift and Stall 183

Trailing edge devices Cl Leading edge devices

Plain flap Droop nose

Slotted flap Slat

Split flap Flaps up

Fowler flap

Fig. 6.15 High-lift devices.

Configuration Effects
The previous discussion applies for a given configuration of the plane
(flaps up, gear up as a baseline). Change of configuration in flight will, of
course, introduce changes in the lift curve and maximum lift coefficient.
The deployment of high-lift devices (Fig. 6.15) will change the a0 , amax ,
and CLmax of the basic flaps-up condition.
Trailing edge devices tend to increase the lift coefficient at a given angle
of attack (reduce a0 ) and increase CLmax , but also reduce amax (stall earlier).
Leading edge devices tend to reduce the lift coefficient slightly at a given
angle of attack (increase a0 ) as well as increase both CLmax and amax (stall
later). Deployment of these devices will therefore alter the airplane pitch atti-
tude at a given airspeed for a given wing loading.
Other configuration changes like spoilers (devices on the wing that kill
lift), landing gear, and even engine thrust/propeller propwash will also
affect the lift produced by the wing and thus impact the angle of attack
(and pitch angle) for a given required lift coefficient.

Required Lift Coefficient vs Airspeed

The required lift coefficient for level flight is inversely proportional to the
square of the airspeed [Eq. (6.7)] with a load factor of 1.0. At the higher
speeds, the rate of change of lift coefficient with airspeed is small, and the air-
craft attitude in level flight (proportional to the angle of attack) will also vary
slowly (around 1 deg per 50 kt). Figure 6.16 provides an example of wing
loading impact on a given wing (defines lift curve) as a way to quantify the
impact. For a wing loading of 75 lb/ft2 , going from an airspeed of 300
KEAS to 350 KEAS will result in a lift coefficient change from 0.246 to
0.181, which translates to a change of angle of attack from 1.11 deg to
0.55 deg.
At the lower airspeeds, the rate of change of lift coefficient with airspeed
is much higher, and the aircraft attitude in level flight (proportional to the
184 Operational Aircraft Performance and Flight Test Practices

2.5 1.8
1.6
1.4
2 1.2
CLmax 1
1.5 0.8
CL

CL
0.6
0.4
200 W/S (psf )
1 150 0.2
75 100 0
50
25 –0.2
0.5 –0.4
–4 –2 0 2 4 6 8 10 12 14 16 18 20
0  (deg)
0 50 100 150 200 250 300 350 400
KEAS

Fig. 6.16 Required lift coefficient vs wing loading and airspeed.

angle of attack) will also vary much more rapidly (as high as 1 deg per 1 kt
near stall). Again, with the example of Fig. 6.16 and a wing loading of
75 lb/ft2 , going from an airspeed of 125 KEAS to 118 KEAS will result in a
lift coefficient change from 1.418 to 1.591, which translates to a change of
angle of attack from 11.8 deg to 15 deg.
One must also consider the angle of attack change at fixed airspeed and
wing loading (defined required lift coefficient) when the airplane configur-
ation is changed. Figure 6.17 shows an example of an airplane going from
a cruise condition (flaps up) lift coefficient, slowing down to 1.3 VSR (see
stall speed discussion in Chapter 7) for that configuration, selecting approach

Landing
flaps
3.5 Takeoff/
approach
3 flaps
2.5

1.23 Vsr 2
1.3 Vsr
1.5
c_l

0.5

Flaps up 0

–0.5

–1
Cruise
–14–12–10–8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20 22 24
–1.5
AoA

Fig. 6.17 Angle of attack change during flap transition.

 CHAPTER 6 Lift and Stall 185

Fig. 6.18 Nose low on approach.

flap, and then slowing down to 1.3 VSR of this new configuration for that flap
setting prior to selecting landing flaps and slowing down to landing reference
speed (VREF , 1.23 VSR ). Note how the angle of attack changes during this
maneuver. For this example, going from the approach flap to the landing
flap produces a major change in angle of attack (pronounced nose-down con-
dition), which must be evaluated by the test team to determine the accept-
ability from a pilot operating point of view. Too much lift capability on
approach may lead to a more nose-down attitude for landing with the nose
landing gear (NLG) becoming lower than the main landing gear (MLG)
(see Fig. 6.18), which increases the risk of NLG touching prior to the MLG.
Furthermore, one should review expected flight attitude for all expected
flight conditions (see Fig. 6.19). Cabin floors where passengers are located are
often parallel to the airplane longitudinal axis. One should verify that the atti-
tude is fairly flat in cruise condition where, for example, the flight attendants
need to push a cart in the aisle.

12

10

8
Expected low speed
Deck angle (deg)

6
W/S = 225 kg/m2
4
Expected normal High speed
cruise speed cruise
2
W/S = 178 kg/m
2

–2
100 110 120 130 140 150 160 170 180 190 200
Airspeed (KCAS)

Fig. 6.19 Deck angle vs airspeed for two wing loadings.

186 Operational Aircraft Performance and Flight Test Practices

Lift vs Weight Coefficients

A distinction must be made between the lift coef- Note: For the study of
ficient, shown in Eq. (6.7), and often what is per- aircraft performance, only
ceived to be a lift coefficient but is actually a weight the net lift is required to
sustain flight; the lift
coefficient CW , shown in Eq. (6.9). The lift coefficient
generated by the wing or
accounts for the load factor normal to the flight path. download from the tail will
The weight coefficient does not account for the not be considered. Stall
load factor. The two coefficients are equal only testing and drag testing are
usually done at the most
when the load factor is equal to 1.0. The only time
adverse CG for the aircraft.
we will omit including the load factor in the lift coef- For a conventional layout,
ficient equation in the book will be during the discus- the forward CG is more
sion on level flight (1 g). penalizing, requiring the
largest generation of lift from
the wing for a given
W aircraft weight.
CW ¼ (6:9)
1
r s V2 S
2 SL

Lift Limit: Aerodynamic Stall

As the saying goes, “All good things must come to an end.” An airfoil’s lift
coefficient will increase with increasing angle of attack up to a point (see
Fig. 6.20). Then the airflow will start separating from the top surface of the
airfoil, and the rate of change on lift coefficient with angle of attack (the
lift curve slope, dcl /da) will reduce. At some angle of attack, the flow will sep-
arate enough that any increase in angle of attack will result in a reduction in
the lift coefficient. The angle of attack at which the lift coefficient is
maximum is the natural wing stall; it is the upper limit of lift capability for
the wing.
Not all of the natural lift may be available to use from an airplane per-
formance point of view. Depending on the flow separation pattern and

Aerodynamic cImax
CI

Fig. 6.20 Aerodynamic lift limit. Source: Flow visualization from NASA video, https://www.
youtube.com/watch?v¼3_WgkVQWtno
 CHAPTER 6 Lift and Stall 187

Fig. 6.21 Expected separation zones based on wing plan form, untwisted wing.

suddenness of the stall or the stall symmetry, the actual maximum may need
to be trimmed back. We discuss this subject extensively in Chapter 7.
The flow separation pattern is highly dependent on the design of the wing
with some expectations for a given wing plan form (see Fig. 6.21). Flow sep-
aration closer to the wing root (such as for a rectangular wing) will have a
greater impact on the generation of lift (lower aerodynamic CLmax ; see the
spanwise lift distribution of Fig. 6.12), but typically will provide good stall
handling characteristics. Flow separation near the wing tip will have less
impact on the aerodynamic CLmax , but may result in unacceptable stall
characteristics, which may force the manufacturer to limit the usable angle
of attack of the airplane in service. Lift distribution near stall may be
tweaked with plan form selection, aerodynamic and geometric twist along
the span, and airfoil selection.
Flap deflection will change the dynamic of the flow pattern over the wing,
which introduces another variable for the manufacturer to deal with to again
get an acceptable flow separation pattern (see Fig. 6.22) and acceptable stall
characteristics. The stall characteristics of the airplane should not be depen-
dent on the airplane configuration beyond what is allowed by regulations.
This needs to be validated by flight test (see Chapter 7).

Stall on Propeller-Driven Planes

Airplanes with propellers face another challenge during stall—the aero-
dynamics of the airplane can be significantly altered by the propeller flow
and most likely result in asymmetrical lift distribution over the wing that
then impacts the maximum usable (certifiable) lift coefficient. This is
especially true for single-engine airplanes (see Fig. 6.23). On multiengine
188 Operational Aircraft Performance and Flight Test Practices

a) b)

Mach number = 0.186

Reynolds number = 1.03 x 106

Fig. 6.22 Stall pattern on a straight tapered wing with (a) flaps deployed and (b) flaps retracted.
Source: [5].

airplanes, some symmetry can be restored from a left wing to right wing
point of view if the manufacturer is willing to have propellers turning differ-
ently on each wing (i.e., if the left wing propeller turns clockwise, then the
right wing propeller turns counterclockwise). This is a design decision, but
also a cost decision (two different propellers increase product cost at some
level).
Propeller power also changes the lift produced by increasing the local
dynamic pressure behind the propeller disk (see the discussion in Chapter
8) and thus increasing the lift produced (but also changing the flow direction
behind the disk, which may change the local angle of attack). See the Note
box in Chapter 7 in the “Thrust Correction” section of the chapter for the
impact of propeller thrust on stall speed.

V
Propeller blade
tip vortex

Fig. 6.23 Propeller flow impact on airplane aerodynamics.

 CHAPTER 6 Lift and Stall 189

High-Altitude Stall
As an airplane gains altitude, the Mach number will increase for a given
calibrated airspeed (see Fig. 3.15 in Chapter 3). This increase in Mach will
have an impact on the lift produced by the wing at a given angle of attack
(as shown in Figs. 6.24 and 14.28 later in the book). One of the ways it will
impact lift is an initial small lift curve slope increase with increasing Mach
number. A detrimental effect is the reduction in the maximum lift coefficient
with increasing Mach, which can be significant (see Fig. 6.24).
The impact on high-altitude impact was clearly shown in the French
Bureau of Enquiry and Analysis for Civil Aviation Safety (BEA) report on
Air France’s Airbus A330-200 Flight AF447 accident. Figure 6.25 is an
extract from the report showing the general impact of Mach (Fig. 6.25b)
and actual flight trace at the initial stall warning activation; initial stall
warning occurred at less than a 5-deg angle of attack. See the following acci-
dent box on the conclusion of the AF447 accident investigation.

1.6

M=0.20
1.4 .30
.40
.45
1.2
.50
.55
1.0
.60
.65
.8
.675

.6
CL

.70
.4

.2

–.2

–.4

–.6

–12 –8 –4 0 4 8 12 16 20 24 28
, deg.

Fig. 6.24 Example of Mach number effects on the lift curve. Source: [4].
 190
Operational Aircraft Performance and Flight Test Practices
Cz 0.8

0.6
Continuous

Initial stall warnings

Mach number computed
0.4
stall warnings
M = 0,3 0.2
Mach number recorded to crash
0

M = 0,8 12
AOA value emitted by IRS3 (>0= UP) (DA)

9 AOA value emitted by IRS2 (>0= UP) (DA)

AOA value emitted by IRS1 (>0= UP) (DA)
Stall warning thoretical threshold
6

3
alphamax M0,8 < alphamax M0,3 Angle of attack

0
02:10:00 02:10:10 02:10:20 02:10:30 02:10:40 02:10:50 02:11:00

UTC (hh:mm:ss)

Fig. 6.25 Impact of high altitude/Mach on maximum lift capability. Source: BEA Final Report, Airbus A330 –203 Flight AF447 accident, 1 June 2009.
 CHAPTER 6 Lift and Stall 191

BEA Final Report, Air France AF447 Accident, 1 June 2009: On 31 May
2009, the Airbus A330 flight AF447 took off from Rio de Janeiro Galeão
airport bound for Paris Charles de Gaulle. ... At 2 h 10 min 05, likely following
the obstruction of the Pitot probes by ice crystals, the speed indications were
incorrect and some automatic systems disconnected. The aeroplane’s flight
path was not controlled by the two copilots. They were rejoined 1 minute
30 later by the Captain, while the aeroplane was in a stall situation that
lasted until the impact with the sea at 2 h 14 min 28.
The accident resulted from the following succession of events:
• Temporary inconsistency between the measured airspeeds, likely following
the obstruction of the Pitot probes by ice crystals that led in particular to
autopilot disconnection and a reconfiguration to alternate law,
• Inappropriate control inputs that destabilized the flight path,
• The crew not making the connection between the loss of indicated
airspeeds and the appropriate procedure,
• The PNF’s late identification of the deviation in the flight path and
insufficient correction by the PF,
• The crew not identifying the approach to stall, the lack of an immediate
reaction on its part and exit from the flight envelope,
• The crew’s failure to diagnose the stall situation and, consequently, the lack
of any actions that would have made recovery possible.
Some of the recommendations to come out of the investigation include:
• EASA reviews the content of check and training programs and makes man-
datory, in particular, the setting up of specific and regular exercises dedi-
cated to manual aircraft handling of approach to stall and stall recovery,
including at high altitude.
• EASA and the FAA evaluate the relevance of requiring the presence of an
angle of attack indicator directly accessible to pilots on board airplanes.
• EASA determines the conditions in which, on approach to stall, the
presence of a dedicated visual indication, combined with an aural warning,
should be made mandatory.

One of the responses by the crew, which was part of all training,
was to advance the throttles to increase engine thrust when an indication
of stall (stall warning) is provided, and the crew did just that. They sub-
sequently did not recognize that the airplane was stalled because
engine power was at maximum and the airplane pitch attitude was
above the horizon, yet the airplane was descending at rates as high at
17,000 ft/min.
An action that came out of this investigation was the creation of FAA AC
120-109, Stall Prevention and Recovery Training, which was first released in
August 2012. The AC provides guidance for training, testing, and checking
192 Operational Aircraft Performance and Flight Test Practices

pilots to ensure correct responses to impending and full stalls. The core of the
AC includes the following:

1. Stall and Stick Pusher Training

1.1. AC developed to provide best practices and guidance for training,
testing, and checking for pilots, within existing regulations, to ensure
correct and consistent responses to unexpected stall warnings and
stick pusher activations.
2. Core principles of this AC include:
2.1. Reduction of AOA is the most important response when confronted
with a stall event.
2.2. Evaluation criteria for a recovery from a stall or
approach-to-stall that does not mandate a predetermined value for
altitude loss and should consider the multitude of external and
internal variables which affect the recovery altitude.
2.3. Realistic scenarios that could be encountered in operational
conditions including stalls encountered with the autopilot engaged.
2.4. Pilot training which emphasizes treating an “approach-to-stall” the
same as a “full stall,” and execute the stall recovery at the first
indication of a stall.
2.5. Incorporation of stick pusher training into flight training scenarios, if
installed on the aircraft.

Display of Angle of Attack in the Cockpit

A quick note on the angle of attack display in the cockpit. Most Part 23
and Part 25 airplanes do not include an angle of attack in the cockpit, even
though it is the primary means to identify whether an airplane is approaching
stall speed. This author recommends the use of an angle of attack indicator
for improved situation awareness for the crew.
The angle of attack (AoA) for maximum lift coefficient varies with alti-
tude, Mach, airplane configuration, and (as we will discuss next) environ-
mental conditions, so the actual AoA provided to the crew is typically a
normalized AoA. The normalization is done by ratioing the measured AoA
with a reference value so the display provides the crew with a sense of
margin to stall no matter what the flight condition/configuration of the
airplane is.
An example of normalization of the angle of attack (AoAN ) could be as
follows:
a  a0
AoAN ¼ (6:9)
aevent  a0
where
a ¼ measured angle of attack (can be local and can be converted to
fuselage AoA)
a0 ¼ zero lift angle of attack for the configuration and flight condition
(in the same reference plane as the measure AoA)
aevent ¼ the AoA value that the crew needs to know
 CHAPTER 6 Lift and Stall 193

Equation (6.9) does linearize the lift curve slope from the zero lift con-
dition all the way to the event, which can be the airplane angle. We have
seen that this lift curve is probably not linear, so the normalized AoA is
not really representing an angle of attack but, as we said, is meant to
provide a margin to an event.
This author prefers the selection of the following events as reference:
• Set the minimum normal flight speed in steady level flight to a normalized
AoA of 0.6. These speeds can be V2 or V2 þ 10 for takeoff (the equivalent
lift coefficient) flaps, VREF for landing flaps, and 1.3 g maneuver margin
(40-deg bank; see Chapter 14) for flaps up.
• Set the stall warning condition to a normalized AoA of 0.8. The angle of
attack for stall warning is a precise value in the airplane’s stall protection
system. This value should be an exact one, as measured.
• Set the normalized AoA of 1.0 near the natural stall (if still identification is
not dependent on a stall prevention device) or at the angle of attack for stall
identification (stick pusher, for example).
A display that can be used for this scheme is shown in Fig. 6.26. (The author
had to be creative given the limited color used in this printing.) The color
scheme follows the accepted aviation colors for the situation. The green
zone is for normal operation; it extends from a normalized AoA of 0.6
(minimum normal speed) to a very low value (high speed). Between 0.6
and 0.8, the color selected is yellow (or amber), which is the color for a
caution; it represents an area of reduced margin to stall warning. The crew
should expect stall warning at exactly 0.8. The area with a normalized AoA
greater than 0.8 is red, which means immediate crew action is required
(reduce AoA).
One of the advantages of this scheme is that when the airplane flies at the
minimum speed and starts maneuvering (increased load factor),
the increased angle of attack of the airplane at the minimum speed will
increase, and the pilot will become aware that the margin to stall warning
is reducing.

Red

Icing 1.0
0.8 Yellow
Icing is a broad subject, and we intend to
provide only enough information to get a
feel for the typical impact of icing on the cre- AoA
0.6
ation of lift by changing the shape of the air-
plane (mass collected, roughness deposited 0.4
Green
on the wing leading edge). The wing and
tail are particularly impacted by icing, and
sometimes just a little ice can cause a lot Fig. 6.26 Normalized angle of
of trouble. We covered the fundamentals attack display for crew.
194 Operational Aircraft Performance and Flight Test Practices

of icing in our previous textbook [1]. We noted the following major

behaviors:
• Ice forms on the airplane because of supercooled water droplets in liquid
form.
W Ice crystals will not collect on the airplane but may block pitot tubes.

W The quantity of water in the air that could form ice on the airplane is

expressed in the form of liquid water content (LWC).

• Ice collects on the leading edges (fuselage, wing, tail) of the airplane.
• Faster speed means increased collection mass.
W The ice catch (mass flow rate of water droplets impinging on the leading

edge) is proportional to the true airspeed V, the frontal area of the

exposed shape Sf , the LWC, and the ice collection efficiency of the
shape h [see Eq. (6.10) and Fig. 6.27].
• Smaller shape/wing means increased collection.
W The wing will collect more ice than the fuselage.

W The wing tip (on a tapered wing) will collect more ice than the wing root.

W The horizontal tail is more susceptible to collecting ice than most of

the wing.
• A sharper leading edge results in increased collection efficiency (more
efficient at collecting ice).

ṁimpinge ¼ h V Sf LWC (6:10)

S
h¼ (6:11)
Sf

Rime ice is generated when supercooled water droplets impinging on the

airplane freeze on contact (cold atmospheric conditions). The ice collected
will be white (because it contains trapped air), and the shape will typically
be limited to the impingement limits (Fig. 6.28). Glaze ice is generated
when the droplets impinge on the airplane and run back some distance
before freezing (warmer conditions, closer to the freezing point). The ice col-
lected tends to be clear, and the shape tends to form horns over time. Mixed-
phase ice contains both rime and glaze ice.
Since the publication of [1], the regulations have evolved significantly.
FAA Part 25, Appendix C, Atmospheric Icing Conditions, is still the norm

Sf

Fig. 6.27 Portion of the frontal area that will impinge the leading edge. Source: [3].
 CHAPTER 6 Lift and Stall 195

Rime ice Mixed-phase ice Glaze ice

Fig. 6.28 Icing types. Source: NASA.

for airplane icing requirements but was expanded with the addition of pre-
scriptive icing scenarios (Part II, Airframe Ice Accretions for Showing Com-
pliance with Subpart B) in Amendment 25-121 (2007), which we introduce in
the following section. FAA Part 25, Appendix O, Supercooled Large Drop
Icing Conditions (SLD), was added in Amendment 25-140 (2015). FAA
Part 33, Appendix D, Mixed Phase and Ice Crystal Icing Envelope (Deep
Convective Clouds), came into effect in Amendment 33-34 (2014).
The atmospheric conditions leading to icing were also clarified in the new
section I of Part 25, Appendix C in Amendment 25-121. Takeoff maximum
icing was added to continuous maximum and intermittent maximum icing.

Preactivation (or Delayed Turn-On – DTO) Ice

Preactivation ice consists of the initial ice catch on the airplane as it enters
icing conditions, prior to the pilot or ice detection system detecting the
condition and activating the ice protection system (IPS) and the system
becoming effective at removing the ice. Per FAA Part 25, Appendix C, Part
II(e), this ice shape means that the airplane is exposed to continuous
maximum atmospheric icing conditions. The resulting ice shape is typically
very thin and covers the leading edges of the airplane (see Fig. 6.29). It is often
simulated in flight testing by the application of sandpaper to the leading
edges, mostly the wing.
Under these conditions, the applicant will verify the airplane’s handling
characteristics with minimum testing consisting of proving that the airplane
is still compliant with §25.143(j), General Handling, and §25.207(h) and (i),
Stall Warning. It should be noted that some airfoils can lose significant lift
with preactivation ice; see Fig. 6.29c, where about 30% of the maximum lift
is lost with only sandpaper on the leading edge of the airfoil.
§25.143 General
(j) For flight in icing conditions before the ice protection system has been
activated and is performing its intended function, it must be
demonstrated in flight with the most critical of the ice accretion(s)
defined in Appendix C, part II, paragraph (e) of this part and Appendix O,
part II, paragraph (d) of this part, as applicable, in accordance with
§25.21(g), that:
 196
Operational Aircraft Performance and Flight Test Practices
a) b) c) 2.0
1.8
Rivulets
1.6
1.4
1.2
1.0
0.8
0.6
0.4 Re = 7.5 × 106, Ma = 0.21
0.2
Clean
0.0 40-grit sandpaper
–0.2 80-grit sandpaper
Ice shape 290
1- –0.4
–0.6
–8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20
1-
Rivulets
 (deg)

Fig. 6.29 Example of preactivation ice. Source: FAA AC 20 –73A, Aircraft Ice Protection.
 CHAPTER 6 Lift and Stall 197

(1) The airplane is controllable in a pull-up maneuver up to 1.5 g load

factor; and
(2) There is no pitch control force reversal during a pushover maneuver
down to 0.5 g load factor.

§25.207 Stall Warning

(h) The following stall warning margin is required for flight in icing
conditions before the ice protection system has been activated and is
performing its intended function. Compliance must be shown using the
most critical of the ice accretion(s) defined in Appendix C, part II,
paragraph (e) of this part and Appendix O, part II, paragraph (d) of this
part, as applicable, in accordance with §25.21(g). The stall warning
margin in straight and turning flight must be sufficient to allow the pilot
to prevent stalling without encountering any adverse flight
characteristics when:
(1) The speed is reduced at rates not exceeding one knot per second;
(2) The pilot performs the recovery maneuver in the same way as for
flight in non-icing conditions; and
(3) The recovery maneuver is started no earlier than:
(i) One second after the onset of stall warning if stall warning is
provided by the same means as for flight in non-icing
conditions; or
(ii) Three seconds after the onset of stall warning if stall warning is
provided by a different means than for flight in non-icing
conditions.
(i) In showing compliance with paragraph (h) of this section, if stall warning
is provided by a different means in icing conditions than for non-icing
conditions, compliance with §25.203 must be shown using the accretion
defined in appendix C, part II(e) of this part. Compliance with this
requirement must be shown using the demonstration prescribed by
§25.201, except that the deceleration rates of §25.201(c)(2) need not
be demonstrated.

The new Part 23 stipulates under §23.2150, Stall Characteristics, Stall

Warning and Spins (underline added by author), “(a) The airplane must
have controllable stall characteristics in straight flight, turning flight, and
accelerated turning flight with a clear and distinctive stall warning that pro-
vides sufficient margin to prevent inadvertent stalling.” Then regulations per-
taining to icing fall under §23.2165.
§23.2165 Performance and flight characteristics requirements for flight
in icing conditions.
(a) An applicant who requests certification for flight in icing conditions
defined in part 1 of appendix C to part 25 of this chapter, or an
applicant who requests certification for flight in these icing conditions
and any additional atmospheric icing conditions. . .
(2) The means by which stall warning is provided to the pilot for flight in
icing conditions and non-icing conditions is the same.

Providing the same stall warning in icing and nonicing does lead to more
standardized training for pilots and improves the chances that the pilots’
reaction to an atmospheric condition low-speed situation will be appropriate.
Then, the responsibility to adjust the stall warning system based on
198 Operational Aircraft Performance and Flight Test Practices

atmospheric conditions falls on the manufacturer. That includes the stall

warning system and the normalized AoA, if installed.

Takeoff, Final Takeoff, and En Route Ice

Takeoff ice is the most critical ice accretion on unprotected surfaces, as
well as is any ice accretion on the protected surfaces appropriate to normal
ice protection system operation occurring between the end of the takeoff dis-
tance and 400 ft above the takeoff surface, assuming accretion starts at the
end of the takeoff distance in the takeoff maximum icing conditions. Final
takeoff ice is the most critical ice accretion on unprotected surfaces, as well
as any ice accretion on the protected surfaces appropriate to normal ice pro-
tection system operation between 400 ft and 1500 ft above the takeoff
surface. Both of these ice shapes are accrued with takeoff flaps down.
En route ice is the critical ice accretion on the unprotected surfaces, as
well as any ice accretion on the protected surfaces appropriate to normal
ice protection system operation during the en route phase. This is in the
flaps-up configuration, one engine inoperative.

Holding Ice
Per Part 25, Appendix C, holding ice is critical ice accretion on unpro-
tected surfaces, as well as any ice accretion on the protected surfaces appro-
priate to normal ice protection system operation during the holding flight
phase. The resulting ice shapes for the holding scenario are expected to be
the largest ice shapes to accumulate on the airplane under normal operation.
FAA AC 25.1419-1 identifies an acceptable methodology to define
holding ice. It states:
The applicant should determine the effect of the 45-minute hold in continu-
ous maximum icing conditions. It should be assumed that the airplane will
remain in a rectangular “race track” pattern, with all turns being made
within the icing cloud. Therefore no horizontal extent correction should
be used for this analysis. The applicant should substantiate the critical
mean effective drop diameter, LWC, and temperature that result in the for-
mation of an ice shape that is critical to the airplane’s performance and hand-
ling qualities. The critical ice shapes derived from this 45-minute hold
analysis should be compared to critical shapes derived from other analyses
(i.e., for climb, cruise, and descent) to establish the most critical artificial
ice shapes to be used during dry air flight tests. Not only is the thickness
important, but the shape, location, and texture may contribute to adverse
aerodynamic characteristics.
Note: The 45-minute hold condition will result in large ice shapes on the
unprotected surfaces that, from visual assessment alone, would appear to
have a drastic effect on airplane performance and handling characteristics.
Operational history and wind tunnel tests have shown that the ice that
forms on the protected and unprotected surfaces before activation and
normal operation of the ice protection system may also have a detrimental
effect on the lift and pitching moment characteristics of the wing and
 CHAPTER 6 Lift and Stall 199

horizontal tail airfoils. Similarly, for airplanes with cyclic ice protection
systems, the ice that accretes on the protected surfaces between cycles
may have more than an insignificant contribution to the total effect of ice
on the airplane performance and handling characteristics, when combined
with the 45-minute holding ice accretion on the unprotected surfaces.

A search of the Internet for pictures of simulated ice shapes installed on test
airplanes produced the images shown in Fig. 6.30, which show the large
shapes that can be expected for worst case holding ice, typically horn-shaped
ice typical of warm temperature glaze ice accretion.

Approach and Landing Ice

FAA Part 25, Appendix C defines approach ice as “the critical ice accre-
tion on the unprotected surfaces, and any ice accretion on the protected sur-
faces appropriate to normal ice protection system operation following exit
from the holding flight phase and transition to the most critical approach
configuration.” Most manufacturers will recommend that the airplane hold
with flaps up near minimum drag airspeed. As the airplane transitions
from the holding pattern to the approach to land condition, the pilot will
reduce airspeed (change in angle of attack), lower the flaps, and extend the
landing gear. This will expose new surfaces to the airflow that may not be
protected by an ice protection system and therefore may collect ice, such
as the leading edge of the flaps. This ice may impact the airplane’s handling
characteristics and stall speeds.
Similarly, landing ice is defined as “the critical ice accretion on the unpro-
tected surfaces, and any ice accretion on the protected surfaces appropriate
to normal ice protection system operation following exit from the approach
flight phase and transition to the final landing configuration.” This adds more
ice to the airplane.

Fig. 6.30 Examples of simulated ice shapes used in flight test.

200 Operational Aircraft Performance and Flight Test Practices

FAA AC 25-25A, Table 4.5, provides a minimum acceptable scenario for

ice accumulation in natural icing conditions for approach and landing ice.
The airplane first accumulates 0.5 in. of ice (on the most critical unprotected
surface) in the holding condition. Then the airplane is slowed down to the
first intermediate flap (typically a takeoff flap condition), where it must
accumulate another 0.25 in. of ice before executing some maneuvers for
handling assessment. Then it repeats the same condition for every other
intermediate flap condition (0.25 in. per intermediate flap). Finally, a last
0.25 in. of ice will accumulate on the landing flap gear-down configuration,
and the handling of the airplane will be assessed, including a 1-kt/s stall.
The more a manufacturer offers intermediate flaps (multiple takeoff flaps
configurations), the more ice that must accumulate.
Prior to executing natural ice testing, a manufacturer must perform simu-
lated ice shapes testing with a best estimate of where the ice may accumulate
in the previously mentioned scenario. This is typically done by using the
holding ice shape as a starting point (more adverse than the natural ice
test initial condition) and adding ice shapes in the expected (by analysis)
locations based on an equivalent maneuver. In this way, the manufacturer
has already collected baseline flight characteristics with holding ice and
then defines a delta behavior from there under controlled conditions.

Failure Ice
Failure ice is the critical ice shape that is developed on the airplane on
normally protected surfaces when the anti-icing/deicing system fails based
on the time for the crew to detect the failure and exit the icing condition.
If no automatic warning is provided to the crew, one could expect this ice
shape to be based on a flight time exposure of as much as 22.5 min. (half
of the 45-min. hold condition).

Intercycle Ice
When an airplane is equipped with a deicing system (discussed later in
this chapter), ice is removed in cycles, and the deicing system often does
not remove 100% of the accumulated ice. The ice that is left onto the wing
(or tail) after the cycling of the deicing system is called intercycle ice (see
Fig. 6.31). This ice is unevenly distributed spanwise and chordwise on the
wing (or tail). The airplane must be shown to have acceptable flight charac-
teristics with intercycle ice.

Ridge Ice
Ridge ice forms behind the ice protection system, typically as a result of
runback and refreeze. This is more likely to occur under “warm” icing
 CHAPTER 6 Lift and Stall 201

Ice limit
0.006 Small rime
feathers
Resid
ual
ice

Clean

Res
idu
ice al Small
rim
feath e Frost
Ice
0.002 ers
limit

Fig. 6.31 Example of intercycle ice. Sources: [6, 7].

conditions (near the freezing point) where the liquid water content (LWC) is
typically higher and the water droplets run back beyond the protection
system before freezing. Upper surface ridge ice can have a more detrimental
impact on the wing lift than ice accumulation on the leading edge. Figures
6.32 and 6.33 show examples of ridge ice and the resulting loss of lift.

Ice-Contaminated Tailplane Stall

This section discusses a special situation of stall that does not impact the
reported stall speed. It is the phenomenon of ice-contaminated tailplane stall
(ICTS). FAA AC 25.1419-1 defines ICTS as:
This phenomenon occurs due to airflow separation on the lower surface of
the tailplane. This can occur if the angle-of-attack of the horizontal tailplane
exceeds the stall angle-of-attack, which can be reduced by even small quan-
tities of ice on the tailplane leading edge. The increase in tailplane
angle-of-attack can result from airplane configuration (e.g., increased flap
extension increasing the downwash angle or trim required for the
center-of-gravity position) and flight conditions (e.g., a high approach
speed resulting in an increased flap downwash angle, gusts, maneuvering,
or changes to engine power setting). ICTS is characterized by a reduction
or loss of pitch control or stability while operating in, or after recently
departing from, icing conditions. For airplanes with unpowered longitudinal
control systems, the pressure differential between the upper and lower sur-
faces of the stalled tailplane may result in a high elevator hinge moment,
forcing the elevator trailing edge down.

One of the first events attributed to ICTS was an accident involving a

Vickers Viscount in Stockholm, Sweden in 1977 (see the following
accident box). Testing for ICTS involves maneuvers at varying load factors
from 0.5 to 1.5 g to evaluate tail effectiveness and try to detect loss of
control authority.
 202
Operational Aircraft Performance and Flight Test Practices
Clean
k = 0.63” (k/c = 0.0035)
k = 0.80” (k/c = 0.0044)
k = 0.94” (k/c = 0.0052) Ice ridges
1.6 k = 0.125” (k/c = 0.0069) 0.12
8
1.4
1.2 0.08 6
1.0 Polished aluminium
4
0.8 0.04

Y (in.)
0.6 2

Cm
CL
0.4 0.00 Heated
0 region
0.2
Polished aluminium
0.0 –0.04 –2
–0.2
–4
–0.4 –0.08 0 5 10 x (in.) 15 20
–0.6
–0.8 –0.12
–12–10–8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20
 (deg)

Fig. 6.32 Ice ridge effects on the lift coefficient. Source: [7].
 0.08
0.06
2.0 0.20 0.04
Clean, ref. 22 EG1162 streamwise
1.8 0.18 NG0671 Runback ridge 0.02

y/c
1.6 0.16 EG1162 Streamwise, ref. 22 0.00
EG1164 Horn, ref. 22 –0.02
1.4 0.14
EG1159 Spanwise ridge, ref. 22 –0.04
1.2 0.00 0.05 0.10 0.15
0.12 0.08 x/c
0.08
Cl 1.0 0.10
0.07 0.06
0.8 0.08 0.04
0.06 EG1164 horn
0.6 0.06 Cm 0.02

y/c
0.4 0.04 0.05 0.00

0.2 0.02 CD 0.04 –0.02

–0.04
–0.0 0.00 0.00 0.05 x/c 0.10 0.15
0.03
0.08
–0.2 –0.02
0.02 0.06
–0.4 –0.04
0.04
–0.6 –0.06 0.01 EG1159 spanwise ridge

y/c
0.02
–0.8 –0.08 0.00 0.00
–8 –6 –4 –2 0 2 4 6 8 10 12 14 16

CHAPTER 6
–8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20
–0.02
 (deg)
 (deg) –0.04
0.00 0.05 x/c 0.10 0.15

Lift and Stall

Fig. 6.33 Aerodynamic impact on simulated ice ridge. Source: [8].

203
204 Operational Aircraft Performance and Flight Test Practices

Vickers Viscount accident in Stockholm, Sweden, Jan. 1977: The aircraft

was completing a regularly scheduled flight from Kristianstad to Stockholm-
Bromma with intermediate stops in Växjö and Jönköping, carrying 19 passen-
gers and a crew of 3 on behalf of Linjeflyg. On approach to Bromma Airport,
the crew encountered marginal weather and icing conditions. At an altitude of
1150 ft and a speed of 137 kt, flaps were deployed at an angle of 40 deg when
control was lost. The airplane entered a nose-down attitude and crashed at an
angle of 110 deg at a speed of 210 kt in a parking lot located in the residential
area of Kälvesta, about 4.5 km short of the runway 12 threshold. The aircraft
was totally destroyed by impact forces and a postcrash fire, and all 22 occu-
pants were killed. At the time of the accident, weather conditions were con-
sidered as marginal with icing conditions and a horizontal visibility of 5 km
with a cloud base at 700 ft.
Probable cause: Loss of control on final approach due to an excessive
accumulation of ice on the horizontal stabilizers after the crew lowered the
flaps from 32 deg to 40 deg. When control was lost, both right engines
number 3 and 4 were not running at full power for undetermined reasons,
which caused the deicing system to produce an insufficient temperature.
Source: https://www.baaa-acro.com/crash/crash-vickers-838-viscount-
stockholm-22-killed

500 Flap 32° Flap 40°

Height above ground, m

400

300

200

100

0 100 200 300 400 500

Horizontal distance, m

Ice Protection Systems

Airplanes that are authorized to operate in known icing condition [flight
into known icing (FIKI)] will have protection systems on the airplane. The
location and type of protection are left to the discretion of the manufacturer
and will greatly impact the ice accretion on the airplane. This is acceptable as
long as the flight test shows that the airplane meets the flight requirements
with the protection selected. At a minimum, all airplanes (even those not
authorized to fly into icing conditions) will have ice protection systems for
the air data probes. Then, part of most of the wing leading edge will be pro-
tected (see Fig. 6.34) along with the engine air intake lip (and possibly part of
the intake duct). The horizontal and vertical tail may or may not be protected,
again at the discretion of the manufacturer and demonstration of acceptable
 AOA
TAT probe vanes
Pitot (RH side only)
static
S-duct probe

Ice
detectors Static ports
Waste water drain mast
Leading edges Standby pitot
Engines (LH side only)
Front and Engine
air intakes Side air inlet
side windows Wing
window (cowls)
Rain repellent leading
Brakes (option) Windshields edges
Window system
heating
Wing anti ice
Engine anti ice
Fixed
Air data leading edge Windshield
Waste water drain mast Side
Probe heating probes root wipers
window
Legend
Pneumatic anti-ice.
Electrical heaters.
Vertical stabilizer
de-icing boot

Elevator horn
heating mat Stabilizer de-ice

Outboard horizontal
stabilizer Electrically heated
Propeller de-icing boot propeller blade
heating mat deicers
Inboard horizontal Windshield
stabilizer wipers Side slip
de-icing boot sensor Pneumatic
leading edge
deicers
Stabilizer de-ice
Electrically
Windshield heating heated
elements Propeller de-ice windshield

CHAPTER 6
Windshield wipers Windshield anti-ice Left Bypass duct
and washers angle deicer
Windshield of attack
Heated static Inboard wing Mid wing wiper sensor
plates de-icing boot Outboard Ice detector
Engine de-icing probe Pneumatic engine
TAT Heated boot wing
intake de-icing Wing de-ice inlet lip deicer
probe AOA vanes hot air boot Ice indicator Wing de-ice Electrically heated
anti-icing
Heated Static pitot-static tubes
Ice pitot plates Probe heat Engine anti-ice TAT sensor
detector head

Lift and Stall

Fig. 6.34 Various protected zones on airplanes.

205
206 Operational Aircraft Performance and Flight Test Practices

flight characteristics with ice shapes if unprotected. The windshield will be

protected to ensure crew forward field of view. Finally, any other surface
deemed necessary by the manufacturer will be protected.
Most Part 25 airplanes and some Part 23 airplanes are certified with an ice
detection system (one or two probes) that are designed to increase crew
awareness of the presence of icing conditions and, should they be proven
reliable enough, automatically turn on the ice protection system of the air-
plane if so designed by the manufacturer.

Anti-icing Systems
An anti-icing system is designed to prevent ice from forming on the
surface it is protecting. This is typically done in the form of an evaporative
system where sufficient heat is provided to prevent the water droplets
hitting the plane from freezing. (The droplets evaporate on contact.) These
systems are energy consumption intensive, and their application tends to
be limited to the most critical sections of the airplane as defined by the man-
ufacturer. Air data probes, engine inlets, and part of the wing are typical
anti-icing system locations. That energy is extracted from the engine, typi-
cally through bleed air for wing and engine inlet protection.
In order to reduce some of the energy consumption from an anti-icing
system, some manufacturers will allow the wing to “run wet” in some
locations by reducing the amount of heat energy provided by the protection
system. The water droplets then hit the surface, partially evaporate, but then
run back to freeze on the unprotected part of the airplane (may create
runback ice). This is an acceptable scenario as long as the resulting ice
shapes are proven to not adversely impact the flight characteristics of the air-
plane in flight testing. An example of runback ice collected in flight testing is
shown in Fig. 6.35.

Fig. 6.35 Ice accumulation under the wing. Source: [4].

 CHAPTER 6 Lift and Stall 207

Deicing Systems
A deicing system is one in which ice is allowed to accumulate for a short
period of time before mechanical work or heat energy is applied to remove
part or all of the accumulated ice. These systems require less energy to
operate, but the airplane does operate under a more ice-contaminated con-
dition. We discuss the two systems in the following sections.

Deicing Boots
Deicing boots consist of a thick rubber membrane with cavities that is
applied over the leading edge of a surface that needs to be protected from
icing conditions. They were invented by B.F. Goodrich Corporation around
1930. In nonicing conditions, the membrane is kept deflated and nearly con-
forms to the original contour of the surface where it is applied (see Fig. 6.36).
In icing conditions, boots are inflated (either manually or automatically) at
regular intervals to break up the accreted ice, which is then partially shed
by the airflow (Fig. 6.37), leaving some residual ice on the leading edge.
After a few cycles, the accretion of ice reaches an equilibrium in general
dimension, the intercycle ice (see Fig. 6.31 earlier in the chapter).
It was noted as far back as 1956 that although the boots provide some
protection, they do not create a clean wing, and their simple activation,

a) b)

9.9
.75 .75
? 5 .75
.75 .7
1
1
5

Boot thickness,
1.2

0.10 in.
1.2
5

Last 6 tu
1 bes may
1 be oper
.75 .7 ated inde
5 .75 penden
tly
.75 .75
21.5 .75 .75
.75 .75 .75
24.5 .75

Fig. 6.36 (a) Deflated boot; (b) inflated boot. Sources: Pictures from DOT/FAA/AR-06/48,
Investigations of Performance of Pneumatic Deicing Boots, Surface Ice Detectors, and Scaling of
Intercycle Ice, 2006; drawings from NACA TN 3564, Effect of Pneumatic De-Icers and Ice Formation
on Aerodynamics Characteristics of an Airfoil, 1956.
208 Operational Aircraft Performance and Flight Test Practices

Just before boot activation Boots inflated

Inflated boot cells

Fig. 6.37 Residual ice on wing deicing boots during activation.

even in nonicing conditions (see Fig. 6.38) leads to some loss of lift and
increased drag simply due to the fact that they do deform the leading edge.
Figure 6.39 was extracted from NACA TN 3564. It can be seen that
deicing boots do provide some protection from icing, but the airplane drag
never fully goes back to the preice encounter value, even after boot

.04
moment coefficient,
Change in pitching-

( C/4 1
Cm - Cm

0
( C/4

Boot deflated
A tubes inflated
–.04 B tubes inflated
A tubes inflated, forward part only
1.0 B tubes inflated, forward part only
Airfoil with standard roughness (ref. 10)

.8 .032
Lift coefficient, CL

.6 .024
Drag coefficient, CD

.4 .016

.2 .008

0 4 8 12 0 .2 .4 .6 .8 1.0
Angle of attack, deg Lift coefficient, CL

Fig. 6.38 Aerodynamic impact of deicing boot inflation in nonicing conditions. Source: [9].
 CHAPTER 6 Lift and Stall 209

a) Icing period, 3.9 minutes b) Icing period, 0.9 minutes

coefficient, moment coefficient,

Change in pitching-

– Cm 1
) C/4
.02
(C/4
0

Change in pitching-moment coefficient,

Cm

.20
Before ice removal
.18 After ice removal

– Cm 1
Lift

(
CL

Clean airfoil

C/4
.028

(
Before ice removal
After ice removal

C/4
A tubes inflated

Cm
B tubes inflated
.024 Boot inoperative .04

.020 0
Drag coefficient, CD

Lift coefficient,
.22

.016
CL
.18

.012 .012
Drag coefficient, CD

.008 .008

.004 .004
0 8 16 24 32 32 40 48 56
Icing time, min

Fig. 6.39 Impact of deicing boot activation on airplane lift and drag. Source: [9].

inflation/deflation, and that some lift is lost prior to the inflation of the boot
and not fully recovered after ice shedding. The drag increase is larger the
longer the interval between boot inflations (more ice accumulating), but
too short of an inflation cycle may lead to greater intercycle ice (boot not
as effective on small accumulations, thus buildup occurs without shedding)
or simply greater basic drag of inflated boot (time-averaged drag over cycle
time).
Some of the flight manuals reviewed by this author for airplanes equipped
with deicing boots include the following statements:

• Do not use below SAT of –408C, minimum speed in icing 140 KIAS, max
speed 226 KIAS (windshield heat). Abrupt maneuvering and steep turns at
low speeds must be avoided because the airplane will stall at higher than
published speeds with ice accumulation. On final approach for landing,
210 Operational Aircraft Performance and Flight Test Practices

increased airspeed must be maintained to compensate for this increased

stall speed. In flight, the boots should be cycled once every time the ice
accumulation is approximately 0.5–1 in. thick.
• Minimum airspeed of 105 KIAS, inflated boots increase stall speed by 10
KIAS. An accumulation of 1 in. of ice on the leading edges can cause a large
(up to 500 FPM) loss in rate of climb, a cruise speed reduction of up to 40
KIAS, as well as a significant buffet and stall speed increase (up to 20 kt).
Even after cycling the deicing boots, the ice accumulation remaining on the
unprotected areas of the airplane can cause large performance losses. With
residual ice from the initial 1-in. accumulation, losses up to 200 FPM in
climb, 20 KIAS in cruise, and a stall speed increase of 5 kt can result.
• Do not cycle the boots during landing (below approximately 500 ft
above ground level) because boot inflation may increase stall speed by as
much as 10 kt.
• The aircraft can sustain a reduction in airspeed from 165 to 105 kt after
repeated operation of the deicing boots in icing conditions, accompanied
by a significant increase in the aircraft’s stall speed from 78 to 92 kt. As
well, the aircraft’s ability to sustain a climb or maintain altitude is reduced,
from an initial climb of 800 fpm to a descent of 100 fpm.
• Wing/tail deice boot off for takeoff and landing, minimum airspeed in
icing 160 KIAS, inflation cycle on auto is 2 min, do not use boot below
TAT ¼ –308C.
• Boot failure, flap 0 landing, minimum speed 134 KIAS, boots ops flap 15
landing. With operational boots, Pusher Ice Mode increases stall speed by
11 kt for flap 0 and 9 kt for flap 15.
Clearly, the use of deicing boots does provide some ice protection capability
to the airplane, but they come with additional operational (speed) restric-
tions. As well, pneumatic boots have proven to be maintenance intensive,
possibly requiring replacement on a yearly basis.

Electro-expulsive Systems
Another, more recent, deicing system is the NASA-developed electro-
expulsive separation system (EESS). The EESS consists of layers of conduc-
tors encased in materials that are bonded directly to the airframe structure.
When ice accumulates on the aircraft, the system is activated and an electric
current is sent through the conductors, causing them to pulse. (They tap the
skin, providing a strong impulse that mechanically breaks the ice bond to the
structure. The ice is then carried away by the airflow.) The EESS technology
has been licensed to Ice Management Systems, Inc. (IMS).
A similar technology that also mechanically removes ice from the surface
by impulse is the Cox Systems Electro-Mechanical Expulsion Deicing System
(EMEDS). The design of the deicing actuator, which is a rolled-up printed
circuit, enables the system to function on substantially less energy. Starting
out as a flat oval, the actuator’s shape changes to a circle when electrical
 CHAPTER 6 Lift and Stall 211

energy is applied. This change causes the actuator to impact the inside of the
leading edge surface, which responds with a small but rapid flex movement
that expels the accumulated ice from the surface of the aircraft’s erosion
shield [10]. EMEDS has been shown to remove ice to within 0.030 in. thick-
ness. The EMEDS’s metal leading edge surface enables it to last for the life of
an airplane.
Both systems can operate continuously after the first sign of ice. (They are
deicing systems, but can typically operate at 1-min intervals rather than 0.25
to 0.5 in. of ice accumulation like the boots.) As well, they require very little
energy, which makes them good candidates for the coming electric airplanes.

Manufacturing Tolerance: An Aerodynamic Perspective

The last subject we will cover in this chapter on lift is manufacturing
tolerances and their impact on the airplane’s aerodynamics. An ideally
built airplane will conform to the engineering drawings, and any airplane
off the production line will match the airplane used in flight testing to
verify the airplane’s compliance to the type design. But of course, conditions
are not always ideal.
One of the first tasks in the design of a new airplane is to define aerody-
namic zones from which to establish production tolerances (see Fig. 6.40).
Two or three zones are typically specified, with zone 1 having the higher
requirements for manufacturing tolerance, zone 2 providing more freedom,
and zone 3 even more. Zone 1 represents the leading edge of a surface, and
the zone’s focus is to avoid the flow becoming turbulent too soon (prevent
drag increase beyond design goal) or causing flow separation (for drag and lift).
The document that contains this information, often labeled “Aerody-
namic Smoothness Specification,” is usually reproduced on an engineering
drawing that becomes part of type design. The document provides definitions
(e.g., what is a forward facing step; see Fig. 6.41). It defines the maximum
expected steps and gaps per zone and what type of fastener flushness can
be tolerated. It defines skin waviness and acceptable deviation from the
design loft. These deviations are measured in thousandths of inches (or frac-
tions of millimeters) with a typical value for zone 1 of around 0.030 in.
(0.76 mm). This author joked with an inspector once who asked him when
looking at the wing tip of a small business jet, “How do you know if this
wing tip is within 0.030 in. of the design?” This airplane had a flexible
wing, so the answer was, “Let me push it down 1 in. Now it is. . .”. What is
important is that one must establish a methodic inspection practice through-
out the build of the airplane to ensure that what is built in production is the
same as the prototypes used in the flight test validation of the design, either
for initial type design or follow-on modification with validation.
Of course, the smaller the tolerance, the better the aerodynamic results;
however, there is a cost for high tolerance. This author has seen a request for
a maximum step in zone 1 of a wing of 0.005 in. at the junction of the leading
 212
Operational Aircraft Performance and Flight Test Practices
Rear spar

FR 58
A/C upper surface
A/C upper surface

A/C lower surface

A/C lower surface

Front spar

Degree of smoothness Degree of smoothness

Zone 1 High sensitivity Rear spar
Zone 1 High sensitivity
Zone 2 Medium sensitivity Rear spar Zone 2 Medium sensitivity
Zone 3 Low sensitivity Front spar
Frame C35

Front spar

Fig. 6.40 Aerodynamic smoothness zones. Source: [11].

 Longitudinal ±5 deg
axis

Step: + 0.1 mm – 2 mm
(+ 0.004 in. – 0.078 in.) As-built local contour
Flow direction
Loft
Tolerance
Step height
Radome Fuselage

Gap: 2 mm ± 1 mm Forward facing step

(0.078 in. ± 0.039 in.)

CHAPTER 6
Fig. 6.41 Aerodynamic smoothness information.

Lift and Stall

213
214 Operational Aircraft Performance and Flight Test Practices

edge to the main wing in order to protect the laminar flow. It took the man-
ufacturing team over a week to install the first leading edge and meet the
requirement; in contrast, if the maximum step had been allowed to be
0.010 in. (for that airplane), they would have been done on the first day.
If part of a zone requires even more tolerance than the general zone, one
should limit the extent to the smallest dimension possible. This could be the
case around a static pressure port for an RVSM-qualified airplane, for
example, where we saw in Chapter 4 that a small deviation in measured
pressure can easily make the airplane not meet the requirements.
During the assembly of the airplane, after individual parts are built to
within tolerances, the focus should shift to building a symmetrical airplane
(or at least the design symmetry). One should verify the wing twist left to
right at the same wing station, for example, and see if there is an acceptable
difference (ideally zero). A wing twist difference left to right will generate roll
trim requirements in flight. We should also remember that stall depends on
handling, and the airplane can have only so much roll before it is an issue.
Most airplanes can take up to a 0.25-deg difference left to right, but that
may need to be reduced on more sensitive airplanes.
Flight controls rigging is another aspect to verify. The rigging is, by defi-
nition, a moving surface, and misrigging will cause an otherwise symmetric
airplane to become unsymmetric. This applies to all surfaces, primary and
secondary.
Finally, the symmetry of the assembled airplane should be verified for
general symmetry of build. One should locate easily identifiable positions on
the airframe (e.g., the joint between two skins at a spar) so the people making
the measurements can pick the proper location to measure (see Fig. 6.42).

HR

TR TL HL
VAR
VAL
VFR VFL
WR

WL

NR
NL

Fig. 6.42 Final symmetry check.

 CHAPTER 6 Lift and Stall 215

This check becomes a verification of the alignment of the major aerodynamic

surfaces, and it is done prior to the first flight of the airplane.

References
[1] Asselin, M., An Introduction to Aircraft Performance, AIAA Education Series, AIAA,
Reston, VA, 1997.
[2] Horn, S., “SJ30-2 Flight into Known Icing Certification Program,” FAA Designee
Conference, Fort Worth, TX, 2006.
[3] Horn, S.Asselin, M.Baumgardner, D., “Observations of Ice Ridge Formation on the
SJ30-2 Leading Edge Slat in Freezing Drizzle Conditions,” SAE Technical Paper
2007-01-3365, 2007.
[4] Abbott, I. H.von Doenhoff, A. E., Theory of Wing Sections, Dover, New York, 1949.
[5] Fujino, M., “Development of the HondaJet,” Honda R&D, https://www.hondarandd.
jp/point.php?pid=463&lang=en.
[6] U.S. Department of Transportation and Federal Aviation Administration, “Effect of
Residual and Intercycle Ice Accretions on Airfoil Performance,” DOT/FAA/AR-02/
68, May 2002.
[7] U.S. Department of Transportation and Federal Aviation Administration, “Charac-
teristics of Runback Ice Accretions and Their Aerodynamic Effects,” DOT/FAA/
AR-07/16, April 2007.
[8] U.S. Department of Transportation and Federal Aviation Administration, “Aerody-
namic Simulation of Runback Ice Accretion,” DOT/FAA/AR-09/26, Dec. 2009.
[0] National Advisory Committee for Aeronautics, “Effect of Pneumatic De-Icers and
Ice Formation on Aerodynamics Characteristics of an Airfoil,” NACA TN 3564,
1956.
[10] NASA, “Deicing and Anti-icing Unite,” NASA Spinoff, https://spinoff.nasa.gov/
spinoff2002/ps_1.html [retrieved 25 Sept. 2020].
[11] Airbus, Flight Operation Support, “Getting Hands-On Experience with Aerody-
namic Deterioration,” Issue 2, Oct. 2001.
Chapter 7 Stall Testing

Chapter Objective
The stall speed is the limiting low speed of the flight envelope. The first image
that usually comes to mind when the words stall speeds are mentioned is that
of maximum lift due to flow separation. Although this is the ultimate goal to
reach to achieve the lowest operational speeds, certification requirements may
dictate a lower maximum lift coefficient if the stall characteristics at or near
the aerodynamic stall angle of attack are not acceptable. This chapter gives
an idea of the steps required to define the stall speeds from an FAA Part 25
(transport category airplane) and Part 23 (small airplane) point of view. It
introduces regulations, guidance material, testing sequence, data reduction,
data expansion, and what information to present to the flight crew. We
address risks related to stall testing and present some mitigations. One
should expect that stall testing will take a major portion of the flight test
time and will define the low speed performance of the plane.

Importance of a Stall Program on Airplane Certification

S
tall testing is an essential part of certification flight testing. The
minimum operational speeds of the airplane are defined by the stall
speed (often a stall speed ratio); certification flight testing for low
speed handling characteristics must be done at these speeds. For Part 23
(small airplane) and Part 25 (transport category airplane), the stall speed
must be determined by flight testing. The stall speeds therefore need to be
determined early in a flight test program for the following reasons (see Fig. 7.1):
• To define the minimum flight airspeeds and how they vary with weight,
altitude, and airplane configuration (stall speeds)

Test
Initial low Flap/slat
program
speed handling optimization Remainder of handling characteristics testing timeline

Stall Remainder of performance testing

First Initial stall
speed testing characteristics and AFM
flight
stall speed testing

Fig. 7.1 Flight test program timeline.

217
218 Operational Aircraft Performance and Flight Test Practices

• To demonstrate that handling qualities are adequate to allow a safe recovery

from the highest angle of attack attainable in normal flight (stall characteristics)
• To determine that there is adequate prestall warning (either aerodynamic
or artificial) to allow the pilot time to recover from any probable high angle
of attack condition without inadvertently stalling the airplane
• To perform the final selection of the flap and slat (if installed) deflection
angles to be used for the remainder of the program
• To determine if the airspeed and altitude indicating systems provide
adequate readings at the low speed, high angle of attack conditions
With this in mind, we take a look at what it takes to define the maximum
usable lift coefficient and the stall speed that can be used in a flight manual.

Defining Stall from a Certification Point of View

When one thinks about stall, the first thing that comes to minds is the
condition where the airflow separates from the wing and the airplane lift
coefficient CL reduces (see Fig. 6.11 in Chapter 6). That is the aerodynamic
stall, and it is the highest value of CL expected to be achieved. The
maximum lift coefficient CLmax that can be used by an airplane operationally
is often lower than this ideal aerodynamic maximum. That value is defined by
a set of regulations under which the airplane got certified; however, no matter
which certification agency oversees the certification of a given airplane, one
should expect that the stall is defined by an acceptable airplane behavior and
a high angle of attack. We will provide a glimpse of what is defined as accep-
table in this regulation-heavy section.

Testing for Repeatability: Stall Speed Part 25 Transport Category

The stall speed is the minimum speed that an aircraft can sustain in level
flight. This speed can vary by a few knots for a given configuration/weight
combination depending on how the aircraft gets there (i.e., entry rate, level
flight vs banked turn, load factor during the maneuver, thrust generated by
the engine/propellers, etc.).
It is therefore very important to define a specific testing methodology and
procedure to get repeatable results. It is also important to see how the stall
speed is defined for several certification agencies, so we offer this quick overview.
To frame our discussion on stall speed, we focus on the certification regu-
lation of Part 25 transport category airplanes and then discuss some differ-
ences with other sets of regulations. The stall speed defined by the U.S.
Federal Aviation Administration (FAA), the European Union Aviation
Safety Agency (EASA), and Transport Canada Civil Aviation (TCCA) for
Part 25 transport category airplanes is defined by paragraph 25.103 as follows:
§25.103 Stall speed. (FAA at amendment 25-121, 2007)
(a) The reference stall speed, VSR , is a calibrated airspeed defined by
the applicant. VSR may not be less than a 1-g stall speed. VSR is
 CHAPTER 7 Stall Testing 219

expressed as:

VCLmax
VSR  pffiffiffiffiffiffiffi
ffi
nzw

Where:
VCLMAX ¼ Calibrated airspeed obtained when the load
factor-corrected lift coefficient
 
nzw W
qS
is first a maximum during the maneuver prescribed
in paragraph (c) of this section. In addition, when
the maneuver is limited by a device that abruptly
pushes the nose down at a selected angle of attack
(e.g., a stick pusher), VCLMAX may not be less than
the speed existing at the instant the device operates;
nzw ¼ Load factor normal to the flight path at VCLMAX
W ¼ Airplane gross weight;
S ¼ Aerodynamic reference wing area; and
q ¼ Dynamic pressure.
(b) VCLMAX is determined with:
(1) Engines idling, or, if that resultant thrust causes an appreciable
decrease in stall speed, not more than zero thrust at the stall speed;
(2) Propeller pitch controls (if applicable) in the takeoff position;
(3) The airplane in other respects (such as flaps, landing gear, and ice
accretions) in the condition existing in the test or performance
standard in which VSR is being used;
(4) The weight used when VSR is being used as a factor to determine
compliance with a required performance standard;
(5) The center of gravity position that results in the highest value of
reference stall speed; and
(6) The airplane trimmed for straight flight at a speed selected by the
applicant, but not less than 1.13VSR and not greater than 1.3VSR .
(c) Starting from the stabilized trim condition, apply the longitudinal
control to decelerate the airplane so that the speed reduction does not
exceed one knot per second.
(d) In addition to the requirements of paragraph (a) of this section, when a
device that abruptly pushes the nose down at a selected angle of attack
(e.g., a stick pusher) is installed, the reference stall speed, VSR , may not be
less than 2 knots or 2 percent, whichever is greater, above the speed at
which the device operates.

Thus, the definition of stall speed for transport category airplanes, per
§25.103, goes much beyond the simple aerodynamic stall (1-g stall speed).
First, it allows the applicant to select a speed equal to or greater than the
1-g stall speed in paragraph 25.103(a). Also note how 25.103(b)(6) and (c)
described the “acceptable” maneuver to be flown to achieve the certification
stall speed. That same maneuver is illustrated in Fig. 7.2. On this same figure
220 Operational Aircraft Performance and Flight Test Practices

we included the traditional stall speed Vs that existed prior to Amendment

108 (2002) for the FAA; we will address this stall speed later in this
chapter. Note that the trim speeds are representative of minimum oper-
ational speeds for takeoff or approach/landing.
Even the rate at which the airplane slows down (1 kt/s) represents the
upper limit of an inadvertent and unrecognized slowdown for an operational
target speed. We will see how stall entry rate (slowdown rate) can impact the
stall speed.
Paragraph 23.103(b) further describes what is an acceptable configuration
to define the stall speed by test including achieving zero thrust. Flight testing
is not done with zero thrust, so acceptable means to correct the thrust to zero
are provided in guidance material or can be offered by the applicant for
acceptance by the certification authority. When an airplane is propelled by
propellers of variable pitch, the pitch setting required by 25.103(b)(2) for
the stall test is the takeoff pitch setting.
The regulation recognizes the impact of the center of gravity (CG)
location on the stall speed. With most airplanes, the CG position that will
satisfy 25.103(b)(5) is the maximum forward CG for the test weight. So
from a Part 25 point of view, the stall speed of the airplane is the one with
the most critical CG for a given weight, which means the airplane will essen-
tially have better capability (slower stall speed) for the majority of operations
because the CG typically falls somewhere between the most forward and
most aft certified values.
What is not directly specified by §25.103 but is implied is that for a stall
speed to be valid, it must meet the requirements of §25.201, Stall Demon-
stration; §25.203, Stall Characteristics; and §25.207, Stall Warning. These
sections require that the weight and CG envelope of the airplane be inves-
tigated, the stall test be done in level flight and in 30-deg bank turns, the
stall be done with power off (idle) and at a higher thrust, and the stall
testing be done at various entry rates up to 3 kt/s. This testing, which we
will briefly describe in the following section, may limit the maximum
angle of attack the airplane may use and meet all these requirements
(Fig. 7.3).

Deceleration at 1 kt/s
Aircraft trimmed with elevator input only Recovery
1.13 and 1.23 VSR
Speed

(1.2 to 1.4 Vs)

1-g stall Stall

identification
Time

Fig. 7.2 Maneuver to define the stall speed.

 CHAPTER 7 Stall Testing 221

1-g stall condition

Aft CG CLmax

CL Fwd CG CLmax
Possible
certification limit

Specific configuration
Flaps: UP, TO, or LDG
Gear: UP or DOWN

(deg)

Fig. 7.3 Certification CLmax .

Part 23 Small Airplane: Stall Speed

We now look at the stall speed as defined by the Part 23 small airplane
(less than 19,000 lb) category. These rules were significantly revised in
2017 in Amendment 64 for the FAA and Amendment 5 for EASA. We
first discuss the pre-Amendment-64 regulations, and then we will review
the new revision.
§23.49 Stalling speed. (FAA at Amendment 23-62, 2012)
(a) VS0 (maximum landing flap configuration) and VS1 are the stalling
speeds or the minimum steady flight speeds, in knots (CAS), at which
the airplane is controllable with—
(1) For reciprocating engine-powered airplanes, the engine(s) idling, the
throttle(s) closed or at not more than the power necessary for zero
thrust at a speed not more than 110 percent of the stalling speed;
(2) For turbine engine-powered airplanes, the propulsive thrust not
greater than zero at the stalling speed, or, if the resultant thrust has
no appreciable effect on the stalling speed, with engine(s) idling and
throttle(s) closed;
(3) The propeller(s) in the takeoff position;
(4) The airplane in the condition existing in the test, in which VS0 and
VS1 are being used;
(5) The center of gravity in the position that results in the highest value
of VS0 and VS1 ; and
(6) The weight used when VS0 or VS1 are being used as a factor to
determine compliance with a required performance standard.
(b) VS0 and VS1 must be determined by flight tests, using the procedure and
meeting the flight characteristics specified in Sec. 23.201.
(c) Except as provided in paragraph (d) of this section, VS0 at maximum
weight may not exceed 61 knots for-
(1) Single-engine airplanes; and
(2) Multiengine airplanes of 6,000 pounds or less maximum weight that
cannot meet the minimum rate of climb specified in Sec. 23.67(a)(1)
with the critical engine inoperative.
222 Operational Aircraft Performance and Flight Test Practices

(d) All single-engine airplanes, and those multi-engine airplanes of 6,000

pounds or less maximum weight with a VS0 of more than 61 knots
that do not meet the requirements of Sec. 23.67(a)(1), must comply with
Sec. 23.562(d).

We see that Part 23 defines the stall speed in a similar way as Part 25 but
reflects the nature of the smaller airplanes. It notes a difference in thrust cor-
rection between reciprocating engines and turbine engines in 23.49(a)(1) and
(2), for example. It also clearly states that the stall speed is dependent on
achieving good stall characteristics in 23.49(b) and that the airplane must
be controllable at stall in 23.49(a). Of course, one big differentiator is that
Part 23 allows single-engine airplanes whereas Part 25 does not, so it is
addressed in §23.49.
One notable difference in the definition of the stall speed between Part 23
(VS , either VS0 or VS1 ) and Part 25 (VSR ) is that Part 23 focuses on the
minimum steady flight speed, which does not necessarily account for the
1-g stall condition. Instead, as we will describe later, the stall speed VS
focuses on the pilot’s recognition that the airplane has stalled. The stall
speed model developed does not need to account for the load factor in the
stall. Part 25 had similar requirements pre-Amendment 108 (2002); it also
used VS and left it to the applicants to decide if they were going to
account for the load factor (adopt a 1-g stall speed).
Part 23 was substantially changed in Amendment 64 (2017) for the FAA
and Amendment 5 (2018) for EASA. From the FAA published final rule,
we get:
The FAA amends its airworthiness standards for normal, utility, acrobatic,
and commuter category airplanes by replacing current prescriptive design
requirements with performance-based airworthiness standards. These stan-
dards also replace the current weight and propulsion divisions in small air-
plane regulations with performance- and risk-based divisions for airplanes
with a maximum seating capacity of 19 passengers or less and a maximum
takeoff weight of 19,000 pounds or less.

These airworthiness standards are based on, and will maintain, the level of
safety of the current small airplane regulations, except for areas addressing
loss of control and icing, for which the safety level has been increased. The
FAA adopted additional airworthiness standards to address certification for
flight in icing conditions, enhanced stall characteristics, and minimum
control speed to prevent departure from controlled flight for multiengine air-
planes. This rulemaking is in response to the Congressional mandate set
forth in the Small Airplane Revitalization Act of 2013.
With this new revision of the requirements, the FAA and EASA moved
the requirements for stall speed from §23.49 to §23.2110 to reflect the
break in the regulation.
§23.2110 Stall speed
The applicant must determine the airplane stall speed or the minimum
steady flight speed for each flight configuration used in normal operations,
 CHAPTER 7 Stall Testing 223

including takeoff, climb, cruise, descent, approach, and landing. The stall
speed or minimum steady flight speed determination must account for the
most adverse conditions for each flight configuration with power set at—
(a) Idle or zero thrust for propulsion systems that are used primarily for
thrust; and
(b) A nominal thrust for propulsion systems that are used for thrust, flight
control, and/or high-lift systems.

The new regulation mandates that the applicant determine a safe minimum
steady flight speed for all configurations and only prescribes the thrust con-
figuration. Two options are presented. The first is the traditional setting for a
propulsion system that provides thrust (same as Part 25 or previous Part 23),
but also adds a thrust setting for propulsion systems that may be used to also
control flight or provide high lift [§23.2110(b)]. This reflects the multitude of
configurations appearing in the 2010s in Part 23 that include some vertical lift
from propellers (eVTOL).
Although the requirements do not prescribe how to obtain the stall speed,
the new Part 23 does require preflight test approval of the proposed means of
compliance from the certification authority. The authorities will accept the
“old” Part 23 approach or the use of industry standards like ASTM, or the
authority may even allow an applicant to define a procedure if they can
be convinced.
For stall speed testing, ASTM F3179, Standard Specification for Perform-
ance of Aircraft, revision 18, provides the following definition of the stall
speed:
Stalling Speed
5.1 VS0 and VS1 are the stalling speeds or the minimum steady flight
speeds in knots (KCAS) at which the aeroplane is controllable with:
5.1.1 The propulsive thrust not greater than zero at the stalling speed,
or, if the resultant thrust has no appreciable effect on the stalling
speed, with engine(s) at minimum flight thrust and throttle(s) closed
with:
5.1.1.1 The propeller(s) in the takeoff position;
5.1.1.2 The aeroplane in the configuration existing in the test, in
which VS0 and VS1 are being used;
5.1.1.3 The center of gravity in the position that results in the
highest value of VS0 and VS1 ;
5.1.1.4 The weight used when VS0 or VS1 are being used as a factor
to determine compliance with a required performance standard.
5.2 VS0 and VS1 shall be determined by flight tests using the procedure and
meeting the flight characteristics specified in the appropriate stall handling
characteristics testing.
Note that, contrary to the certification requirements of FAA/EASA/
TCCA, ASTM standards are not public domain documents; the applicant
needs a subscription to get it. We simply provide an extract of the ASTM
to show the evolution of the approach to certification testing for stall
speed. This information may not reflect the latest revision.
224 Operational Aircraft Performance and Flight Test Practices

What we observe with this revision of ASTM F3179 is that this accep-
table means of compliance is very close to what the requirements of
§23.49 used to be. This ASTM revision also points to the fact that the
stall speed determined in flight testing is valid only if the stall character-
istics of the airplane are acceptable; this is also similar to what was speci-
fied in §23.49.

Stall Characteristics
Before we delve into testing for stall speed and the creation of a perform-
ance stall speed model, it is worth taking some time to understand the level of
effort needed for a good stall program. As was stated previously, the stall
speeds that are determined in flight testing are valid only if the flight charac-
teristics at stall are acceptable. Our focus again will be on Part 25 as
a baseline.
FAA 14 CFR/EASA CS/TCCA CAR Part 25 provides the following rules.
(We start with the wings-level test condition and have underlined some key
passages for discussion):
§25.203 Stall Characteristics
(a) It must be possible to produce and to correct roll and yaw by unreversed
use of the aileron and rudder controls, up to the time the airplane is
stalled. No abnormal nose-up pitching may occur. The longitudinal
control force must be positive up to and throughout the stall. In addition,
it must be possible to promptly prevent stalling and to recover from a
stall by normal use of the controls.
(b) For level wing stalls, the roll occurring between the stall and the
completion of the recovery may not exceed approximately 20 degrees.

Figure 7.4 shows a stall test point with acceptable stall characteristics. Note
that the stick force required to slow down up to stall identification (stall
ID, point where the pilot terminates the stall maneuver) remained positive.
Also note that as the airplane reached 1-g stall and continued the maneuver
to stall ID, there was no “abnormal nose-up pitching,” and an elevator
increase was required to continue slowing down. Finally, when the elevator
was pushed (beyond stall ID), there was a prompt decrease in pitch and
angle of attack, showing good controllability in the stall with normal use
(push to lower nose) of the controls.
On some airplanes, such as the example shown in Fig. 7.5, as the stall is
approached, the control forces may start reducing while the pilot is trying to
maintain a 1-kt/s slowdown. This control force reduction can come from a
reduced nose-down pitching moment from the airplane, either because the
horizontal tail comes into the wake of the wing and/or engine nacelles
(change in downwash angle and/or dynamic pressure on the tail) or by a
reduced nose-down pitching moment from flow separation on the wing or
both. Swept wing aircraft often have wing tip stalls prior to wing root; as
 1-g Stall 1-g Stall
stall ID stall ID
KIAS (kt) Nzw (g)
150 1.4 Elevator (deg)
140 1.2 15
130 1.0

Nzw (g)
120 0.8
KIAS

10
(kt)

110 0.6 n
dow
100 0.4 5 Slow

Elevator
(deg)
90 0.2
80 0 0 t
Elevator (deg) Stick force (Ib) ou
sh
Pu
30 40 –5
30
20 20

Stick force
–10
Elevator

10
(deg)

(lb)
10 0 Stick force (lb)
–10 50
0 –20
–30 40

Pull force
–10 –40
30

Stick force
Pitch (deg) BoomAoA (deg) Slow down

(lb)
20
30
Boom AoA (deg)

20 10
Pitch (deg)

10 0
0 Push out

CHAPTER 7
–10 –10
10 12 14 16 18 20 22 24
–20
40 50 60 70 80 90 100 Boom AoA (deg)
Time (s)

Stall Testing
Fig. 7.4 Stall with acceptable pitch characteristics.

225
226 Operational Aircraft Performance and Flight Test Practices

the flow starts separating at the wing tip, it moves the wing center of pressure
CP forward, thus reducing the nose-down pitching moment.
For an airplane with reversible flight controls (where the aerodynamic
forces on the elevator are transmitted to the control column), the changes
in the pressure distribution over the horizontal tail and elevator may
impact the control force for the pilot (it may result in force gradient
change and possibly reversal), even if the elevator deflection required to con-
tinue to slow down is in the correct direction. This was the case for the air-
plane shown in Fig. 7.5 where the elevator deflection was still in the correct
direction (although reduced gradient), but there is a clear change in force gra-
dient slope just before 1-g stall. The requirement of §25.203(a) is that “the
longitudinal control force must be positive” (i.e., a pull force is required to
slow down) up to the stall (not the 1-g stall, the pilot-identified stall—we
will expand on this item later). The case shown in Fig. 7.5 is an acceptable
but marginal case. The change in the aerodynamic forces becomes less of a
problem with irreversible flight controls in which the artificial feel system
provides a control force proportional to the stick displacement.
Part 23 requirements are similar to those in Part 25, §25.203.
MIL-F-8785C, 3.4.2.1.2, adds “it is desired that no pitch-up tendencies
occur in unaccelerated or accelerated stalls.”
On the lateral-directional side, §25.203 states that the use of correct flight
control inputs is required (unreversed use), which means one must use ailer-
ons (lateral control) to pick up a dropping wing in the stall. This author had a
discussion with a test pilot once who said that when you near stall you should
use the rudder to roll the airplane. It is true that aileron deflection (down-
going aileron to pick up dropping wing) increases the risk of wing tip stall,
but the regulation implies that the pilot does not know that the airplane is
getting close to stall, and the normally expected behavior is to instinctively
use the ailerons. The applicant should therefore design the plane with this
aspect in mind to provide proper roll control margin.
Paragraph 25.203(b) requires that the airplane roll be controllable to
approximately 20-deg bank on either side of wings level. It is expected to

30
1-g stall
20
Push force Pull force

Stall ID
CG
Stick force (lb)

10

0 CP Sta
ll p
rog
–10 res
sio
n
–20

–30
10 12 14 16 18 20 22
Boom AoA (deg)

Fig. 7.5 Marginal longitudinal control characteristics in stall.

 CHAPTER 7 Stall Testing 227

25
Max per §25.203(b)

Max absolute roll angle (deg)

20
Company
FAA
15

10

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
MSR

Fig. 7.6 Typical test scatter in roll (well-behaved airplane).

have some roll (uneven stall pattern on the wing) during the testing up to stall
and during the recovery. There is the possibility of small wing oscillation that
would drive the local angle of attack to be uneven. The airplane may not be
built perfectly symmetric. And for airplanes with a larger stall angle, the pilot
may lose sight of the horizon during the maneuver. The data of Fig. 7.6 rep-
resent a typical scatter of a test campaign where the airplane had certifiable
roll characteristics. The data are plotted as maximum roll angle seen during a
stall vs the Mach number at stall, where the higher Mach number conditions
represent higher altitude stall conditions.
On some airplanes, the roll characteristics in stall may prove unaccepta-
ble. The airplane may oscillate too much, requiring exceptional piloting skills
to stay within the limits of the certification requirements. Other times, one
wing tip may stall first and sharply, resulting in a very rapid increase in roll
rate in one direction (wing drop) that is beyond the ability of the pilot to
recognize the roll and counter it (see Fig. 7.7).
If this unacceptable behavior is seen in flight test (i.e., late in the program),
the options left to the airplane designer, short of redesigning the wing, are
either to reduce the usable angle of attack range until the airplane has accep-
table stall characteristics or to try to “dress up” the wing with various airflow
control devices like stall strips, vortilons, stall fences, leading edge droop,
and so forth (see Fig. 7.8). Although these devices may have an acceptable
impact on stall characteristics and stall speed, they may have an adverse
impact on cruise performance.

Stall Speed (Vs1g ) vs Stall Identification (VSID )

“Why can’t the pilot stop the stall demonstration prior to the unaccep-
table characteristics if it is beyond 1-g stall?” This author was asked this
228 Operational Aircraft Performance and Flight Test Practices

KIAS (kt) AOS (deg)

140 30
130 25
AoA
120 20

AoA (deg)
KIAS (kt)

110 15
100 10
90 KIAS 5
80 0
70 –5
Roll (deg) Roll Rate (deg/s)

20
Roll rate (deg/s)

0
Roll (deg)

Max §25.203(b)
–20
Roll rate Roll angle
–40

–60
Aileron (deg)
15
Alleron (deg)

10
5
0
–5
0 5 10 15 20 25 30
Time (s)

Fig. 7.7 Unacceptable roll behavior in stall.

question by a CEO once during a flight test program of a small Part 23 jet
after the results of the stall testing were presented and it was shown that the
airplane had unacceptable rolling tendency after the 1-g stall point. This is
where we spent time explaining §23.201, Stall Demonstration, to the CEO.
There is a distinction between the time when the aircraft reaches the
1-g stall condition and the point at which the crew recognizes that the air-
craft has stalled. The 1-g stall is the point where level flight cannot be
maintained any longer (the wing cannot sustain the flight path). If the
“stall” is gentle, the flight crew may not notice the small deviation from
the flight path (small reduction in the load factor and start of a gentle
descent). In this situation, the flight crew could still inadvertently reduce
the airspeed.
Stall identification is the point at which the crew has a clear and distinct
indication that the airplane is stalled. If the 1-g stall is gentle enough, the stall
identification could be several knots slower. Certification requirements do
provide guidance to determine what is an acceptable stall identification.
For Part 25 airplanes, this is defined as part of §25.201:
§25.201 Stall Demonstration
(d) The airplane is considered stalled when the behavior of the airplane gives
the pilot a clear and distinctive indication of an acceptable nature that
the airplane is stalled. Acceptable indications of a stall, occurring either
individually or in combination, are—
 Embraer EMB-145
Leading edge droop
Learjet 60

Wing fences Vortilon

Button head screws

Boundary layer
control triangles Boundary layer
energizers Boundary layer
Wing fence
control devices

Hondajet HA420

CHAPTER 7
Stall strips

Stall Testing
Fig. 7.8 Wing dressing to address unacceptable stall characteristics.

229
230 Operational Aircraft Performance and Flight Test Practices

(1) A nose-down pitch that cannot be readily arrested;

(2) Buffeting, of a magnitude and severity that is a strong and effective
deterrent to further speed reduction; or
(3) The pitch control reaches the aft stop and no further increase in pitch
attitude occurs when the control is held full aft for a short time before
recovery is initiated.

The “nose-down pitch” is the point where the pitching motion cannot be
arrested by the application of nose-up elevator, not necessarily the first indi-
cation of a nose-down pitching motion (see Fig. 7.9). On this chart, you see a
first nose-down pitching motion as the airplane reaches the 1-g stall con-
dition (as seen by the distinct change in slope of the load factor, the
g-break); however, a short moment after this nose-down pitching motion is
initiated, the pilot starts applying more stick input (change in the slope of
the elevator deflection and the elevator control force). In doing so, the
pitch (and AoA) start increasing again. This stall was terminated when a
stick pusher activated, creating a nose-down pitching motion by moving
the elevator and providing a force feedback (stall identification) to the pilot.
The guidance material of FAA Advisory Circular (AC) 25-7D states that
(underline added by author):
During this testing, the angle-of-attack should be increased at least to the
point where the behavior of the airplane gives the pilot a clear and distinctive
indication through the inherent flight characteristics or the characteristics
resulting from the operation of a stall identification device (e.g., a stick
pusher) that the airplane is stalled. That device must provide an effective
deterrent to further speed reduction.

For the conditions where the control stick reaches the aft stop, it must be shown
that continued pressure on the stick will not result in unacceptable character-
istics. AC25-7D suggests the following as a sign of a good test condition:
The pitch control reaches the aft stop and is held full aft for two seconds, or
until the pitch attitude stops increasing, whichever occurs later. In the case of
turning flight stalls, recovery may be initiated once the pitch control reaches
the aft stop when accompanied by a rolling motion that is not immediately
controllable (provided the rolling motion complies with § 25.203(c)).

For the deterrent buffet, AC25-7D suggests:

The airplane demonstrates an unmistakable, inherent aerodynamic warning
of a magnitude and severity that is a strong and effective deterrent to further
speed reduction. This deterrent level of aerodynamic warning (i.e., buffet)
should be of a much greater magnitude than the initial buffet ordinarily
associated with stall warning. This deterrent level of aerodynamic warning
(i.e., buffet) should be of a much greater magnitude than the initial buffet
ordinarily associated with stall warning

Pusher-Defined Stall: Flight Test Example

A pusher is a device that abruptly pushes the control column forward to
reduce the pitch and AoA of the airplane to prevent any further speed
 Pusher activation
KIAS (kt) Nzw (g) 1-g stall Stall ID
140 1.4
130 1.2
120 g-break 1.0

Nzw
KIAS
(kt)

(g)
110 0.8
100 0.6
90 0.4
80 0.2
Pitch (deg) AoA (deg)
40
30
Pitch (deg)
AoA (deg)

20
10
0
–10
–20
Start of pitch down
Elevator (deg) Stick force (lb)
but easily countered by more elevator

Stick force
20 80
Elevator
(deg)

CHAPTER 7
(Ib)
10
40
0
–10 0
0 10 20 30 40 50
Time (s)

Stall Testing
Fig. 7.9 Nose-down pitching motion at natural stall followed by clear stall ID (pusher).

231
232 Operational Aircraft Performance and Flight Test Practices

reduction (i.e., it provides a clear indication of stall). Stick pushers are

installed on airplanes either when the natural stall characteristics are not cer-
tifiable (roll-off greater than allowed or pitch-up tendency at stall) or when a
risk of deep stall exists; it is installed less often as only a stall ID device.
Airplanes with rear fuselage–mounted engines and a T-tail configuration
(e.g., Douglas DC-9, Canadair CRJ, most business jets) are likely candidates
because the tail may be blanked by the wing and/or engine wake at high
AoA, thereby reducing its pitch control effectiveness.
The requirements for qualification of a stick pusher vary among certifica-
tion agencies. Transport Canada (TCCA) Advisory Circular (AC) 525-020,
Stall Compliance—Stick Pusher, specifies the following (underline added
by author):

Para 5.4 (a)(i) 1) For aeroplanes which incorporate a stick pusher. . .

the system for operating the device should be such that the system is auto-
matically armed, and will remain armed, in each configuration in which
operation of the system is necessary to show compliance with the stalling
requirements, . . .
Para 5.4 (b)(i) The operation of the device, in straight and turning flight
stalls should be such that the aeroplane with the system operating normally
complies with all relevant stalling requirements . . .
Para 5.4 (b)(iv) The characteristics of the stick-pusher should be such that
it is unlikely that a member of the flight crew will prevent or delay
its operation.
The required stick force, rate of application and stick travel will depend on
the aeroplane’s stall and stick force characteristics but a force in the region of
50 to 80 pounds applied virtually instantaneously has previously been
accepted as providing this characteristic.

The FAA provides the following for stick pushers in AC 25-7D, §42.1, Design
and Function of Artificial Stall Warning and Identification Systems (under-
line added by author):

. . .some airplanes require a stall identification device or system (e.g., stick

pusher,) to compensate for an inability to meet the stalling definitions of §
25.201 or the stall characteristics requirements of § 25.203.
42.1.3.2 Stall identification systems should be armed any time the airplane
is in flight.
42.1.3.2.1 The arming should take place automatically and may be pro-
vided by the same ground/air sensing system used for arming the stall
warning system. The stall identification system may be inhibited during
the takeoff rotation, but should become functional immediately after main
gear liftoff.
42.1.3.2.2 Stall identification systems may incorporate automatic disarm-
ing in flight regimes where the risk of stalling is extremely remote or
where their unwanted operation would pose a threat to continued safe
flight; examples of such inhibits would be high airspeed, and “g” cutouts
(typically 0.5 g), and while the pilot is following windshear recovery flight
director guidance.
42.1.3.2.3 A means to quickly deactivate the stall identification system
should be provided and be available to both pilots. It should be effective at
all times, and should be capable of preventing the system from making any
 CHAPTER 7 Stall Testing 233

input to the longitudinal control system. It should also be capable of cancel-

ing any input that has already been applied, from either normal operation or
from a failure condition.
42.1.6.2 The characteristics of stall identification systems, which by design
are intended to apply an abrupt nose-down control input (e.g., a stick
pusher), should make it unlikely that a flightcrew member will prevent or
delay its operation. The required stick force, rate of application, and stick
travel will depend on the airplane’s stall and stick force characteristics, but
a force of 50 to 80 lbs applied virtually instantaneously has previously been
accepted as providing this characteristic.
42.1.6.3 Normal operation of the stall identification system should not
result in the total normal acceleration of the airplane becoming negative.

For Part 23 airplanes, the FAA specifies in AC 23-8C, Flight Test Guide, the
following (underline added by author):
Airspeed Margins. The airspeed margin between unsatisfactory stall charac-
teristics and the minimum stick pusher actuation speed, for identical flight
conditions, should be evaluated. The following information is provided as
a guide. For airplanes with unsatisfactory, hazardous or unrecoverable aero-
dynamic stall characteristics, the minimum speed margin between aerody-
namic stall and minimum stick pusher systems actuation speed should not
be less than five knots. For other airplanes with known and less hazardous
aerodynamic stall characteristics, the speed margin may be reduced to not
less than two knots.

AC 23-8C adds that a “less hazardous aerodynamics stall characteristic”

would be one where the airplane may not meet the roll-off limit of 15 deg,
for example (thus the stick pusher is needed to show compliance), and the
more hazardous stall characteristics would result in an unsafe flight condition
(a roll angle beyond 60 deg, for example).

Pusher-Defined Stall: Prestall Pusher

The applicant can elect to prevent the airplane from reaching natural stall
when the natural stall characteristics are noncompliant. Figure 7.10 presents
the results of a flight test condition with a prestall activation of a stick pusher.
Note how the lift coefficient CL keeps increasing until the pusher activates
and rapidly moves the elevator in the opposite direction, creating a nose-
down pitching moment to initiate a recovery and minimize additional
speed decay. There was still a small speed decay beyond the activation of
the pusher before the airplane reacted to the nose-down motion and
started reaccelerating, during which time the airplane characteristics still
have to meet applicable requirements.
The stick pusher should stop pushing the nose down as soon as the angle
of attack is below a preset value (at least equal to the firing angle) to allow the
pilot to start pulling out of the nose-down condition and recover the airplane.
The pusher should not stay on long enough to push the airplane to a negative
load factor. In Fig. 7.10, the total pusher activation time was about 1.5 s, as
seen by the rapid release of the nose-down elevator at time 61 s.
234 Operational Aircraft Performance and Flight Test Practices

Pusher
KIAS (kt) CL Nz (g)
activation
200 2.5
190
CL
180 2.0
170
160 1.5
KIAS (kt)

Nz (g)
150

CL
140 1.0
NZ
130
KIAS
120 0.5
110
100 0.0

Elevator (deg) Pitch (deg) Boom AoA (deg)

20

15
AoA
10
Boom AoA (deg)
Elevator (deg)
Pitch (deg)

Pitch
5

–5
Elevator
–10

–15
30 40 50 60 70 80
Time (s)

Fig. 7.10 Prestall pusher flight test example.

When a pusher is used in this fashion, note there is no aerodynamic stall,

and the pusher becomes a “stop slowing down” device. Note the rapid change
in load factor during the activation of the stick pusher. Figure 7.11 shows

1
Company
Minimum load factor in the stall (g)

0.9
FAA
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
MSR

Fig. 7.11 Flight test example of minimum load factor in stall, pusher-equipped airplane.
 CHAPTER 7 Stall Testing 235

typical test scatter of stall testing with a stick pusher (mix of pre- and poststall
pusher activation). The rapid change could be an issue at higher altitude
where the stall angle of attack is typically much lower than the low altitude
condition. The guidance material points to not having negative load factor
as an acceptable characteristic and also the possibility of adding an inhibit
pusher function when the load factor is lower than a given value.
The crew should not resist the stick pusher any more than the initial reac-
tion force when the pusher fires and the pilot is holding the stick, because
going beyond the pusher activation angle of attack will result in unacceptable
handling characteristics. Not resisting may be difficult for the pilot to do
when close to the ground, such as during the approach to land, as seen in
the Colgan Dash-8 Q400 accident.

NTSB/AAR-10/01 Loss of Control on Approach, Colgan Air

DHC-8 2 400, N200WQ, Flight 3407, February 12, 2009: The airplane
was cleared for the runway 23 instrument landing system approach to
Buffalo Niagara International Airport. The weather consisted of light snow
and fog with wind of 15 kt. The deicing system had been turned on 11 min
after takeoff. Shortly before the crash, the pilots discussed significant ice
buildup on the aircraft’s wings and windshield.
Following the clearance for final approach, landing gear and flaps (5 deg)
were extended, and the airplane slowed to 145 kt. The captain then called for
the flaps to be increased to 15 deg with the airplane slowing to 135 kt. Six
seconds later, the aircraft’s stick shaker activated as the speed continued to
slow to 131 kt. The captain responded by abruptly pulling back on the
control column, followed by increasing thrust to 75% power, instead of low-
ering the nose and applying full power. That action pitched the nose up
even further, increasing the load factor. The stick pusher activated, but the
captain overrode the stick pusher and continued pulling back on the
control column. The aircraft pitched up 31 deg, then pitched down 25 deg,
then rolled left 46 deg and snapped back to the right at 105 deg, eventually
crashing and killing all on board.

Pusher-Defined Stall: Poststall Pusher

Airplanes that have good handling qualities in natural stall may not
require a stick pusher except maybe for stall identification and for deep
stall protection. This is often the case for airplanes with slats extended, for
example. But to take full advantage of the 1-g stall rule, the stick pusher
should fire at a speed at least 2% or 2 kt (whichever is larger) slower than
the natural stall speed to allow the applicant to use the latter speed as the air-
plane VSR per §25.103(d).
Figure 7.12 is an example of a poststall pusher application. In this case,
the 1-g stall condition is reached at about time 68 s. We then see a drop in
 236
Operational Aircraft Performance and Flight Test Practices
KCAS (kt) NZW (g) CL 1-g Pusher
stall activation
140 2.5

1.855302
130 1.584568 2.0
CL

120 1.5

Nzw (g)
KCAS (kt)
0.9296512

CL
0.7544255
NZW
110 1.0

KIAS
100 0.5
100.56
98.06
90 0.0

Elevator (deg) Pitch (deg) Boom AoA (deg)

30
18.4577 23.7377
11.67
20
AoA 10.30664
Boom AoA (deg)
Elevator (deg)
Pitch (deg)

10
Pitch –4.997829
0

Elevator
–10

–14.62796
–20
30 40 50 60 70 80 90
Time (s)

Fig. 7.12 Poststall pusher flight test example.

 CHAPTER 7 Stall Testing 237

the load factor (g-break) and a steepening of the elevator input (a sign of an
increasing nose-down pitching moment, good characteristics). We also note
that the buffeting (as measured by the Nzw load factor) is not significant, and
as the pilot keeps pulling on the control stick, the AoA keeps increasing. In
this case, the pusher fired about 5 deg of AoA later and a little over 2 kt
slower.
Here again, the stick pusher is a “stop slowing down” device meant to
provide a clear indication of stall (stall identification), but this time the man-
ufacturer can take full credit for the lower aerodynamic stall speed and the
resulting shorter field performance.

Aft Stick Elevator Stop–Defined Stall: Flight Test Example

Some airplanes have good stall characteristics and a well-tuned longitudi-
nal control system that will allow the plane to reach maximum travel at or
just beyond natural stall without creating unacceptable pitch-up tendencies.
Figure 7.13 shows a typical elevator travel limited stall speed flight test point.
Note how the lift coefficient increases until 1-g stall. As the airspeed is further
decreased, the lift coefficient and the load factor start decreasing and some
buffet appears. To further reduce the airspeed, the pilot must put in more ele-
vator. We observe that as the airplane pass the 1-g stall, the rate of input of
elevator per knot change in airspeed goes up. At one point, the elevator
becomes travel limited. FAA AC 25-7D provides the following guidance:
“The airplane is considered to be fully stalled when the pitch control
reaches the aft stop for at least 2 seconds, or the pitch attitude stops increas-
ing, whichever occurs later.”
Note that the exact point at which the stick reaches the aft stop and the start
of the 2 s may be difficult to define because the stick may also be shaken by a
stick shaker. The stall speed for non-load-factor-corrected condition VS is
the condition after the 2 s for Part 25 or the point where the stick reaches
the aft stop for Part 23 (see next section). For load-factor-corrected stall, the
stall speed then becomes defined by VCLmax (VS1g corrected to 1-g condition).

Stall Demonstration and Identification: FAA 14 CFR/EASA CS/TCCA CAR Part 23

FAA Part 23 pre-Amendment 64 and EASA CS pre-Amendment 5 had
similar, but slightly different, stall identification requirements under
§23.201, Wings Level Stall:
§23.201(b) The wings level stall characteristics must be demonstrated in
flight as follows. Starting from a speed at least 10 knots above the stall
speed, the elevator control must be pulled back so that the rate of speed
reduction will not exceed one knot per second until a stall is produced, as
shown by either:
(1) An uncontrollable downward pitching motion of the airplane;
(2) A downward pitching motion of the airplane that results from the
activation of a stall avoidance device (for example, stick pusher); or
(3) The control reaching the stop.
 238
1-g Push

Operational Aircraft Performance and Flight Test Practices

CLcor stall out
Nzw (g) KCAS (kt)
2.0 200
190
180
1.5
CLcor 170
160

KCAS (kt)
Nzw (g) Load factor
CLcor, 1.0 150

g-break 140

Light 130
0.5
KCAS buffet 120
110
0.0 100
Pitch (deg) Boom AoA (deg) Elevator (deg)
30
Recovery
20
Elevator AoA (deg)

Boom AoA
Boom AoA (deg)
Pitch (deg)

10
Pitch
0

Elevator
–10

–20
30 40 50 60 70 80 90 100
Time (s) Aft stop

Fig. 7.13 An aft stop-limited stall.

 CHAPTER 7 Stall Testing 239

Part 23 also specified slightly difference characteristics than those required

for Part 25, reflecting the nature of the smaller airplanes.
§23.201(c) Normal use of elevator control for recovery is allowed after the
downward pitching motion of paragraphs (b)(1) or (b)(2) of this section
has unmistakably been produced, of after the control has been held against
the stop for not less than the longer of two seconds or the time employed
in the minimum steady flight speed determination of Sec. 23.49.
§23.201(d) During the entry into and the recovery from the maneuver, it
must be possible to prevent more than 15 degrees of roll or yaw by the
normal use of controls except as provided for in paragraph (e) of
this section.
§23.201(e) For airplanes approved with a maximum operating altitude at or
above 25,000 feet during the entry into and the recovery from stalls
performed at or above 25,000 feet, it must be possible to prevent more
than 25 degrees of roll or yaw by the normal use of controls.
Although different, one can still see the attributes of a safe minimum flight
speed. Finally, the configuration required for the stall demonstration was
also slightly different than Part 25:
§23.201(f) Compliance with the requirements of this section must be shown
under the following conditions:
(1) Wing flaps: Retracted, fully extended, and each intermediate normal
operating position, as appropriate for the phase of flight.
(2) Landing gear: Retracted and extended as appropriate for the altitude.
(3) Cowl flaps: Appropriate to configuration.
(4) Spoilers/speedbrakes: Retracted and extended unless they have no
measurable effect at low speeds.
(5) Power:
(i) Power/Thrust off; and
(ii) For reciprocating engine powered airplanes: 75 percent of
maximum continuous power. However, if the power-to-weight
ratio at 75 percent of maximum continuous power results in
nose-high attitudes exceeding 30 degrees, the test may be carried
out with the power required for level flight in the landing
configuration at maximum landing weight and a speed of 1.4 VSO ,
except that the power may not be less than 50 percent of maximum
continuous power; or
(iii) For turbine engine powered airplanes: The maximum engine
thrust, except that it need not exceed the thrust necessary to
maintain level flight at 1.5 VS1 (where VS1 corresponds to the
stalling speed with flaps in the approach position, the landing gear
retracted, and maximum landing weight).
(6) Trim: At 1.5 VS1 or the minimum trim speed, whichever is higher.
(7) Propeller: Full increase r.p.m. position for the power off condition.
For FAA Amendment 64 and EASA Amendment 5 and above, Part 23 is now
less prescriptive; the authorities expect the applicant to select/define an
acceptable means of compliance (MOC) for stall demonstration. The focus
now is on selecting that MOC and getting the authorities to accept it prior
to the start of the stall test program.
One acceptable MOC for stall characteristics for the new Part 23 is ASTM
F3180, an extract of which is shown next. The new regulations and standards
now call for enhanced stall warning and airplane handling characteristics that
240 Operational Aircraft Performance and Flight Test Practices

demonstrate resistance to inadvertent departure from controlled flight. The

ASTM standard introduces the concept of a low speed flight characteristic
score SLSC that allows the applicant more flexibility on how they can protect
the airplane from unacceptable low speed characteristics.
The following is an extract from ASTM F3180, 4. Low-Speed
Characteristics:
4.1 Low-Speed Flight Characteristics Score—The applicant shall demon-
strate that the aeroplane has acceptable stall characteristics, stall
warning, and spinning characteristics, if applicable, by compliance with
performance criteria specified for stall characteristics and spinning,
as applicable, and by accumulating a number of “points” from stall
warning, departure characteristics, and safety-enhancing features.
4.1.1 The sum of the Stall Warning Score (SSW ), Departure Character-
istics Score (SDC;SE or SDC;ME , if required), and the safety-
enhancing features score (SSEF , if required) is called the
Low-Speed Flight Characteristics Score (SLSC ).
NOTE 1—The rationale for the development of the
Low-Speed Flight Characteristics Score is provided in Borer’s [1]
“Development of a New Departure Aversion Standard for Light Air-
craft.”, ref. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/
20170005881.pdf.
4.1.2 The minimum Low-Speed Flight Characteristics Score is dependent
on the certification level, number of engines, and whether or not the
aeroplane is approved for aerobatics, as summarized in Table 7.1.
We will discuss stall warning later in this chapter. For reference, the airplane
certification levels are defined in §23.2005 as follows:
§23.2005 Certification of normal category airplanes
(a) Certification in the normal category applies to airplanes with a
passenger-seating configuration of 19 or less and a maximum
certificated takeoff weight of 19,000 pounds or less.
(b) Airplane certification levels are:
(1) Level 1—for airplanes with a maximum seating configuration of 0
to 1 passengers.
(2) Level 2—for airplanes with a maximum seating configuration of 2 to
6 passengers.
(3) Level 3—for airplanes with a maximum seating configuration of 7
to 9 passengers.
(4) Level 4—for airplanes with a maximum seating configuration of 10
to 19 passengers.
(c) Airplane performance levels are:
(1) Low speed—for airplanes with a VNO and VMO  250 Knots
Calibrated Airspeed (KCAS) and a MMO  0.6.
(2) High speed—for airplanes with a VNO or VMO . 250 KCAS or a
MMO . 0.6.
(d) Airplanes not certified for aerobatics may be used to perform any
maneuver incident to normal flying, including—
(1) Stalls (except whip stalls); and
(2) Lazy eights, chandelles, and steep turns, in which the angle of bank is
not more than 60 degrees.
(e) Airplanes certified for aerobatics may be used to perform maneuvers
without limitations, other than those limitations established under
subpart G of this part.
 CHAPTER 7 Stall Testing 241

Table 7.1 Summary of Minimum Requirements for Low-Speed nigh Characteristics

Certification Level 1 2, 3, 4 1, 2 3, 4 1, 2, 3, 4
Engine(s) Single Single Multi Multi Any
Aerobatic? No No No No Yes
4.1 Minimum Low-Speed 150 200 150 50 50A
Flight Characteristics Score
(SLSC )
4.2 Stall Characteristics Pass all Pass all Pass all Pass Pass all
all
4.3 Stall Warning Score Min 50 Min 50 Min 50 Min 50 Min 50
(SSW ) Max 100 Max 100 Max 100
4.4 Departure Characteristics Min 50B Min 50B N/A N/A N/AA
Score – Single Engine Max 100 Max 100
(SDC;SE )
4.5 Departure Characteristics N/A N/A Min 50 N/A N/A
Score – Multiengine Max 100
(SDC;ME )
4.6 Spinning N/A N/A N/A N/A Pass allA
4.7 Safety-Enhancing A/R to A/R to A/R to N/A N/AA
Features Score (SSEF ) meet meet meet
SLSC SLSC SLSC
Note: I—N/A: Not Applicable: A/R: As Required
A If spinning is requested for fewer than all possible conditions in 4.2.1 for single-engine aeroplanes, the
scores and requirements are pro-rated based on the number of conditions approved for spinning vs. not
approved for spinning, as appropriate to Certification Level and Engines.
BS
DC;SC may be less than 50 as outlined in 4.4.2.6(3), which requires particular equipment from 4.7.3
be installed.

Stall Identification: MIL-F-8785C

We open a small door to another acceptable means of compliance for
military/government airplanes. That MOC is MIL-F-8785C, which provides
the following guidance for the definition of the stall speed, including its
identification:
3.4.2.1. Stalls. The stall is defined in terms of airspeed and angle of attack in
6.2.2 and 6.2.5 respectively. It usually is a phenomenon caused by airflow sep-
aration induced by high angle of attack, but it may instead be determined by
some limit on usable angle of attack. The stall requirements apply for all Air-
plane Normal States in straight unaccelerated flight and in turns and pullups
with attainable normal accelerations up to nL . Specifically, the Airplane
Normal States associated with the configurations, throttle settings and
trim settings of 6.2.2 shall be investigated; also, the requirements apply to
Airplane Failure States that affect stall characteristics.
...
6.2.2 Speeds:
VS – The stall speed (equivalent airspeed), at 1 g normal to the flight path,
defined as the highest of:
242 Operational Aircraft Performance and Flight Test Practices

(a) speed for steady straight flight at CLmax , the first local maximum of the
curve of lift coefficient (L/qS) vs. angle of attack which occurs as CL is
increased from zero
(b) speed at which uncommanded pitching, rolling or yawing occurs
(3.4.2.1.2)
(c) speed at which intolerable buffet or structural vibration is
encountered
...

6.2.5:

aS – The stall angle of attack at constant speed for the configuration, weight,
center of gravity position and external-store combination associated with a
given airplane normal state; defined as the lowest of the following:
(d) Angle of attack for the highest steady load factor, normal to the flight
path, that can be attained at a given speed or Mach number
(e) Angle of attack, for a given speed or Mach number, at which
uncommanded pitching, rolling of yawing occurs (3.4.2.1.2)
(f) Angle of attack, for a given speed or Mach number, at which
intolerable buffeting is encountered.

Note how much closer to the edge the stall condition (intolerable rather than
deterrent buffet, exact angle at which an uncommanded pitching motion
occurs, etc.) is defined under MIL-STD. This is partly a reflection of the
greater training required by the crew of these airplanes and partly due to
the acceptance of higher risk. In the end, this is just another way to define
acceptable stall handling characteristics.

Stall Identification: Fly-by-Wire (FBW) Airplanes with Envelope Protection

The Airbus A320 introduced FBW to Part 25 airplanes in the late 1980s,
and it has progressively been used by more and more airplane manufacturers.
Contrary to conventional controls where the pilot has direct control of a
flight control, with FBW the pilot commands a computer to change flight
condition (pitch up/down, roll, etc.), and the computer decides which
flight controls (elevator, ailerons, rudder, etc.) need to be moved to match
the commanded input. FBW has provided new low speed protection capa-
bility to manufacturers in that, through control laws, the airplane can be pre-
vented from exceeding a maximum angle of attack in operation (a for high
alpha protection, Fig. 7.14).
At the time of writing of this chapter, Part 25 had not yet formally
adopted regulations to define what this high alpha protection capability
needed to be. Instead, each manufacturer has defined acceptable means of
compliance with their respective certification authorities to certify the exist-
ing products using FBW. In 2017, the Flight Test Harmonization Working
Group (FTHWG) brought forward some material that is in general agree-
ment with the participating members. The following quotes some of the defi-
nitions it is proposing and new regulations to allow future certification of the
FBW low speed protection.
 CHAPTER 7 Stall Testing 243

CL

High alpha protection

Aero stall
Reference stall


 for Vmin1g  for VSR

Fig. 7.14 FBW low speed envelope protection.

The proposed update for Part 25 to account for the new low speed pro-
tection keeps the definition of VSR from §25.103 but will add a step in the
execution of the test for VCLmax :
§25.103(b)(7) If installed, the High Angle-of-Attack Limiting Function
(HALF) disabled or adjusted, at the option of the applicant, to allow reaching
the angle of attack corresponding to VSR .
So the applicant still needs to define and demonstrate a safe maximum angle
of attack that is no greater than the angle of attack for aerodynamic stall (see
Fig. 7.14 earlier in the chapter). The working group also is proposing to intro-
duce new regulations §25.202 and §25.204.
§25.202 Handling demonstration for high angle-of-attack limiting functions
is aimed at offering an alternative to §25.201 stall demonstration where the
focus is now to show that the HALF will protect the airplane from stall.
The requirements also provide steps to show that if such protection is failed/
OFF, that sufficient warning is provided to the crew to allow recovery under
safe characteristics.
§25.204 Flight characteristics for high angle-of-attack limiting functions
states:
(a) Applicability: If a High Angle-of-Attack Limiting Function is installed
and compliance is being shown to §25.202 in lieu of §25.201, the high
angle-of-attack flight characteristics during the handling demonstrations
required by §25.202 must meet the requirements of paragraphs (b)
through (f) [author: of §25.204] in lieu of §25.203.

It therefore defines the new acceptable high angle-of-attack characteristics of

the airplane with HALF operating. Paragraph (b) will state:

(b) Throughout maneuvers with a deceleration of not more than 1 knot per
second, both in straight flight and in 308 banked turns, and with the High
Angle-of-Attack Limiting Function operating normally, the airplane’s
characteristics must be as follows:
244 Operational Aircraft Performance and Flight Test Practices

(1) There must be no abnormal nose-up pitching;

(2) There must be no uncommanded nose-down pitching indicative of
stall. Reasonable attitude changes associated with stabilizing the
angle-of-attack at the AOA-limit as the longitudinal control
reaches the stop are acceptable;
(3) There must be no uncommanded lateral or directional motion
indicative of stall, and the airplane must exhibit good lateral and
directional control by conventional use of the controls throughout
the maneuver; and
(4) The airplane must not exhibit buffeting of a magnitude and severity
that would act as a deterrent from completing the maneuvers.

So overall, the new regulations will provide another means of defining an

acceptable high angle-of-attack capability and minimum speeds of the air-
plane that will then be used to define operational speeds. It is still a prescrip-
tive approach, but one accepted by industry and authorities.

Bureau d’Enquêtes et d’Analyses (BEA) 20081127-0 FBW Slow-flight low

altitude accident-Airbus A320–232 D-AXLA, 27 Nov. 2008: Flight
GXL888T from Perpignan–Rivesaltes aerodrome was undertaken in the
context of the end of a leasing agreement, before the return of D-AXLA to
its owner. The programme of planned checks could not be performed in
general air traffic, so the flight was shortened. In level flight at FL320, angle
of attack sensors 1 and 2 stopped moving and their positions did not
change until the end of the flight [Author’s note: The FBW HALF uses
angle-of-attack information to provide low speed envelope protection.] After
about an hour of flight, the aeroplane returned to the departure aerodrome
airspace and the crew was cleared to carry out an ILS procedure to runway
33, followed by a go around and a departure towards Frankfurt/Main
(Germany). Shortly before overflying the initial approach fix, the crew
carried out the check on the angle of attack protections in normal law.
As the A320 reached a nose-up pitch of 18.6 deg with the speed dropping
to 99 kt, the stall warning sounded. The crew moved the thrust levers immedi-
ately to “takeoff/go-around” power. Flight-recorder data show that the aircraft
rolled left, then 50 deg right, as the captain made lateral and longitudinal side-
stick inputs to counter. Its pitch reduced to 11 deg, and the A320’s pitch and
roll flight-control laws passed from “normal” to “direct”—preventing the auto-
matic systems from offering protection. The crew eventually lost control of the
airplane and crashed at sea.

Stall in Icing
FAA Part 25 Amendment 25-121 (2007) introduced new airworthiness
standards to evaluate the performance and handling characteristics of trans-
port category airplanes in icing conditions. The action sought to improve the
level of safety for new airplane designs when operating in icing conditions
and harmonizes the U.S. and European airworthiness standards for flight
in icing conditions. These actions were driven by several National
 CHAPTER 7 Stall Testing 245

Transportation Safety Board (NTSB) recommendations following a series of

airplane accidents in icing, including:
• NTSB Safety Recommendation A-91-0872 recommended requiring flight
tests where ice is accumulated in those cruise and approach flap
configurations in which extensive exposure to icing conditions can be
expected and requiring subsequent changes in configuration to include
landing flaps. This safety recommendation resulted from an accident that
was attributed to tailplane stall due to ice contamination (Jetstream 31).
• NTSB Safety Recommendation A-96-056 recommended revising the icing
certification testing regulation to ensure that airplanes are properly tested
for all conditions in which they are authorized to operate or are otherwise
shown to be capable of safe flight into such conditions. Additionally, if safe
operations cannot be demonstrated by the manufacturer, operational
limitations should be imposed to prohibit flight in such conditions, and
flight crews should be provided with the means to positively determine
when they are in icing conditions that exceed the limits for aircraft
certification (ATR 72-212).
• NTSB Safety Recommendation A-98-094 recommended that
manufacturers of all turbine-engine–driven airplanes (including the
EMB-120) provide minimum maneuvering airspeed information for all
airplane configurations, phases, and conditions of flight (icing and nonicing
conditions). Also, the NTSB recommended that minimum airspeeds
should take into consideration the effects of various types, amounts, and
locations of ice accumulations, including thin amounts of very rough ice,
ice accumulated in supercooled large droplet icing conditions, and
tailplane icing.
• NTSB Safety Recommendation A-98-096 is also a result of the same
accident discussed under Safety Recommendation A-98-094. The NTSB
recommended the FAA require, during type certification, that
manufacturers and operators of all transport category airplanes certificated
to operate in icing conditions install stall warning/protection systems that
provide a cockpit warning (aural warning and/or stick shaker) before the
onset of stall when the airplane is operating in icing conditions.
With these updates, and the ice shapes defined in Chapter 6, most of the
stall characteristics testing must now be repeated to show the airplane has
acceptable handling to the maximum angle of attack available to the operator
(stall ID or HALF). The maximum AoA in icing can be reduced once icing
conditions are detected, as long as it is done automatically (through a reliable
primary ice detector or by the activation of an ice protection system).
It is not our intention to cover this additional material in this chapter; we
simply want to emphasize again that no stall speed can be valid if the
maximum usable angle of attack does not have acceptable handling charac-
teristics. Those characteristics need validation prior to performing stall speed
testing, or there is a risk (cost and schedule) of repeating the testing.
246 Operational Aircraft Performance and Flight Test Practices

Stall Warning ≥5%

25.207(c)
Airplanes that have FBW envelope
protection for high angle of attack
≥2%
(HALF) are supposed to prevent the If pusher ≥3%
pilot from achieving an angle of installed 25.207(d)
attack that would have unacceptable 25.103(d)
(uncertifiable) handling character-
istics, as demonstrated by flight
testing. For airplanes in which HALF VStall ID VSW
VSR
failed and for non-HALF-protected 25.201 25.103(a) 25.207
airplanes, the crew should be provided
with a warning prior to reaching the
critical AoA (certification stall) that Fig. 7.15 Stall warning margins.
will give them enough time to react
and counter the increasing angle-of-attack condition. Section 25.207 provides
the minimum certification requirements margin (critical passages underlined
by the author). Figure 7.15 provides a visual depiction of this stall warning
margin. It is important to note that the regulation clearly states that the stall
warning provided to the crew must be the same (stick shaker or buffeting)
in icing and in nonicing. One cannot change the warning from a stick
shaker to a buffeting to show compliance because one should not expect the
crew to pick up the difference that is weather dependent.
§25.207 Stall warning

(a) Stall warning with sufficient margin to prevent inadvertent stalling with
the flaps and landing gear in any normal position must be clear and
distinctive to the pilot in straight and turning flight.
(b) The warning must be furnished either through the inherent aerodynamic
qualities of the airplane or by a device that will give clearly
distinguishable indications under expected conditions of flight.
However, a visual stall warning device that requires the attention of the
crew within the cockpit is not acceptable by itself. If a warning device is
used, it must provide a warning in each of the airplane configurations
prescribed in paragraph (a) of this section at the speed prescribed in
paragraphs (c) and (d) of this section. Except for the stall warning
prescribed in paragraph (h)(3)(ii) of this section, the stall warning for
flight in icing conditions must be provided by the same means as the stall
warning for flight in non-icing conditions.
(c) When the speed is reduced at rates not exceeding one knot per second,
stall warning must begin, in each normal configuration, at a speed, VSW,
exceeding the speed at which the stall is identified in accordance with
Sec. 25.201(d) by not less than five knots or five percent CAS, whichever
is greater. Once initiated, stall warning must continue until the angle of
attack is reduced to approximately that at which stall warning began.
(d) In addition to the requirement of paragraph (c) of this section, when the
speed is reduced at rates not exceeding one knot per second, in straight
flight with engines idling and at the center-of- gravity position specified in
Sec. 25.103(b)(5), VSW, in each normal configuration, must exceed VSR by
not less than three knots or three percent CAS, whichever is greater.
 CHAPTER 7 Stall Testing 247

Figure 7.16 shows a flight test example of stall warning verification. To

ensure that one aims for the smallest margins, which then provides the
smallest possible takeoff and landing speeds, one should record the
moment the shaker and the pusher (or other devices) are active. This
helps with data reduction and in showing compliance because crew notes
(hand-recorded information) under these test conditions may not
be optimal.
Observation of Fig. 7.16 shows that the speed margin between stall
warning and VCLmax was a little more than 3 kt (once corrected to 1 g,
VSW ¼ 103.1 KCAS and Vs1g ¼ 99.8 KCAS), so VS1g could be used for
VSR . We also note that there is more than 5 kt between Vpush (VSID )
and stall warning and more than 2 kt between pusher (VSID ) and VSR .
Note that one does not correct the pusher speed by the load factor at
the time of the pusher activation, especially if the pusher is used as a post-
stall ID. This author had several discussions with the FAA on the subject,
and the reason is clear. Section 25.207 requires that sufficient warning be
provided to the crew of an impending stall. The requirement states that,
under a 1 kt/s slowdown, there must be at least 5 kt or 5% between VSW
and VSID (Vpush ); speeds were never load factor corrected and correspond
to an equivalent of at least 5 s of warning to the crew. In the example
from Fig. 7.16, there is almost 12 s warning; it meets the intent of
§25.207. The difficulty comes when comparing VSW to VSR or VSR to
Vpush where VSR is load factor corrected. If one were to correct the
Vpush of Fig. 7.12, where the airplane is clearly stalled (significant
reduction in the load factor), then one would compute a speed of 102.8
KCAS, only 0.3 kt slower than the 1-g VSW , yet there is clearly sufficient
warning time. It is therefore this author’s approach that one first compares
the actual margins between as-tested VSW and VCLmax and Vpush under a
1-kt/s deceleration maneuver. The minimum margins must be 3 kt (or 3%)
between the first two, 2 kt (or 2%) between the following two, and 5 kt
between VSW and VSID (Vpush ). Then VSR (not less than VS1g ) is com-
puted and compared to the load-factor-corrected VSW , and the
minimum margin of 3 kt (or 3%) must remain. The analysis is simpler
for prestall pusher where the load factor is essentially one all the way
to pusher activation.
The requirements of §25.207 are minimums. To meet these minimums
for all test conditions, one should expect that the average margin will be
higher. Figure 7.17 shows an example of test scatter.
The stall warning requirements of Part 23, pre-Amendment 64, are
similar and listed under §23.207 (underline added by author):

(c) During the stall tests required by §23.201(b) and §23.203(a)(1), the stall
warning must begin at a speed exceeding the stalling speed by a
margin of not less than 5 knots and must continue until the stall occurs.
(d) When following procedures furnished in accordance with §23.1585,
the stall warning must not occur during a takeoff with all engines
 248
Operational Aircraft Performance and Flight Test Practices
KCAS (kt) Nzw (g)

110 1.6
1.4
105 0.9151342 1.2
0.9499419 0.831414

KCAS (kt)

Nzw (g)
1.0
100
0.8
100.5582
0.6
95
95.75 0.4
93.69
90 0.2
Pitch (deg) AoA (deg)
30
16.44048 21.11987
20
26.32106
Pitch (deg)
AoA (deg)

10
13.36 12.96
0 11.1101

–10

–20
Shaker 2 Pusher 2
1=ACTIVE 1=ACTIVE
1.2
1 1 1
1.0
Pusher 2 (1=ON)
Shaker 2 (1=ON)

0.8
0.6
0.4
0 0
0.2
0.0 r ax she
r
ake Lm Pu 1805 35
10 15 20 Sh 0097 25 VC 9329 40
.7 .0 .4
22 Time (s) 30 34

Fig. 7.16 Stall warning margin verification.

 1.13 7 8
1.12
1.11 6 7
1.1
6
1.09 5

VSR-Vpush (kts)

VSW-VSR (kts)
VSW/Vstall ID

1.08 5
4
1.07
4
1.06 5% margin
3
1.05 3
Minimum §25.207(c) Minimum §25.207(d)
1.04 2
1.03 Minimum §25.103(d) 2
1.02 1
1
1.01
1 0 0
0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.22 0.23 0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.22 0.23 0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.22 0.23
MSR MSR MSR

Fig. 7.17 Stall test results compared to the minimum margins of §25.207 and §25.103(d).

CHAPTER 7
Stall Testing
249
250 Operational Aircraft Performance and Flight Test Practices

operating, a takeoff continued with one engine inoperative, or during an

approach to landing.
(e) During the stall tests required by §23.203(a)(2), the stall warning must
begin sufficiently in advance of the stall for the stall to be averted by pilot
action taken after the stall warning first occurs.

Part 23 at FAA Amendment 23-64, as discussed previously, introduced a

point system to allow the applicant to decide which type of warning is best
for the airplane configuration, with the focus being on stall prevention. It
also asks that the stall warning in nonicing be the same as that in icing,
per the latest Part 25 requirements.
We also note that for the requirements of MIL-F-8785C, there was not
only a minimum stall margin requirement, but also a specification that a
maximum margin must be respected:
1.4.2.1.1.1 Warning speed for stalls at 1 g normal to the flight path. Warning
onset for stalls at 1 g normal to the flight path shall occur between the fol-
lowing limits when the stall is approached gradually:

Flight phase Minimum speed for onset Maximum speed for onset
Approach Higher of 1.05VS or VS þ 5 kt Higher of 1.10VS or VS þ 10 kt
All other Higher of 1.05VS or VS þ 5 kt Higher of 1.15VS or VS þ 15 kt

Stall Speed Testing

Now that we have identified what an acceptable stall is, we proceed to
planning and executing a stall program with, for this book, a focus on stall
speed testing. After the maximum usable angle of attack has been defined
during the stall characteristic testing, the applicant then moves on to certifi-
cation stall speed testing to produce a stall model for the airplane flight
manual (AFM). The model must define a maximum usable lift coefficient
CLmax , either corrected for 1-g conditions [Eq. (7.1)] or not [Eq. (7.2)],
depending on the certification basis of the airplane.
nW
CLmax ¼ ð7:1Þ
1
r sV 2 S
2 SL
W
CLmax ¼ ð7:2Þ
1
r sV 2 S
2 SL

VS vs VCLmax : From a Lift Coefficient Point of View

Equations (7.1) and (7.2) both allow the user to create a maximum lift
coefficient model for certification—one accounts for the load factor during
the stall maneuver and the other does not. Figure 7.18 shows the resulting
 Nzw (g) CLcor CL KCAS (kt)
2.0 200
190
180
1.5
170
160

KCAS (kt)
CLcor, CL
Nzw (g)
1.0 150
140
130
0.5
VCLmax 120
110
Vs
0.0 100
Pitch (deg) Boom AoA (deg) Elevator (deg)
30

20
Boom AoA (deg)
Elevator (deg)
Pitch (deg)

10

CHAPTER 7
0

–10

Stall Testing
–20
30 40 50 60 70 80 90 100
Time (s)

Fig. 7.18 Lift coefficient with and without load factor correction.

251
252 Operational Aircraft Performance and Flight Test Practices

lift coefficient (CLcor being the load factor–corrected lift coefficient) for the
same maneuver.
From Fig. 7.18 we note the following:
• The aircraft was initially trimmed at 140 kt.
• Then the airplane decelerated at about 1 kt/s.
• Conditions for VCLmax were reached at about 115 kt and a NZW  1.0 g.
• Conditions for VS were reached at about 109 kt.
Note that in both cases (it was the same maneuver, after all), the pilot
continued the maneuver until it was perceived that the airplane was stalled
per the requirements of §25.201. From a pilot’s perspective, the loss of lift
at the natural stall may not be noticeable, except for the possible presence
of light buffeting.
Because the load factor at the point of reaching VCLmax (the first
maximum, at the g-break) is about 1.0, the 1-g stall speed Vs1g is also 115 kt.
We get a stall speed ratio (Vs /Vs1g ) of 0.948. At first glance, using the
uncorrected lift coefficient method seems to provide a distinct advantage in
certification stall speed compared to the load factor corrected method.
Under JAR 25.103(b)(2), prior to change 15 (mid-1990s), the Joint Avia-
tion Authority (JAA, the European agency monitoring airplane certification
prior to the establishment of EASA in 2002) had identified an additional
requirement that the stall speed VS should not be smaller than 94% of
Vs1g . The JAA had identified that modern airplanes could achieve
minimum flight speed of the order of 0.95 to as low as 0.92 Vs1g (some man-
ufacturers had very good stall test pilots that were able to extract the
maximum from the airplane) and had therefore imposed an additional
requirement over those of FAA 14CFR. The JAR regulations included the
1-g stall rule starting at Part 25, Change 15.
This meant that, although the regulations were not mandating the use of
1-g stall rules, they effectively asked the manufacturer to take no more
credit of the non-load-factor-corrected method than 94% of the 1-g stall. To
provide some relief on operational speed impact, recognizing that history
had shown that the stall speed ratio defined takeoff and landing speeds had
provided a good safety record, the speeds were adjusted to the following:

Takeoff safety speed V2 = 1:2 VS = ð0:94  1:2ÞVSR = 1:13 VSR

Landing speed VREF = 1:3 VS = ð0:94  1:3ÞVSR = 1:23 VSR

With the addition of this 1-g stall verification requirement and the adjust-
ment of the operational speeds plus the eventual incorporation of a formal
1-g stall rule in the regulation (EASA and FAA), most Part 25 airplanes
have now moved to the 1-g stall analysis. We will focus on this approach
for the rest of the chapter. Just remember that some airplanes certified
prior to 2000 for Part 25 and a lot of airplanes still being certified under
 CHAPTER 7 Stall Testing 253

Part 23 did not/do not use the 1-g stall regulation and that their stall speed is
the minimum steady flight speed.

Test Scope, Preparation, and Instrumentation

The stall speed of every airplane configuration Note: In 2004, the Sino
(combination of flap settings, gear up/down, Swearingen SJ30-2 became
spoiler positions, altitude, weight, and CG) must the first Part 23 aircraft to
adopt the Part 25 1-g stall
be determined. From a certification point of view,
rule as a means of
only the stall speed for the most adverse CG is pub- compliance. The 1-g stall
lished [§25.103(b)(5)]. The greater the number of rule was also adopted by the
flap settings, the greater the scope of testing that HondaJet team in 2010. Both
efforts were led by
must be expected. And because performance engin-
this author.
eers like data, each test condition for stall speed is
typically repeated at least once (two stalls per test
condition) to help provide a better average.
The guidance material of FAA AC 25-7D specifies that “a sufficient
number of stalls (normally four to eight) should be accomplished at each
critical combination of weight, altitude, CG, and external configuration.
The intent is to obtain enough data to determine the stall speed at an
entry rate not exceeding 1.0 knot/second.”

Test Preparation
Stall speed testing is done to collect data for the AFM; thus, careful prep-
aration is required to document the test, ensure the airplane configuration is
the intended end configuration, and make sure the proper data are collected
to show compliance. The requirements of Part 21, Type Certification, specify
the following (underline added by author):
§21.33 Inspection and tests
(a) Each applicant must allow the FAA to make any inspection and any flight
and ground test necessary to determine compliance with the applicable
requirements of this subchapter. However, unless otherwise authorized
by the FAA—
(1) No aircraft, aircraft engine, propeller, or part thereof may be
presented to the FAA for test unless compliance with paragraphs
(b)(2) through (b)(4) of this section has been shown for that aircraft,
aircraft engine, propeller, or part thereof; and
(2) No change may be made to an aircraft, aircraft engine, propeller, or
part thereof between the time that compliance with paragraphs (b)(2)
through (b)(4) of this section is shown for that aircraft, aircraft
engine, propeller, or part thereof and the time that it is presented to
the FAA for test.
(b) Each applicant must make all inspections and tests necessary to
determine—
(1) Compliance with the applicable airworthiness, aircraft noise, fuel
venting, and exhaust emission requirements;
(2) That materials and products conform to the specifications in the type
design;
254 Operational Aircraft Performance and Flight Test Practices

(3) That parts of the products conform to the drawings in the type
design; and
(4) That the manufacturing processes, construction and assembly
conform to those specified in the type design.

§21.35 Flight tests

(a) Each applicant for an aircraft type certificate (other than under §§21.24
through 21.29) must make the tests listed in paragraph (b) of this section.
Before making the tests the applicant must show—
(1) Compliance with the applicable structural requirements of this
subchapter;
(2) Completion of necessary ground inspections and tests;
(3) That the aircraft conforms with the type design; and
(4) That the FAA received a flight test report from the applicant (signed,
in the case of aircraft to be certificated under Part 25 [New] of this
chapter, by the applicant’s test pilot) containing the results of his tests.
(b) Upon showing compliance with paragraph (a) of this section, the
applicant must make all flight tests that the FAA finds necessary—
(1) To determine compliance with the applicable requirements of this
subchapter; and
(2) For aircraft to be certificated under this subchapter, except gliders
and low-speed, certification level 1 or 2 airplanes, as defined in part
23 of this chapter, to determine whether there is reasonable
assurance that the aircraft, its components, and its equipment are
reliable and function properly.
(c) Each applicant must show for each flight test (except in a glider or a
manned free balloon) that adequate provision is made for the flight
test crew for emergency egress and the use of parachutes.

The last subparagraph is important because it requires the manufacturer to

explain why it selected the level of protection it offers to the crews flying
the test airplane. Some manufacturers elect to offer emergency egress and
airplane stall parachute.
Weight is a major variable in the computation of the stall lift coefficient,
so one should plan to have a pre- and postflight weight of the plane. The
proper configuration should be selected (ballast, experimental equipment,
etc.) and kept in very tight control. FAA §25.21 Proof of compliance specifies
the requirements:
§25.21 Proof of compliance

(a) Each requirement of this subpart must be met at each appropriate

combination of weight and center of gravity within the range of loading
conditions for which certification is requested. This must be shown—
(1) By tests upon an airplane of the type for which certification is
requested, or by calculations based on, and equal in accuracy to, the
results of testing; and
(2) By systematic investigation of each probable combination of weight
and center of gravity, if compliance cannot be reasonably inferred
from combinations investigated.
...
(b) Parameters critical for the test being conducted, such as weight, loading
(center of gravity and inertia), airspeed, power, and wind, must be
 CHAPTER 7 Stall Testing 255

maintained within acceptable tolerances of the critical values during

flight testing.

The guidance material of FAA AC 25-7D states:

. . .weight tolerances that have been found acceptable for the specified flight
tests are ‘Stall speeds +5%’. Many flight tests need to be conducted at or very
near the maximum operating weight for the airplane configuration, particu-
larly those tests used to establish airplane flight manual (AFM) performance
information. The purpose of the test tolerances is to allow for variations in
flight test values, not to routinely schedule tests at less than critical weight
conditions or to allow compliance to be shown at less than the critical
weight condition.

AC 25-7D goes on to say:

It can be difficult or impossible to conduct testing at the airplane’s minimum
allowable weight with an airplane configured for conducting a flight test
program. If the minimum weight cannot be obtained (within the specified
tolerance limit) and compliance at the minimum weight cannot be clearly
deduced from the results at the tested weight, the testing should be con-
ducted on a production airplane (or other airplane on which the minimum
weight can be obtained).
The same goes for the center of gravity (CG) position. On larger planes, the
manufacturer may be able to install movable ballast in flight (water ballast
where the liquid is moved at various locations along the fuselage), but
smaller airplanes have to contend with fixed ballast, and then the CG travels
as the fuel is burned. To allow for a reasonable amount of testing with this
latter condition, FAA AC 25-7D states:
A test tolerance of +7 percent of the total CG range is intended to allow for
inflight CG movement. This tolerance is only acceptable when the test data
scatter is on both sides of the limiting CG or when adjusting the data from
the test CG to the limit CG is acceptable. If compliance with a requirement
is marginal at a test condition that is inside of the CG limits, the test should
be repeated at the CG limits.
Figure 7.19 represents the weight and CG location of a stall speed test cam-
paign for a small Part 23 jet to illustrate the +7% total travel envelope vs the
test conditions.

Instrumentation
The minimum instrumentation required to support the testing would be
to measure airspeed and altitude, preferably calibrated (already corrected for
instrument and position errors, and ideally corrected for any lag if significant
lag exists in the air data system). To solve Eq. (7.1), one must also have:
• Accelerometer for 1-g stall determination
• Aircraft weight, either by ZFW plus fuel on board or weight at engine
startup minus fuel burned
• CG position, preflight computed for a given weight or in-flight derived
from fuel quantity and location (assuming no one moves in the aircraft)
256 Operational Aircraft Performance and Flight Test Practices

14,000

13,000

12,000
Company
Weight (Ib)

FAA
11,000

10,000

9000

8000
12 14 16 18 20 22 24 26 28 30 32 34 36
CG (% MAC)

Fig. 7.19 Allowable tolerance of CG for stall speed test campaign.

To help with the validation that the stall had good

characteristics (to supplement the pilot’s notes in the Note: Recent years have seen
flight test log), one should also record: the introduction of
all-electric airplanes in flight
• Measurement of pitch attitude, free stream angle of testing. These airplanes have
attack (boom), and local angle of attack being a fixed weight and CG from
recorded by the stall protection system takeoff to landing. This
greatly reduces the
• Flight control positions and pilot forces (for stall contribution of weight to the
characteristics) test data scatter.
• Vibration at pilot station (for excessive vibration)
• Shaker and pusher signals to indicate when they fired to ensure the system
is functioning as required
Useful information for data reduction include:
• Engine setting (for thrust correction)
• Flap and gear position
Instrumentation should be verified prior and after each flight to know
whether it recorded proper information. The instrumentation should have
been calibrated within a manufacturer-defined period and verified that it
was still within valid calibration timeframe.

Test Execution
Testing for stall should obviously be done at a safe altitude (depending on
the airplane, typically at least 10,000 ft above ground minimum altitude) and
 CHAPTER 7 Stall Testing 257

with minimum crew. This author has had a chance to talk to crews from many
test centers, and there is no agreement on what minimum crew means. This
may vary from a single pilot on board for the light sport airplane (LSA) cat-
egory to two pilots with the right seat pilot acting as a flight test engineer
(FTE) to a standard two pilots and a dedicated FTE, to even two pilots and
multiple FTEs in the back of the airplane. It remains the responsibility of a
manufacturer’s flight testing organization to define what minimum crew
means, knowing that this crew will be exposed to some high-risk testing.
The requirements state that the airplane performing stall speed testing
be trimmed at a stall speed ratio. Obviously, the stall speeds have not
yet been formally determined, so the tester must use their best guess (educated
guess based on analysis and possibly wind tunnel test) prior to stall testing. The
crew is then provided with the best estimates in a clear format to help expedite
the execution. The requirements of §25.103(b)(6) are wide enough (between
1.13 VSR and 1.3 VSR ) to allow the tester some flexibility.
Stall speed testing requires that the stall be done at an entry rate of no
more than 1 kt/s [§25.103(c)]. Up to FAA AC 25-7C, the guidance material
suggested that the entry rate be computed using the average deceleration
between the minimum steady speed VS and a speed 10% greater (see
Fig. 7.20). With the implementation of the 1-g stall rule and the determi-
nation that the entry rate had significantly less impact on the load factor
corrected lift coefficient, this guidance was removed from AC 25-7 at revision
D, leaving each manufacturer to define the means to compute the entry rate
for a nominal 1-kt/s test condition. This guidance is still part of the AC 23-
8C flight test guide for small airplanes because the use of VS is still
acceptable.

KCAS (kt) Nzw (g)

140 1.2
1.1

130 1.0
0.9
KCAS (kt)

Nzw (g)

120 0.8
0.7

110 0.6
En
try
rat 0.5
e
100 0.4
130 140 150 160 170 180 190
Time (s)
1.1 Vs Vs

Fig. 7.20 Recommended practice for computation of entry rate.

258 Operational Aircraft Performance and Flight Test Practices

Verify
KIAS1 (kt) Boom KCAS (kt) Pressure altitude (ft) entry rate
Verify

Altitude (ft) (103)

18
production airspeed
BKCAS (kt)
KIAS1 (kt)

140
17 against ref airspeed
120
16
100
15
Elevator Stick force New Check elevator
(deg) (lb) (g)
40 1.5 and stick forces
Stick force (lb)
Elevator (deg)

new (g)
20 1.0 Verify
0 0.5 min g
–20 0.0 Pusher defined
Pitch
(deg)
BoomAOA CL1g
CL1g
stall (no g-break
(deg)
40 before)
BAOA (deg)

15
Pitch (deg)

CL1g
20
0 10

–20
Roll (deg)
20 Roll remained
10 within ±20 deg
(deg)
Roll

0
–10
–20
Stall warning Stick pusher
Warning (1=ON)
Pusher (1=ON)

(1=ON) (1=ON)
1.0
Verify
margins
0.5
(5 kt or 5%)
0.0
10 20 30 40 50 60
Time (s)

Fig. 7.21 Example of stall test monitoring.

To help with test execution, the crew should be provided with a decelera-
tion rate indication in the field of view (looking outside) of the pilot executing
the stall. This will help the crew best gauge the airplane entry rate while the
pilot focuses on handling the airplane coming to the stall point.

Telemetry Monitoring: Providing Feedback to the Flight Crew

To ensure repeatability of testing and to help ensure that the proper data
are being collected, a telemetry crew can provide timely feedback to the flight
crew on the acceptability of each test point as it is being flown or immediately
after being flown, before the flight crew moves on to the next test point.
Feedback should include things like entry rate, minimum load factor in
the maneuver, maximum roll angle reached (confirmed by the flight crew),
maximum aileron used, load factor at shaker (for the dynamic entry and
turning stalls), and minimum speed seen in the stall. Figure 7.21 shows an
example of such monitoring. Obviously, a lot of review must be done for
every single test condition, and multiple eyes on the data trace does help.
Timely feedback will also ensure quicker testing because it is often faster
to repeat the same configuration a second time than to come back to this
configuration from another one.
 CHAPTER 7 Stall Testing 259

Proposed Test Card

Stall testing should be documented carefully and diligently during its
execution so as not to unnecessarily repeat high risk testing. Figure 7.22
shows a suggested test card for data collection while performing stalls. If
used by both the flight crew and the telemetry crew, verification of the test
point flown can be accomplished in real time. It also provides an early vali-
dation of the data being collected. It serves as a record should the data collec-
tion system fail in flight, and it contains the crew assessment of the stall
handling characteristics as each stall is performed.

Stall Test Risks and Mitigations

Testing for stall characteristics and stall speed will bring the airplane
to the edge of the envelope. The manufacturer should have a good,

Performance stall test card

Test point: Weight: CG:
Start time: Vtrim: Altitude:
Configuration
Flap: Gear: [ ] Up [ ] Down Spoilers: [ ] Ext [ ] Ret
Thrust setting: [ ] Idle [ ] Power N1 %:
Test results
Target/Crew Recorded TM Feedback
Entry rate:
V warning:
AoA warning:
Stall warning defined by: [ ] Shaker [ ] Buffet
Vstall:
AoA stall:
Altitude stall:
Altitude at recovery:
Stall defined by: [ ] Pusher [ ] Nose down [ ] Elevator aft stop [ ] Buffet
Crew comments:

Telemetry comments:

Fig. 7.22 Proposed test card to record stall speed test results.
260 Operational Aircraft Performance and Flight Test Practices

pre–flight test, expectation of the airplane’s characteristics at the maximum

angle of attack and address the expected test risks with proper mitigations.

Risk Identification
Even with an airplane with gentle stall characteristics, there is always a
high risk of something going wrong while performing stall tests. Some of
the most common risks for performing a stall include:
• Risk of departure from controlled flight, including spins
• Risk of engine flameout due to inlet distortion at high angle of attack
• Risk of deep stall where pitch control becomes ineffective
• Risk of excessive vibration causing structural damage to the airplane
The excessive vibration requirements are supposed to be a pilot’s call (cri-
teria defined by AC 25-7D, discussed earlier), but with proper instrumenta-
tion (e.g., accelerometer on the seat rail of the pilot), some calibration of the
event is possible. On some large transport airplanes, this excessive vibration
could lead to structural damage under certain conditions.
There is also, of course, the risk of not being compliant (stick force,
roll-off, etc.), but this is more program related than a crew safety concern.
These represent a risk of repeat testing, which translates into a schedule
impact and an additional cost to the program.

Deep Stall
Deep stall is a condition where, as the flow separates from the wing and/or
the engine nacelles, it blanks the horizontal tail, rendering the primary pitch
control (elevator) ineffective. T-tail airplanes are typically prone to this con-
dition. A typical sequence of events for a deep stall (see Fig. 7.23) starts with
an airplane trimmed at a lower AoA and slowing down. As the airplane
stalls (flow separation), there may be a rapid pitch attitude (and AoA) increase
not immediately countered by the pilot. Then, the airplane will pitch up (AoA
increase) to another stable condition (zero pitching moment) at a much higher
AoA, and the crew will not have any ability to push out of the situation. (A full
nose-down elevator will not create a restoring nose-down pitching moment.)

CM
Trim Deep stall
point region Trim
Stall

points

V 
Full nose-
down elevator

Fig. 7.23 Deep stall region.

 CHAPTER 7 Stall Testing 261

1800
1600

Riser tension (kgf )

1400
1200
1000
800
600
150
400
Ship airspeed

Airspeed too 200

100 low to record 0
24 25 26 27 28 29 30 31 32
50 Time (s)

Controls regained
60

Chute released
Boom AoA

40 AoA not decreasing

20

0
50

0
Pitch

–50
Stall chute rocket fired

Max nose down

Stall chute deployed

–100 pitch ≈80 deg

40
Elevator

20

0 Max elevator
Start of rapid pitch up

applied
–20
4 Peak g ≈3.2
3
NZstall

1
0
14,000
12,000
Altitude

10,000
Peak ROD ≈20,000 fpm
8000
6000
30 40 50 60 70 80 90 100

15:28:00:081 Elapsed time (s)

Fig. 7.24 Flight test deep stall event.

The flight test verification that the airplane does not have that tendency is
clearly stated in §25.201(d)(3): “The pitch control reaches the aft stop and
no further increase in pitch attitude occurs when the control is held full aft
for a short time before recovery is initiated.” Flight test airplanes that may
be prone to deep stall are usually fitted with a stall chute to pull on the tail
and reduce the AoA in case of emergency.
Figure 7.24 shows an example of a deep stall event that occurred during
a flight test program. Note how rapidly the angle of attack increased
262 Operational Aircraft Performance and Flight Test Practices

beyond the onset of the pitch-up. The instrumentation pegged at 50 deg, so

the exact “stable AoA” in this stall regime was not known. The pilot
pushed the elevator to full nose-down, but the airplane did not react
much (small pitch angle reduction, but no reduction in the AoA below
the saturation angle). The crew fired the stall chute at time 60 s, and it
was fully deployed within about 2 s. It still took over 7 more seconds for
the stall to break and the airplane to reach 80 deg nose-down attitude
and an AoA within flight range, at which time the parachute was let go;
the rate of descent was already near 20,000 fpm. The airplane safely recov-
ered and returned to base.
Deep stall was an unknown in the 1960s when the first accidents
occurred (see the accident boxes for the BAC 1-11 and the Hawker Siddeley
Trident). Even with the knowledge gained since then, deep stall remains a
risk, and an airplane that can be prone to such an event needs to be
tested thoroughly.

Stall Test Accident—Hawker Siddeley Trident, 3 June 1966: Aircraft

G-ARPY departed the Hawker Siddeley factory’s airfield at Hatfield in the
late afternoon to conduct stall testing over East Anglia. On the fourth stall
test, the recovery systems (stick shaker and stick pusher) were deliberately
made inoperative as part of the test. During the stall, the recovery action
from the pilots was left too late, and the aircraft first entered a deep stall
from which it then entered a flat spin and could not be recovered. All four
test crew were killed.
The crash site showed the telltale signs of the deep stall event with the air-
plane’s parts in essentially the same location (signs of a large vertical speed
with little to no forward speed).

Stall Test Accident-BAC 1–11, 22 Oct. 1963: The accident happened on 22

Oct. 1963 when a BAC One-Eleven (registration G-ASHG) took off from
Wisley Airfield. The BAC One-Eleven was on a test flight to evaluate the air-
craft’s stability and handling characteristics during the approach to and recov-
ery from the stall with a center of gravity in varying positions. The aircraft was
on its fifth stalling test. The flight crew initiated a stall of the BAC One-Eleven
 CHAPTER 7 Stall Testing 263

at a height of about 16,000 ft and with 8 deg of flaps; the plane entered a stable
stall (deep stall). The aircraft began to descend at a high vertical speed and in a
substantially horizontal attitude, and eventually struck the ground with very
little forward speed. The aircraft broke up and caught fire, killing all seven
crew on board.
The cause of the accident was that during the stalling test the aircraft
entered a stable stalled condition, recovery from which was impossible. This
was the first accident to be attributed to the phenomenon known as deep
stall. Lessons learned were applied to the Douglas DC-9 design [3].

Spins: Single-Engine Part 23 Airplanes

Part 23 single-engine propeller-driven airplanes present a special chal-
lenge in that the spinning propeller will create an unsymmetrical airflow
onto an otherwise symmetric airplane. As well, a sudden increase in engine
power will also lead to a large torque that may spin the airplane around
the propeller axis. This is exemplified in Fig. 7.25 where, during the recovery
from stall, the student pilot applied power as he was just starting to lower the
nose. This large torque from the propeller and the resulting asymmetric aero-
dynamics while at high angle of attack resulted in a rapid departure from
controlled flight.
FAA AC 23-8C provides the following guidance for recovery from stall
(underline added by author):

Recovery. The flight tests include a determination that the airplane can be
stalled, and flight control recovered, with normal use of the controls.
Section 23.201(a) requires that it must be possible to produce and correct
roll by unreversed use of the roll control and to produce and correct
yaw by unreversed use of the directional control. The power used to
regain level flight may not be applied until flying control is regained. This
is considered to mean not before a speed of 1.2 VS1 is attained in the recovery
dive.

Part 23 regulations, §23.143(b), require that it be possible to affect a “smooth

transition” from a flying condition up to the stalled flight condition and
return without requiring an exceptional degree of skill, alertness, or strength.
Any need for anticipated or rapid control inputs exceeding that associated
with average piloting skill is considered unacceptable.
264 Operational Aircraft Performance and Flight Test Practices

- RPM = 2500
- RPM = 1470

- RPM = 1050
100

90
Corrected KTAS Airspeed
Pitch angle (deg), roll angle (deg), estimated sideslip (deg)

80 discussion
Power
Recorded KTAS increase
70 Inverted
airspeed true recorded (KTAS), KTAS corr

60

50

40 Significant airspeed
recording error
30

20
Recorded pitch

10

–10 Estimated sideslip angle

Recorded roll
–20
2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039
Time (s)

Fig. 7.25 Cirrus SR22 accident, NTSB Docket ERA09FA169.

Risk Mitigation: Attitude Recovery Device

The loss of control during a stall test or a spin test (spin test for Part 23)
is always a risk. That risk can be minimized by the use of spin chutes (also
called stall chutes), the preferred option to allow an airplane to recover from
a deep stall or an uncontrollable spin. A stall chute is installed in the most
aft section of the airplane. It is a small-diameter parachute that is typically
fired by pyrotechnic charges to pull on the airplane’s tail and pitch the air-
plane downwards (reduce the angle of attack or stop a spin) to allow its
crew to recover. The spin chute is jettisoned as soon as the crew regains
control of the airplane so as not to overload the tail with the drag force
from the parachute.
During a stall test, the crew should be limited to the minimum required to
successfully complete the testing. The crew should be fitted with parachutes
and helmets, and the airplane should be equipped with a crew escape system
to allow fast egress in an emergency situation. The surface winds in the test
area should be limited to 15-25 kt if the crew does bail out.
Sometimes the safety equipment does not work as intended or the test
risk reduction does not call for it and an event occur. Such was the case
for the following three flight test accidents, the Canadair CL600 in 1980,
the Canadair CRJ100 in 1993, and the Cessna 162 Skycatcher in 2008. The
flight test community and the spin chute providers have learned from
these accidents.
 CHAPTER 7 Stall Testing 265

NTSB LAX80FA073, Stall Test Accident—Canadair Challenger 600,

3 April 1980: During dedicated stall testing, the airplane departed and
entered a deep stall. Crew deployed the stall chute, the airplane pitched
down sharply (70 deg nose down), and control was regained. The aircrew
then tried to jettison the stall chute, but the jettison mechanism failed to
release the chute. The aircraft failed to regain controlled flight and crashed
into the desert floor at nearly 17,000 ft/min descent rate, almost wings
level. At about 3500 ft above ground, the pilot called for the crew to bail
out, 2 of the 3 crew members were able to bail out in time. [Note: This
author had a chance to work with one of the surviving members of this

NTSB CHI93MA276, Steady Heading Sideslip Test Accident—Canadair

CRJ100, 26 July 1993: The crew was performing a steady heading sideslip
at low speed. The procedure required the pilot to stop at stall warning or
15 deg sideslip, whichever occurred first. The pilot continued to maximum
rudder deflection (about 21 deg sideslip), past stall warning. One wing
stalled first, and the airplane rolled rapidly through 360 deg and entered
deep stall. The copilot attempted to deploy the stall chute, but it had not
been set up correctly prior to the maneuver, and the stall chute departed
the airplane. Full control of the airplane was not regained before impact. All
crew members were lost.

Risk Mitigation: Test Progression

One does not go to the stall on the first attempt. A careful stall campaign
should be planned so that one can ease into stall testing from expected,
266 Operational Aircraft Performance and Flight Test Practices

more gentle stall configurations to more interesting configurations. Testing

should be started preferably at midweight and mid-CG, and then progress
to forward CG and heavier weights before going to aft CG and light
weight (for a traditional airplane layout with an empennage in the back of
the plane).
Each time a new combination of configuration, weight, and CG is
approached, the crew should perform an AoA buildup in which the airplane
is brought to an angle of attack less than the expected stall angle (say 3 deg
less), and then recovered by normal use of controls. This allows the crew
to verify controllability and high AoA, find signs of flow separation/buffeting,
and note possibly developing unacceptable characteristics. With proper
characteristics demonstrated, the crew can then inch their way up by
another degree or so and again check the characteristics. This is continued
until the airplane is considered stalled per §25.201. The airplane may stall
as predicted, earlier, or later.
The crew should establish abort criteria and a minimum test altitude (and
minimum bailout altitude). Engine ignition should be on continuously, at
least during the initial part of the stall testing, to observe the engine charac-
teristics in a stall. If an auxiliary power unit (APU) is available to help with
engine start, it should be running during the test.

NTSB DFW08FA234, Cessna 162 Skycatcher N162XP, 18 Sept. 2008: The

flight originated at approximately 1100, to obtain an assessment of the air-
craft’s spin characteristics. Sixteen spins were planned, with the aircraft in
four different configurations. The aircraft had successfully completed the
first configuration with a set of four spins. The pilot then completed three
of the four spins in the second set of spin testing. With the aircraft at
10,000 ft, the pilot initiated a spin to the left. Once the spin was established,
the pilot made the planned control inputs. The airplane failed to recover
from the spin. Despite several attempts by the test pilot to recover the aircraft
from the maneuver, the aircraft continued to spin. At the planned altitude of
6000 ft, the pilot elected to deploy the aircraft’s recovery parachute and pulled
the chute’s activation handle. The aircraft’s parachute did not deploy, and the
aircraft continued to spin. The pilot then elected to jettison the cabin door,
exit the airplane, and deploy his parachute. The airplane impacted the
ground in a wooded area.
Control continuity was established to the respective flight controls. An
initial examination of the aircraft’s parachute system revealed that the
rocket system used to deploy the chute had activated; however, the parachute
failed to be pulled from its canister.
During flight testing and the subsequent review of test data, it was discov-
ered that an unrecoverable spin could develop. The accident airplane was a
preliminary configuration, and the manufacturer abandoned the configur-
ation for production airplanes.
 CHAPTER 7 Stall Testing 267

Creating a Stall Speed Model

The stall speed data have been collected, and it is now time to perform the
data reduction so as to come up with a model that will represent the air-
plane’s characteristics as closely as possible under all flight conditions.
Once the model is complete, it is used in the data expansion phase to
create AFM data for the crew.

Data Reduction
The data reduction phase, as just mentioned, consists of creating a model
for the flight testing done. The data were collected under the best conditions
possible, as specified in the test plan, with natural test scatter present. The
data must then be corrected by removing the effects of various parameters
such as entry rate, thrust, center of gravity position, and so forth. For each
test point, one must verify that all the required certification requirements
and margins are met. One must also verify that the air data system provides
accurate (certifiable) information to the crew.

Configuration Effects
There can be significant differences in the way the airplane stalls depend-
ing on the airplane configuration (see Fig. 7.26). For example, highly swept

3.5

3
Flaps landing
Pusher
2.5

2 Pusher
Flaps takeoff
CL

1.5
Flaps UP Pusher
1

0.5

0
–2 0 2 4 6 8 10 12 14 16 18
Angle of attack (deg)

Fig. 7.26 Different expected maximum angles of attack per flap configurations.
268 Operational Aircraft Performance and Flight Test Practices

wings tend to have flow separation at the wing tip, first resulting in a tendency
to pitch up as they near aerodynamic stall. But the same wing in the landing
flap configuration with slats deployed could actually have a stall pattern start-
ing inboard that creates a nose-down pitching moment as the airplane’s aero-
dynamic stall is approached. Also, the maximum usable angle of attack tends
to decrease with increasing flap setting (with exception again based on air-
plane configuration). It is, therefore, the normal approach to define stall
speed based on flap configuration. Additional impact may be seen by
having a landing gear up vs down, and that, too, can be addressed in the mod-
eling. Once the configuration is defined as specified in §25.103(b)(3), one can
proceed with the stall speed analysis.

Calibrating the Air Data in the Stall

The stall model that will be created is based on a corrected lift coefficient
CLcorr . This corrected lift coefficient is a function of several parameters, all of
which need to be validated before the coefficient can be used.
nzw W
CLcorr ¼ ð7:3Þ
1
r V2 S
2 SL EAS
The first parameter of importance is, of course, the airspeed. The airplane
will be flying at large angles of attack; the flow on the nose of the airplane will
be substantially different than it was at the minimum airspeed for calibration,
as discussed in Chapter 4 and shown in Fig. 7.27. For this reason, the tester
must find a reliable source of air data, the truth source, to define the
stall speed.
Manufacturers often use a trailing cone as the static pressure truth
source. Depending on the installation selected (see Fig. 7.28), a trailing
cone that was providing good information during the airspeed calibration

Probe 2 Probe 1

 = 0 deg  = 6 deg

 = 3 deg  = 10 deg

Fig. 7.27 Local flow variation (CFD analysis, pressure map and flow lines).
 CHAPTER 7 Stall Testing 269

(b)

(a)

(c)

Fig. 7.28 Example of location of trailing cone attach point with airplane at high AoA.

exercise (steady airspeed, medium to low angle of attack; see Chapter 4) may
not always work as well during stall testing, because it may now come into or
near the wake of the plane. For example, the cone at the top of the vertical tail
for an airplane with a low horizontal tail and low engines (Fig. 7.28a) may still
provide quality data during stall testing, whereas the cone attached to the
fuselage of a high wing and propeller-driven airplane (Fig. 7.28c) may not
be good in stall. The instability of the cone once it gets into the wake of
the plane is often seen as pulses in the data trace. (The cone starts whipping
around and records some dynamic pressure.) The tester can verify the stab-
ility of the trailing cone early in the testing if an experimental camera is
installed looking aft at the expected location of the cone. One must also
remember that the longer the trailing cone, the greater the lag response
time; because a stall is a dynamic condition, one must correct for this effect.

Production Air Data System Check in a Stall

The readings from the production air data system must be compared to a
calibrated system to ensure that the airspeed reported to the crew as stall is
approached does not provide misleading cues. The EASA regulation support-
ing this is:
EASA CS25.1323(d) From 1.23 VSR to stall warning speed the IAS must
change perceptibly with CAS and in the same sense, and at speeds below
stall warning speed the IAS must not change in an incorrect sense.

EASA also provides an acceptable means of compliance (AMC): “An accep-

table means of compliance when demonstrating a perceptible speed change
between 1.23 VSR to stall warning speed is for the rate of change of IAS with
CAS to be not less than 0.75.”
Similar requirements were added in 2003 to FAA 14 CFR 25.1323 under
para (d) at Amendment 25-109 and for TCCA 525.1323, also under para (d).
Obviously, an increasing indicated airspeed while the actual (calibrated) air-
speed is decreasing could lead to crew confusion and eventual unintentional
stall. Figure 7.29 shows an acceptable production air data system output
270 Operational Aircraft Performance and Flight Test Practices

KIAS KCAS Nzw Shaker 2

(kt) (kt) (g) 1 = ACTIVE
120 1.2

1.0
nzw
110

Stall warning
KCAS
0.8
KIAS

Natural stall

Shanker 2
Nzw (g)
KCAS (kt)
KIAS (kt)

Recovery
100 0.6

KIAS changes 0.4

perceptibly KIAS does not
90 with KCAS change in an
incorrect way
0.2

80 0.0
60 65 70 75 80 85 90
Time (s)

Fig. 7.29 Comparing the airplane air data system to a truth source during stall.

(KIAS) compared to a truth source from a nose boom (KCAS). The reader
must remember that the boom airspeed may still need to be calibrated and
lag corrected.
Figure 7.27 shows an example of a possible source of air data error with
varying the angle of attack. Probe 2 is located behind and above Probe 1 on
this forward fuselage. CFD analysis shows that as the angle of attack is
increased, Probe 2 may come to be in the wake of Probe 1. This will
impact the air data reading of Probe 2.

Accelerometer Correction
When performing the analysis for 1-g stall, one must know the value of
the acceleration normal to the flight path nzw . For the 1-g stall rule, the
load factor corrected lift coefficient CLcorr is computed per Eq. (7.3) in
terms of equivalent airspeed or as follows when considering true airspeed:
nzw W
CLcorr ¼ ð7:4Þ
1
r sV 2 S
2 SL
 CHAPTER 7 Stall Testing 271

The load factor nzw used must be perpendicular to the flight path;
however, accelerometers installed in an airplane are fixed in orientation
with respect to the airframe with typical components being along the longi-
tudinal axis nx , lateral axis ny , and vertical axis nz . The load factor normal to
the flight path nzw can be computed from the fuselage fixed vertical and
longitudinal acceleration to a flight path vertical acceleration via the follow-
ing transformation:

nzw ¼ nz cosðaÞ þ nx sinðaÞ ð7:5Þ

where a represents the airplane true angle of attack with respect to the longi-
tudinal axis (the axis parallel to the reference nx ; see Fig. 7.30). This angle can
be measured with a calibrated nose boom angle of attack vane or can be com-
puted by flight path reconstruction methods. Care must be taken to ensure
that the signs of the measured acceleration are correct for Eq. (7.5). For
example, Fig. 7.31 shows that if nx is positive with a positive value of pitch
attitude in level flight and nz is also positive, then both shall be summed
up per Eq. (7.5).
Note in Fig. 7.31 that the recorded accelerations (nx and nz ) may be noisy
and that their use to compute an nzw may lead to an unusable value for the
purpose of performing stall analysis (symbols on bottom trace of Fig. 7.31,
labeled Nzw). Therefore, some filtering of the recorded signals may be
necessary to compute a more appropriate value (solid line on bottom trace
of Fig. 7.31, labeled Nzstall), but filtering may introduce some lag in the
data and should be validated prior to use.
Alternatively, the applicant can install a single accelerometer perpendicu-
lar to the average angle of attack of all stall configurations. A small error on
the stall angle (1–4 deg) results in less than 0.25% error on the load factor nzw
(see Fig. 7.32).
The maximum aircraft lift coefficient CLmax is the maximum value of
CLcorr measured during a stall maneuver. One should not account for tran-
sient peaks in CLmax due to noisy accelerometers or dynamic increase in load
factor due to such things as a control input. These do not reflect the capa-
bility of the wing to generate lift. There is usually a noticeable break in the

V
x
z

Fig. 7.30 Acceleration axis transformation.

 272
Nx (g) Nz (g)

Operational Aircraft Performance and Flight Test Practices

1.5

1.0

Nx (g)
Nz (g)
0.5

0.0

–0.5

Pitch (deg) Boom AoA (deg)

30
Boom AoA (deg)

20
Pitch (deg)

10

–10
Nzw (g) Nzstall (g)

1.4
Nzstall (g)

1.2
Nzw (g)

1.0
0.8
0.6
20 30 40 50 60 70
Time (s)

Fig. 7.31 Correcting fuselage-mounted accelerometers for normal acceleration to flight path.
 CHAPTER 7 Stall Testing 273

Ideal nzw angle 15 deg Ideal nzw angle 13 deg

from airframe vertical from airframe vertical

 = 17 deg  = 15 deg =
V
 = 2 deg 10 deg  = 13 deg Horizon
 = –3 deg
V

Flaps UP, gear UP stall Flaps Landing, gear DOWN stall

Fig. 7.32 Selecting the average angle of attack for a single accelerometer.

recorded load factor (see Fig. 7.33) at the time of natural stall, indicating that
the aircraft has started “falling.”
The advantage of having a good nzw during stall testing can be seen in
Fig. 7.34. The data presented show a stall test done in a specific configuration
(flap and gear) at two different weights by two different pilots. In stall
method 1, the pilot favored maintaining the pitch attitude as the stall was
approached, letting the flight path angle (FPA) decrease more rapidly; this
helped maintain a deceleration rate closer to 1 kt/s. In stall method 2, the
pilot favored a more constant elevator input that resulted in a small pitch
rate increase near stall; this resulted in a small increase in the deceleration
rate. For both methods, the approach to stall met the intent of the guidance
material. Once both maneuvers are corrected with the nzw , one sees that the
resulting lift coefficients are essentially the same (within the scatter of the
test). The load factor–corrected lift coefficient clearly shows the lift capa-
bility of the wing.

Verifying the Test Weight

Another major impact to the stall speed computation is the airplane
weight. The weight is not measured in flight; it is estimated based on a pre-
flight measured weight (weighing of the plane) and an estimated fuel burn
at the time of the test. (As we have noted previously, electrical airplanes do
not have to contend with weight change in flight.) That estimate of fuel
burn can be through the measurement of a fuel volume (see Fig. 7.35) con-
verted to a fuel weight with a standard fuel density possibly corrected for
fuel temperature or via a fuel flow rate measurement (also a fuel volume
used vs time corrected for a standard density and possibly fuel tempera-
ture). If one measures fuel burn via a fuel volume, one can possibly intro-
duce scatter in flight data by having an instantaneous “apparent” weight
variation (increase or decrease) in flight due to the fuel movement inside
the fuel tank (fuel movement away from or onto the probes). In the
example shown in Fig. 7.35, this impact is clearly seen as the airplane
initiates the stall maneuver and the weight of the plane (as recorded)
274 Operational Aircraft Performance and Flight Test Practices

110

KCAS

100
VCLmax

90
g-break

1.0

nzw

0.8

CLmax

CLcorr

Time

Fig. 7.33 A g-break at natural stall.

starts increasing. This weight change has a direct impact on the inflight
measured lift coefficient that is used to determine VCLmax (last data trace
in Fig. 7.35). Before using a weight source for the computation of the
AFM maximum lift coefficient, one must ensure that all corrections to
the weight are done appropriately, including using best source data like
pre- and postflight weight of the airplane.
The flight test results of Fig. 7.35 show a variation of the weight, as
recorded, of +15 lb. Is this big? By itself, a weight scatter does not mean
much. Once put in perspective of the test weight (nominally 11,830 lb), it rep-
resents an error of about 0.13%, which may seem reasonable. So what is the
value of a pound of error? We use the chart shown in Fig. 7.36 to quantify the
impact of such error.
Figure 7.36 shows three different airplane designs (three different wing
loadings) at two different stall conditions. One can represent a flaps-up stall
 Stall method 1 Stall method 2
KIAS (kts) KIAS (kts) 1.5
150 150
S = 320 ft2
140 140 1.4
130 130
KIAS

KIAS
120 120 1.3
W = 19,760 lbs W = 21,000 lbs
110 110
100 100
1.2
Pitch (deg) FPA (deg) Pitch (deg) FPA (deg)

CL
30 30
20 20 1.1
Pitch (deg)

Pitch (deg)
FPA (deg)

FPA (deg)
10 10
0 0
1
–10 –10
–20 –20
–30 –30 0.9
Nz (g) Nx (g) Nz (g) Nx (g)
1.5 1.5 0.8
1.0 1.0 10 15 20 25 30
Nx (g)

Nx (g)
Nz (g)

Nz (g)

Angle of attack (deg)

CHAPTER 7
0.5 0.5

0.0 0.0 Stall method 2 Stall method 1

20 30 40 50 20 30 40 50 60 70
Time (s) Time (s)

Stall Testing
Fig. 7.34 Benefit of nzw on data reduction for CL:

275
276 Operational Aircraft Performance and Flight Test Practices

(CLmax near 1.25) and one a landing flap configuration (CLmax near 2.5).
The solid lines for each wing loading represent the stall speed for a
given maximum lift coefficient. Error margins on weight, in the form of
a wing loading, were added as dashed lines with the magnitude of the
error being +1 lb/ft2 . Figure 7.36 shows that a weight error around
1 lb/ft2 for an airplane with a baseline wing loading of 50 lb/ft2 would
have a stall speed error around 1.1 kt in the flaps-up configuration. (The
speed error is for a given fixed stall lift coefficient.) But the same weight
error in the landing configuration would yield a smaller airspeed error.
The chart also shows that if the wing loading is increased, the speed
errors reduce.
So, to get a 1-k error for the airplane in Fig. 7.35 with a wing loading of
50 lb/ft2 (the airplane weight was nominally 11,830 lb, so this would give a
wing area of about 235 ft2 ), one would need to misjudge the weight by

120
110
Airspeed

100

90
80
1.5

1
NZstall

0.5

0
36

34
CG

32

30
11,900
Weight

11,850

11,800
2.5
CL|Ref CG

1.5

Fig. 7.35 Weight growth in the data stream.

 CHAPTER 7 Stall Testing 277

180

W/S error (dashed lines) ±1 lb/ft2

160
VSR ≈ ±0.8 kt

W/S = 100 lb/ft2

140
Stall speed (kt)

W/S = 75 lb/ft2

120
VSR ≈ ±1.1 kt VSR ≈ ±0.5 kt

100

80 VSR ≈ ±0.8 kt

W/S = 50 lb/ft2
60
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
CLmax

Fig. 7.36 Impact of a weight computation error on max lift coefficient and stall speed.

about 225 lb in the flaps-up configuration and by about 295 lb in the landing
configuration.
In the end, weight is a controllable variable in the data reduction for
stall and under the control of the performance engineer during the test plan-
ning and execution. Good execution results in less scatter in the stall
speed model.

CG Correction
The CG location during the stall speed test will have an impact on the
airplane lift coefficient. On a traditional airplane layout with the horizontal
empennage in the back, the most adverse CG location is the forward one
(least lift available). Because the maximum forward CG in the envelope
most likely changes with weight, the flight measured lift coefficient
should be corrected to a reference CG (see Fig. 7.37) during the data
reduction phase to help reduce the data scatter. This will help provide
for a clearer demonstration of the effects of various other parameters
such as thrust, entry rate, gear, spoilers, and so on.
The guidance material of FAA AC 25-7D provides the following equation
as an acceptable means of correcting the lift coefficient:
   
MAC
CL jstd ¼ CL jtest 1 þ ðCGstd  CGtest Þ ð7:6Þ
‘t
 278
Operational Aircraft Performance and Flight Test Practices
a) b) c)
Reference CG

40,000 7% total travel

17,980
MRW 38,650 lb (17,530 kg) tolerance

Reference CG
MTOW 38,600 lb (17,465 kg) 24,000 22,000
38,000
16,980 21,000

Reference CG
36,000 22,000 20,000
15,980
Aircraft weight (lb)

19,000
34,000
20,000

Aircraft weight (kg)

18,000

Weight (lb)

Weight (lb)
MLW 33,750 lb (15,310 kg) 14,980
32,000 17,000
18,000
13,980 16,000
30,000 Fuel burn curve
15,000
12,980 16,000
28,000 14,000
MZFW 26,100 lb (11,840 kg)
11,980 14,000 13,000
26,000
12,000
24,000 10,980 0 5 10 15 20 25 30
12,000
CG (% MAC)
MFW 23,100 lb (10,480 kg)
22,000
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 9980 10,000
% MAC
5 10 15 20 25 30 Smaller corrections
CG (% MAC)

Fig. 7.37 Selecting a reference CG for data reduction.

 CHAPTER 7 Stall Testing 279

where
CL jstd ¼ corrected lift coefficient at the standard or reference CG
location for the data reduction
CL jtest ¼ computed lift coefficient for the CG at the time of the test
MAC ¼ wing mean aerodynamic chord
‘t ¼ moment arm of the horizontal tail (quarter chord of wing
MAC to quarter chord of tail MAC)
CGstd ¼ standard or reference CG location for the data reduction
CGtest ¼ CG at the time of the test
The reference CG line should typically be the most forward CG point of
the envelope. The guidance material does specify that Eq. (7.6) should be
used for small corrections and not be used to correct test conditions from
extreme CG position (from aft CG to forward CG, for example). This
author’s position is that a 5–10% MAC CG correction is acceptable; if the
forward CG line has more travel than this (Fig. 7.37c), then a more mid-
forward reference CG should be used for the data reduction.
It may sound counterintuitive to correct the test CG to a more forward
value, thereby decreasing the recorded value, but the correction does help
remove some of the test scatter and creates a better stall model (lift coefficient
at stall model). What is allowed at the time of expansion of the data to create an
AFM model is to use Eq. (7.6) in reverse and compute the equivalent maximum
forward CG lift coefficient at the actual forward CG line of the weight and CG
envelope, thus regaining any loss of CLmax in the reduction.

Thrust Correction
The AFM stall speed corresponds to a zero- Note: Thrust effect on FBW
thrust speed [see §25.103(b)(1)]. The certification HALF-protected airplane, a
authorities do not ask that stall testing be done pilot’s report: Flying the
A400 M: Flying the Airbus
with engines off, but rather that the stall speed be
A400 M at a weight of about
defined with no thrust force contribution to the 225,000 lb and with flaps 4
lift; however, they do require that the impact of selected, we pulled the thrust
the propulsion system on the ability of the wing to levers back to idle and
maintained altitude, causing
produce lift be considered. This is an important
the aircraft to decelerate.
point to consider during the stall speed data With full aft stick, we
reduction because some aircraft can have a signifi- reached alpha max at 98
cant increase in lift coefficient with engines provid- KIAS. After setting
maximum climb power, we
ing power. The most obvious case is wing-mounted
continued to increase nose
engine propeller–driven aircraft with their high- attitude to slow the aircraft at
speed slipstream over the wing; see the following about 1 kt/s with each
Note box on the A400 M pilot report. engine producing about 7900
shp. With the stick all the
Let’s address jet-driven airplanes first. Testing
way aft, the aircraft
for stall speeds is normally done with the thrust decelerated to 78 KIAS.
lever at flight idle. Then, the thrust contribution to
280 Operational Aircraft Performance and Flight Test Practices

the lift is generally small. Regardless, the guidance material suggests that the
impact of thrust on the stall speed be verified by flight testing or by analysis.
The guidance material of FAA AC 25-7D proposes the following
approach to determine if the thrust has a significant (more than 0.5-kt)
impact on the stall speed:
1. At least three stalls should be conducted at one flap setting, with thrust set
to approximately the value required to maintain level flight at 1.5 VSR in
the selected configuration. These test points and the ones at idle thrust
setting are then plotted on a chart (all testing performed at similar weight
and altitude); see Fig. 7.38.
† These data may then be extrapolated to a zero thrust condition to
eliminate the effects of idle thrust on stall speeds, using Eq. (7.7).
   
CLmax T ¼0 ¼ CLmax CGcorr; test  mTidle ð7:7Þ

where m is the slope of the data collected in Fig. 7.38, and Tidle is the
idle thrust computed posttest. Note we already applied a first correction
(CG correction) to the test CLmax prior to applying the
thrust correction.
For airplanes where engine thrust has little impact on the airplane’s aero-
dynamics (this is often the case for aft fuselage–mounted engines with a
T-tail; see Fig. 7.39), it may be possible to correct the effects of thrust by
analysis. The thrust component to the lift coefficient is
T sinðaT Þ
DCLT ¼ ð7:8Þ
1
r V2 S
2 SL EAS
aT ¼ a þ iT ð7:9Þ
where a is the airplane angle of attack (with respect to the longitudinal axis),
and iT is the engine thrust line incidence with respect to the longitudinal axis.

Thrust for
1.5 VSR

CLmax
Idle

Zero thrust
extrapolation

Thrust

Fig. 7.38 Flight test correction for zero thrust.

 CHAPTER 7 Stall Testing 281

T

Fig. 7.39 Thrust angle of attack.

Then, the thrust-corrected lift coefficient becomes

 
CLstd ¼ CLtest CG corrected  DCLT ð7:10Þ

The guidance material specifies that “If the difference between idle thrust
and zero thrust stall speed is 0.5 knots or less, the effect may be considered
insignificant.” Then, no correction to the lift coefficient is needed.
Correction to zero thrust is also allowed for propeller-driven airplanes as
long as the engine is at idle (throttle closed) or near idle (power required to
get zero thrust) and the propeller during the stall is at the takeoff setting
(typically lowest blade pitch for forward flight). The guidance material does
not allow a substitute for propeller setting other than in the takeoff configur-
ation so as to capture the aerodynamic impact of the propellers in
this configuration.
Let’s provide a numerical example of what an acceptable means of com-
pliance would be to determine the zero thrust setting for the stall test.
Assume the expected stall speed is 60 KTAS at the test altitude and test
weight. The minimum engine setting (idle) RPM is 800, and the propeller
diameter is 6 ft. One can compute a minimum advance ratio (see Chapter
8) J ¼ 1.266. Then one must refer to the propeller manufacturer’s propeller
performance chart, an example of which is shown in Fig. 7.40. For that pro-
peller, the takeoff setting would be 15 deg, and an advance ratio of 1.266
would actually provide negative thrust at the test condition. To bring the
thrust coefficient to zero with a propeller blade pitch of 15 deg, one would
need to lower the advance ratio to 0.81. The speed is expected to be 60
KTAS at stall, so one can alter the advance ratio by increasing power to
the point where zero thrust is produced. In this case, the engine power
should be increased to 1250 RPM.
This correction applies for piston-prop and electric motor-prop combi-
nations. For turboprop-equipped airplanes, one must contend with the pro-
peller effect and the residual thrust from the gas turbine. The guidance
material suggests that the effects of engine power on stall speeds for a tur-
bopropeller airplane should be determined with engines idling, throttles
282 Operational Aircraft Performance and Flight Test Practices

.14

.12

.10

.08 45º Blade angle at 0.75R

CT 40º
35º
.06
30º
.04 25º
20º
15º
.02

0 .2 .4 .6 .8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
V/nD

Fig. 7.40 Finding the blade pitch for zero thrust at stall [2].

closed, and the propellers in the takeoff position. Engine torque, engine
RPM, and estimated propeller efficiency can be used to predict the
thrust associated with this configuration (from the propeller and engine
exhaust).

Entry Rate Correction

Although the entry rate has been shown to have little impact on the load
factor corrected lift coefficient, it does have a significant impact on the non–
load factor corrected lift coefficient. Essentially, the faster one would slow
down the airplane, the lower the speed (inertia effects) would be at the
time of the stall identification.
For the 1-g method, AC 25-7D states: “Because CLMAX is relatively
insensitive to stall entry rate, a rigorous investigation of entry rate
effects should not be necessary.” Some manufacturers still correct for
entry rate even if this is often less than 1/10 of a knot. Some habits
are hard to let go.
For airplanes performing stall testing without correcting for the load
factor, it is suggested that a spread of entry rates for a given configura-
tion be varied from about 0.5 kt/s to 1.5 kt/s to be able to develop an
entry rate correction trend (see Fig. 7.41). At least two test points per
target entry rate should be done to ensure data scatter does not skew
the trend.
For airplanes that have a stall ID defined by a prestall stick pusher, the
entry rate has little impact on CLmax because the airplane is not stalled at
pusher activation. Thus, entry rate investigation for the purpose of com-
puting a stall speed is not required. That is not to say that dynamic
 CHAPTER 7 Stall Testing 283

entry into stall is not needed for stall characteristics requirements but
simply that the stall speed will not be measurably impacted by the
entry rate. Review of the stall characteristics data will give the manufac-
turer a better picture of whether entry rate investigation is required at
the time of performing stall speed testing.
For a practical example, assume the airplane of Fig. 7.41 had a wing
loading of 75 lb/ft2 at the time of the stall. Using the curve-fitted data, that
airplane would have a stall speed VS1g of 109.33 kt for a 1-kt/s entry rate.
If the entry rate had been 10% higher (1.1 kt/s), the computed stall speed
would have been 109.26 kt, a difference of less than 0.1 kt. At 1.5 kt/s, the
resulting difference is less than 0.4 kt. So, per the guidance material, if
testing is performed around 1 kt/s, the test scatter will be sufficiently close
to the 1 kt/s to not warrant entry rate investigation.

Altitude and Test Weight Effects

Altitude and weight will impact the airplane stall speed by introducing
some Mach effects, even at these low speeds, with this effect (increasing alti-
tude and weight) reducing the available CLmax . The applicant must either
address this effect or determine a conservative stall speed for the airplane.
If the applicant elects to schedule the AFM stall speed as a function of alti-
tude, then sufficient testing must be done to determine the effect of altitude.
If, on the other hand, the stall speed is not defined in terms of altitude, the

2.05

2
Not load factor corrected
y = 0.1078x + 1.8542

1.95
CLmax

1.9
Load factor corrected

1.85
y = 0.0252x + 1.8282

1.8

1.75
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Entry rate (kt/s)

Fig. 7.41 Entry rate impact on the measured CLmax .

284 Operational Aircraft Performance and Flight Test Practices

guidance material of FAA AC 25-7D recommends that the stall speed testing
be conducted at a nominal altitude of no less than 1500 ft above the
maximum altitude for AFM takeoff and landing performance (maximum
certified field altitude). FAA AC 25-7D and AC 23-8C also recommend
plotting CLmax against aircraft weight, noting that there is an expected
decrease in CLmax as the weight is increased. Figure 7.42 shows typical
trends of CLmax with weight for a given altitude (Fig. 7.42a) vs W/d
(Fig. 7.42b). A convenient way to include altitude effects for AFM is to
plot CLmax vs Mach number, and then include W/d effects. In terms of
Mach number and W/d,
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u  
u W
u
u d
Mstall ¼ u
t1 ð7:11Þ
gpSL SCLmax
2
The slope of the curve shown on both graphs in Fig. 7.42 will be flatter for
an airplane with flaps down if that airplane has a hard leading edge (no slats
or other moveable device), plain flaps, and a rigid wing. The slope will be
larger for an airplane with a slatted leading edge, flaps with slots, and a flex-
ible wing. (Wing deflection tends to change the wing twist.) A flaps-down
configuration typically has a maximum altitude of 15,000 to 20,000 ft, as
selected by the manufacturer, which is also the typical test altitude for stall
testing for most large planes (for flight test safety—provides altitude
margin for recovery from stall), so the option to test at a single altitude is a
possibility. We note that the flaps-up flight envelope goes much higher in
most cases, so programming stall with altitude is a must in this case.
The lowest altitude in flight testing is dictated by safety concerns whereas
the lowest test weight is dictated by the amount of test equipment that is
carried in the aircraft. (The minimum test weight usually is much larger
than a production aircraft minimum flight weight, so the minimum test
W/d will be much larger than the minimum production value.) In fact, the

a) Test at one altitude b) Test at multiple altitude

(max airport altitude + 1500 ft) and weight combinations
Light weight lower alt

Light weight Heavy lower altitude

Max
CLmax
AFM Heavy higher altitude
CLmax
Heavy weight
Light weight higher alt

Airspeed (KCAS) Mach

Fig. 7.42 Altitude and weight effects on CLmax .

 CHAPTER 7 Stall Testing 285

test minimum W/d may even be larger than the W/d corresponding to the
maximum takeoff weight at sea level. Once all stall characteristics testing is
done and the stall speed testing above the minimum test altitude
(maximum airport altitude plus 1500 ft) is accomplished, the manufacturer
may elect to perform additional stall speed testing at a lower altitude to
extend the slope of the CLmax vs Mach to a lower value of W/d. In this
case, the airplane should be configured in its lightest test weight possible,
and testing should be executed as low as test safety will allow. (This author
suggests no less than 8000 ft above ground for most airplanes.)
Because of the trend of increasing CLmax with decreasing W/d (or Mach),
and because the data collected must be expanded to lower W/d for AFM pur-
poses, the simple extrapolation of CLmax with decreasing W/d could result in
higher CLmax than demonstrated in flight testing. The guidance material only
allows the expansion to the highest CLmax demonstrated, as shown in
Fig. 7.42b.

Production Tolerances
Once certification testing is done, the manufacturer will start producing
the airplane as certified and of course start getting revenue. Each plane pro-
duced will undergo production flight testing (two to six flights) during which
it will be checked against a reference to ensure all is working as expected and
certified. One part of the production flight test plan will involve production
stalls to validate the build of the airplane and calibration of the onboard
instrumentation, which future crew will rely on for safe operation.
Production stalls are interesting in that the crew will now rely exclusively
on production instrumentation, and we have seen previously that the air-
speed may not even decrease beyond stall warning. As well, the production
airplane will have no means to adjust the CG in flight and may not have
full forward CG at the time of the test. With this in mind, the performance
engineer must provide some guidance to create the production flight test
plan and execute the stalls. Forward CG is favored, and the crew will be
asked to provide the load sheet (the weight and balance sheet of the pro-
duction plane) as part of the test results for postflight data reduction. The
production test plan will ask for a nominal 1-kt/s entry rate, but the crew
will have no experimental gauge to calibrate it, and the entry rate will
remain unmeasured. Most larger planes have a means to measure accelera-
tion with respect to the airplane principal axes but no experimental angle
of attack; the performance engineer would then need to derive the flight
path angle at the time of the stall by other means, but that data may not
be available to download from the production system. The crew will be
asked to record the airspeed at the time of stall warning and stall identifi-
cation, another source of scatter.
Finally, although built to strict standards, each airplane produced will
have a small deviation from the airplane used for certification. With all this
286 Operational Aircraft Performance and Flight Test Practices

in mind, the performance engineer must provide expected indicated stall

warning and stall identification speeds for the production stalls. Speed toler-
ances are provided (typically +3 KIAS) for rapid check during the flight,
knowing that with some data reduction by the performance engineer, the
deviation for expected numbers will come down. The crew will check the
margin from VSW and VSID are as expected and that the handling is also
acceptable (no more than 20-deg roll, no pitch up or pitch force reversal).

Data Expansion
This phase takes the stall speed model created and applies regulatory
margins on top. We discussed these during the data reduction explanation;
here we just remind the reader that the data expansion phase can include:
• CG correction from the reference used for data reduction to the
production forward CG for a given weight (if this is the CG that
provides the highest stall speed); and
• Altitude correction (W/d) if the manufacturer elects to do so

Presenting the Information to the Flight Crew

The data having been reduced to an approved model by the certification
agency, the applicant must now present the data to the flight crew in a
manner that will be easy to understand. The reference stall speed of the air-
plane must be provided to the crew in the AFM, per §25.1587(b)(2):
§25.1587 Performance information
(b) Each Airplane Flight Manual must contain the performance information
computed under the applicable provisions of this part (including
§§25.115, 25.123, and 25.125 for the weights, altitudes, temperatures,
wind components, and runway gradients, as applicable) within the
operational limits of the airplane, and must contain the following:
(2) VSR determined in accordance with §25.103.

It is unlikely that the crew will actually consult the AFM in flight to determine
the minimum flight speed; instead, they are more likely to rely on information
displayed in the cockpit and use operational speeds such as takeoff or landing
speeds. This noncriticality in the requirements allows the manufacturer to be
flexible with the format of the stall speed in the AFM.
One way to present the stall speed is in the form of a chart that accounts for
weight and altitude effects (if the applicant’s model accounts for such effects);
see Fig. 7.43. Depending on the resolution of the grid provided and on the
number of altitude lines provided, the precision of the airspeed derived from
these charts will vary from 0.5 kt to 1 kt. This format would need to be repeated
for every airplane configuration (at least every flap setting).
Another way to present the data is in a table format (see Fig. 7.44). This
format provides a more precise number at a specific weight/altitude
 CHAPTER 7 Stall Testing 287

110 Pressure
altitude (ft)
20K
105
15K
10K
100 5K
SL
VSR (KCAS)

95

90

85

80
10,000 10,500 11,000 11,500 12,000 12,500 13,000 13,500 14,000
Aircraft weight (lb)

Fig. 7.43 AFM stall speed chart format.

combination but leaves the crew to interpolate four ways for an altitude and
weight combination that is not provided. Alternatively, it can be rec-
ommended to the flight crew to use a conservative number by selecting
the bottom right value of a four-way integration.

Exercise
(1) An airplane is performing a stall test and the following is recorded. The
airplane weight at the time of the test is 110,000 lb, and the reference

Weight Pressure altitude (ft)

(lb) 0 5000 10,000 15,000 20,000
10,000 84.6 84.6 84.7 85.4 86.7
10,500 86.7 86.7 86.8 87.8 89.2
11,000 88.7 88.8 88.9 90.1 91.8
11,500 90.7 90.7 91.1 92.5 94.4
12,000 92.6 92.6 93.2 94.6 96.8
12,540 94.5 94.5 95.4 96.9 99.4
13,000 96.1 96.1 97.2 98.8 101.6
13,500 97.8 97.9 99.2 100.9 103.9
14,000 99.4 99.8 101.1 103.1 106.2

Fig. 7.44 Table format presentation of AFM stall speeds.

288 Operational Aircraft Performance and Flight Test Practices

wing area is 1200 ft2 . Is this a good stall? Why? If a good stall, what is the
stall speed (VSR )? Is the stall warning setting appropriate?

KIAS (kt) KCAS (kt) Stall

warning Pusher
120
115
KCAS (kt)
KIAS (kt)

110
105
100
95

Roll (deg) Aileron (deg)

40 10
30

Aileron (deg)
5
Roll (deg)

20
10 0
0
–5
–10
–20 –10

Elevator (deg) Stick force (lb)

80
Stick force (lb)
Elevator (deg)

40

–40

Pitch (deg) Boom AoA (deg)

30
Boom AoA (deg)

20
Pitch (deg)

10
0
–10
–20

Nzw (g) CL
2.0 2.0
1.5
Nzw (g)

CL

1.0 1.5
0.5
0.0 1.0
90 95 100 105 110 115 120
Time (s)

References
[1] Borer, N. K., “Development of a New Departure Aversion Standard for Light Aircraft,”
AIAA-2017-3438, 17th AIAA Aviation Technology, Integration, and Operations
Conference, Denver, CO, 2017, https://ntrs.nasa.gov/search.jsp?R ¼ 20170005881
[retrieved 20 Dec. 2017].
[2] Hartman, E. P.Biermann, D. “The Aerodynamic Characteristics of Full-Scale Propellers
Having 2, 3, and 4 Blades of Clark Y and R.A.F. 6 Airfoil Sections,” NACA TR640, 1938.
[3] “The DC-9 and the Deep Stall,” Flight International, Vol. 442, 25 March 1965.
 Chapter 8 Thrust and Drag
Modeling

Chapter Objective
The objective of this module is to define basic models for both thrust and drag.
We will find that the airplane’s drag has good general trends for which we can
readily define typical change with speed, altitude, and weight. Then we turn
our focus to the production of thrust, and we see how the combination of
various systems generate the required propulsive force. When it comes time
to combine the thrust and the drag models, the results may not always be
as expected due to the interference of the airframe in the production of
thrust and vice versa, so we introduce the concept of excess thrust
management.

Drag Model

A
n airplane moving through an air mass will be subjected to an aero-
dynamic force. In Chapter 6, we discussed the force component per-
pendicular to the flight path vector, the lift. In this chapter, we will
cover the force that is parallel to the flight path vector, the drag force (see
Fig. 8.1). The drag is a dissipative force that tends to reduce the airplane’s
energy. It can be written in the same general format as the lift equation,
that is,
1 1
D ¼ rSL s V 2 S CD ¼ g pSL d M2 S CD ð8:1Þ
2 2
The drag comes from two main sources:
1. Friction drag due to viscosity: A tangential force to the local surface of the
airplane
2. Pressure drag: A normal force to the surface of the airplane
These drag sources combine in different ways to produce the total drag.
The drag tree shown in Fig. 8.2 provides a convenient way to analyze the
various forms of drag and is this author’s preferred approach to breaking
down the total drag. As a performance problem is tackled, one always

289
290 Operational Aircraft Performance and Flight Test Practices

V D

Fig. 8.1 Lift and drag vectors.

needs to consider what could impact the drag model generated and deter-
mine the limits of that model.
For performance analysis throughout this book, the drag will be
expressed in terms of induced drag (drag due to the generation of lift) and
parasite drag (drag not due to lift). We may also elect to add a separate

Total drag
Drag not due to lift Drag due to lift

Parasite drag Induced drag

Interference drag

Excrescence
drag

Wave Vortex Trim

drag drag drag

Profile drag

Wetted area Shape, volume

Skin friction
Form drag
drag

Viscosity Pressure

Fig. 8.2 Drag tree.

 CHAPTER 8 Thrust and Drag Modeling 291

wave drag element to the discussion when discussing drag rise, knowing that
this drag parameter affects both the parasite drag and induced drag.

Drag Sources
This section does not provide an exhaustive review of the various sources
of drag; other references, such as [1], are completely dedicated to this subject.
We cover enough about drag, however, to create a model and support the
performance analysis discussions. It will allow us to perform sensitivity analy-
sis with the airplane performance model.

Skin Friction Drag

The skin friction drag of the airplane is proportional to its wetted area, the
total airplane surface in contact with the moving air. This drag component is
the dominant component of the parasite drag (drag not due to lift). A con-
venient way to estimate this component uses the following equation, parasite
drag coefficient (CDo ):
f
CDo ¼ ð8:1Þ
S
where f represents the equivalent parasite area of the airplane, and S is the
reference wing area. Roskam [2] shows that the equivalent parasite area cor-
relates very well with the airplane’s wetted area via Eq. (8.2)
logð f Þ ¼ a þ logðSwet Þ ð8:2Þ
Coefficient a of Eq. (8.2) correlates to the airplane’s equivalent skin fric-
tion coefficient cf , which is itself a measure of the airplane’s smoothness. The
coefficient typically varies from 0.0020 (a very streamlined airplane) to over
0.0150. Roskam [2] provides many examples of airplanes’ estimated equival-
ent parasite area. With the use of cf , Eq. (8.2) can be written as
Swet
CDo ¼ cf ð8:3Þ
S
Excrescence Drag
Excrescence drag or roughness drag is a grouping of drag components
generated from items not part of the baseline aero design. This can include
antennas, steps and gaps, and air flow leaks [outflow from pressurization
system, air leaks around doors, auxiliary power unit (APU) exhaust flow,
APU door]. One way to visualize excrescence drag is to compare the differ-
ence between what one would find on a wind tunnel model and on the full-
size airplane (Fig. 8.3).
Excrescence drag typically affects the parasitic drag most (i.e., drag not
due to lift, thus higher impact at higher speeds). It can be minimized by:
• Minimizing the number of excrescences and their sizes
292 Operational Aircraft Performance and Flight Test Practices

Fig. 8.3 Wind tunnel vs airplane, Boeing B-17.

• Moving the roughness towards the tail of the airplane

• Having the roughness on the bottom of the wing rather than the top of
the wing
• Reversing for the horizontal tail—on the top surface rather than the
lower surface
This author has seen many airplanes first developed as a regional jet or a
business jet that were transformed (missionized) by the addition of external
shapes and multiple antennas to serve a new mission for which they were not
previously intended. Those additional shapes can have an important impact
on the airplane’s baseline performance.
We also had a chance to discuss with a colleague and ex-student from the
FAA Anchorage Aircraft Certification Office (ACO) the work that goes on to
develop a Supplemental Type Certificate (STC) to allow bush planes on floats
to carry external loads such as fishing boats (see Fig. 8.4) or four-wheel
all-terrain-vehicles attached to the floats. This is an interesting challenge!

Vortex Drag
Vortex drag forms the majority of the drag due to lift component of the
total drag. There is a necessary transfer of momentum to the air to generate
lift; this transfer of momentum (proportional to a mass times the change of
airspeed perpendicular to the incoming air, that is, the induced speed dVv ) is
distributed along the span of the wing (see Fig. 8.5a). For a wing of infinite
span, this induced speed tends towards zero (i.e., dVv  0). As we have
seen in Chapter 6, this induced speed is the downwash v.

Fig. 8.4 A de Havilland DHC-2 Beaver with and without external load (canoe) on floats.
 CHAPTER 8 Thrust and Drag Modeling 293

a) b)
b

L
L
v
dVv

Side view Front view

Fig. 8.5 Generation of lift through transfer of momentum

For a wing of finite span, we define a stream tube of air proportional to

the wing span b. For a given lift force generated by this wing of finite span,
the downwash v produced will need to increase as the span is reduced;
this increase will be proportional to the span squared (i.e., the strength of
the downwash must increase as fast as the reduction in the span squared).
 p 2
L ¼ ṁ DV ¼ r V b v ð8:4Þ
4

The average induced angle ai is half the downwash angle (e  v/V for
small angles)

1 v
ai ¼ ð8:5Þ
2 V

This induced angle tilts the aerodynamic force vector, creating two com-
ponents (when broken down into the wind axis). One is the lift as described
previously; the other component is parallel to the incoming flow and is there-
fore a drag component (see Fig. 8.6).

L F


V i Di


Fig. 8.6 Induced angle.

294 Operational Aircraft Performance and Flight Test Practices

Mathematically, we can see that

L
F¼ ð8:6Þ
cosðai Þ

Di ¼ F sinðai Þ ¼ L tanðai Þ ð8:7Þ

With the assumption of small angle again, tan(ai )  aI , and from
Chapter 6, we know that
CL
ai ¼ ð8:8Þ
p AR
where AR is the wing aspect ratio, the ratio of the wingspan square to the
wing area S, or b 2 /S. Thus, combining these equations, we find that the
induced drag Di is equal to

L2 W2 W2
Di ¼ ¼ ¼ ð8:9Þ
p q b2 p q b2 p AR q S
The second part of Eq. (8.9) (assumption L ¼ W) being true in steady level
flight. The induced drag is thus proportional to the square of the span loading
of the plane (W/b)2 ; the larger the wingspan, the smaller the induced drag.
From this last equation we extract the induced drag coefficient CDi .

CDi ¼ K CL2 ð8:10Þ

where
1
K¼ ð8:11Þ
p AR e
Here we introduce the Oswald coefficient e. The actual spanwise lift dis-
tribution has a strong impact on the induced drag coefficient. The theory
indicates that an elliptical lift distribution would result in the lowest
induced drag for a given size S and aspect ratio AR. Not all wings can have
such a distribution; the presence of a fuselage, engine nacelles, propeller slip-
stream, or wing flap deflection can significantly impact the baseline spanwise
lift distribution. The Oswald coefficient e allows the induced drag coefficient
[Eq. (8.10)] to keep the same general format while providing a means to
adjust the drag computed for a given amount of lift produced. The Oswald
coefficient will take on a value of 1.0 for an ideal elliptic lift distribution
and a smaller value when the lift distribution deviates from ideal.
This represents an acceptable mathematical model for the vortex drag
(the induced drag to generate the lift) of the airplane. Another way to visual-
ize the impact of generating the lift is to consider that there is a difference of
pressure from below the wing to the top of the wing. This difference of
pressure must be equal at the wing tip (pressure continuity), and thus
there will be a spanwise pressure distribution that will generate a flow in
 CHAPTER 8 Thrust and Drag Modeling 295

that direction. The lower surface air will tend to go towards the wing tip
whereas the upper surface air will tend to go towards the wing root, creating
a vortex sheet at the trailing edge of the wing.
The vortex sheet is unstable and strongest near the wing tip (for flaps-up
conditions). It will tend to roll up into two large vortices. This roll-up motion
requires energy that is extracted from the airplane total energy; this is
lift-induced drag.

Trim Drag
The lift and weight vectors are not necessarily aligned with each other and
that some cancellation (trim) of the resulting pitching moment is usually pro-
vided by the horizontal tail to maintain steady level flight. We also know that
the airplane’s aerodynamic pitching moment MWB and the thrust vector
location with respect to the vertical center of gravity can have a significant
impact on the trim required by the tail (see Fig. 8.7).
The “extra” lift required by the wing to compensate for the download of
the horizontal tail to trim will generate a vortex drag. The down lift of the
horizontal tail will also generate a vortex drag. The sum of the “additional”
vortex drag is part of the trim drag—the drag generated specifically to trim
the airplane.
In practice, all surfaces generating a trim lift will generate a vortex drag,
the strongest generally being the one for trimming the airplane longitudin-
ally. If that vortex drag is not generated for the purpose of maintaining the
flight path, it is then rolled into trim drag. Consider the case of an engine-out
(failed) flight [one engine inoperative (OEI)]. The asymmetry created by the
loss of thrust on one side of the airplane centerline (see Fig. 8.8) will require a
force generated by the vertical tail (a side lift) to balance out the yawing
moment generated. This side force typically being above the center of
gravity, a rolling moment will be generated from the vertical tail that will
need to be countered by the deflection of ailerons and/or roll assist spoilers.
All of these forces generate extra drag not normally present during

CL, CM

CL
L = W + Lt
MWB
Trim point A.C.

D Lt
 T
CM W

Fig. 8.7 Trim of moments.

296 Operational Aircraft Performance and Flight Test Practices

FR

T
Fa

FR
D Fa

Fig. 8.8 Asymmetric forces generated during OEI condition.

symmetric flight; they are a form of trim drag. This particular case will be
explored further in Chapter 13.
Trim drag is not normally part of the performance equation directly;
rather, one typically uses the trimmed drag polar of the airplane (the drag
equation with a specific center of gravity location) to perform the analysis.
For certification performance analysis, the regulations of Part 25 (transport
category airplanes) and Part 23 (small airplanes) specify that the center of
gravity (CG) “shall be at the most adverse location” for the condition for
which the compliance to the specific regulation must be shown; it is up to
the applicant to demonstrate that is the case. For cruise performance analysis
(not airplane certification related), the manufacturer has more flexibility, and
the location of the CG used to define the drag equation will typically be one
that represents an average condition for the flight.

Wave Drag
Wave drag Dw is the term used to capture the increase in drag at higher
Mach numbers when the airflow around the airplane goes transonic and
starts producing shock waves. As the air goes through a shock wave, the
result will be a loss of momentum, which is an additional source of drag
and a deviation from the baseline drag model (Fig. 8.9).
The wing is typically the main source of wave drag in the early part of the
transonic zone as the flow is accelerated over the top surface to create lift and
thus goes supersonic sooner (see Fig. 8.9). One of the impacts of a shock wave
is also a rapid pressure rise from just before to just after the shock. If this
shock forms over a surface (such as the wing), this will result in some flow
separation behind the shock. This flow separation will lead to some loss of
lift (for a wing), which must be compensated for by an increase in angle of
attack to maintain level flight. In turn, this higher angle of attack will intensify
the strength of the shock and lead to more losses. As the airplane continues
 CHAPTER 8 Thrust and Drag Modeling 297

Shock
M>1 M<1 Shock-induced
flow separation
M<1

Transonic flow

Supersonic flow

Fig. 8.9 Wave drag and shock waves. Photo source: NASA.

to accelerate towards Mach 1.0 (sonic speed), other parts of the airplane will
also develop shock waves and increase the drag of the airplane.
From a drag point of view, the freestream Mach number that corresponds
to the first point on the surface to reach Mach 1.0 is call the critical Mach
number Mcrit . As the Mach number keeps increasing, weak shock waves
will form initially and will gradually change the general behavior of the
drag for a given lift coefficient (see Fig. 8.10). At a certain Mach, the drag
will have deviated enough from the general trend, and this will be defined
as the drag rise Mach number MDR . The drag deviation from the trend
where one considers having reached a drag rise condition differs from

0.07
CL
0.06 0.5

0.4
0.05

0.3
0.04
0.2
CD

0.1
0.03

0.02

0.01

0
0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1
Mach

Fig. 8.10 Drag rise.

298 Operational Aircraft Performance and Flight Test Practices

manufacturer to manufacturer, but a generally accepted value is 20 drag

count (DCD ¼ 0.0020). One will also notice from Fig. 8.10 that the value of
MDR tends to decrease with increasing lift coefficient. This is typical for a
design in which most of the wave drag comes from the wing; the higher
the lift coefficient, the harder the wing is working, producing a larger
pressure gradient and stronger shock sooner (at lower Mach).

Proposed Drag Model for Basic Performance Assessment

The trimmed drag coefficient variation with variation of lift coefficient
that will be used for airplane performance analysis will have the following
general shape:

CD ¼ CDo þ KCL2 ð8:12Þ

where CDo is the parasite drag coefficient (per the drag tree, Fig. 8.2) and KC2L
is the induced drag coefficient (or drag due to lift). In this form, minimum
drag coefficient occurs at zero lift coefficient. This general shape is referred
to as the airplane’s drag polar. Figure 8.11 illustrates the behavior of the
drag coefficient vs lift coefficient for CDo ¼ 0.02 and K ¼ 0.05.
Of course, this is not always the case; the minimum drag could occur at a
nonzero lift coefficient value. A more general form of the drag polar
equation is
 2
CD ¼ CDmin þ K CL  CLo ð8:13Þ
where the lift coefficient CLo for minimum drag coefficient CDmin is nonzero.

1.2

0.8
CL

0.6

0.4

0.2

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
CD

Fig. 8.11 Example of drag polar.

 CHAPTER 8 Thrust and Drag Modeling 299

Drag coeff

CDo

MDR Mach

Fig. 8.12 Drag coefficient change beyond drag rise Mach number.

For this book, the drag model used will be based on the following
assumptions:
• The airplane is a point mass that is stable and controllable (3 deg of
freedom model) and upon which four forces act (lift, drag, thrust, and
weight).
• The drag polar will have the form specified by Eq. (8.12). We will also
assume that CDo and K are constant for a flight Mach number that is less
than the drag rise Mach number (M , MDR ).
• As a first assumption, one can assume that MDR will not change with CL
for initial analysis.
• Beyond MDR , both coefficients increase; for some airplanes, that increase
can be very sharp (see Fig. 8.12).
• The increase in both coefficients can be captured in a wave drag
coefficient CDw added to Eq. (8.12) or by varying the coefficients of
Eq. (8.12).
• We only cover the drag up to Mach 1.0 for this book, and we expect the
drag to increase from MDR to Mach 1.0.
• On the low-speed side, one would expect an increase in drag coefficient as
the airplane nears stall and the airflow starts separating from the airplane,
and that the drag would start increasing [deviating from the model of
Eq. (8.12)]. For initial analysis, we can assume that the model is
unimpacted.

Steady Level Flight Assumption

Using the drag model just developed, we have our first look at the forces
involved to sustain level flight (see Fig. 8.13). To do this, we will use the fol-
lowing assumptions:
300 Operational Aircraft Performance and Flight Test Practices

T
D
T Horizon

Fig. 8.13 Level flight, equilibrium of forces.

• No acceleration (constant airspeed and ground speed, therefore no wind

change).
• Flight along the local horizontal (constant geometric and pressure
altitudes, therefore constant atmospheric conditions).
• The drag model that will be developed will neglect the effects of the
airplane’s angle of attack on the thrust line and will be for a trimmed
airplane.
An equilibrium of forces exists, where:
• The airplane weight W is counteracted by the lift L to maintain the altitude,
both forces acting along the local vertical.
• The airplane drag D must be counteracted by the thrust T to maintain
airspeed.
L ¼W ð8:14Þ

T ¼D ð8:15Þ
leading to
T D
¼ ð8:16Þ
W L

Aerodynamic Efficiency
The aerodynamic efficiency (E) of the airplane under a given flight con-
dition (airspeed, altitude) and configuration (flap, gear, etc.) is defined as
the lift-to-drag ratio
L
E¼ ð8:17Þ
D
 CHAPTER 8 Thrust and Drag Modeling 301

Remembering the definition of the lift and drag equations [Eqs. (6.7) and
(8.1)], one gets
CL
E¼ ð8:18Þ
CD
And from the steady level flight conditions, Eq. (8.16), one gets
T 1 W
¼ ! Tr ¼ ð8:19Þ
W ðL=DÞ E
The drag D and the thrust required Tr to maintain a given flight condition
are synonymous and will be used alternatively. Equation (8.19) clearly shows
the impact of the airplane’s weight on the thrust required to sustain flight and
ultimately on the energy consumption to do so.

Minimum Drag Condition

The minimum thrust required for a given weight and configuration will
occur when the airplane is flying at its maximum lift-to-drag ratio condition
(or maximum aerodynamic efficiency, Em ). This condition can be determined
by differentiating the lift-to-drag equation with the lift coefficient
  " #
@E @ CL
¼ ¼0 ð8:20Þ
@CL Em @CL CDo þ K CL2
Em

We can then solve Eq. (8.20). The resulting lift and drag coefficients and
lift-to-drag ratio are
rffiffiffiffiffiffiffiffi
CDo
CLEm ¼ ð8:21Þ
K
CDEm ¼ 2 CDo ð8:22Þ
1
Em ¼ pffiffiffiffiffiffiffiffiffiffiffiffi ð8:23Þ
2 CDo K
The maximum lift-to-drag ratio therefore occurs at a condition where
the total drag coefficient CD is equal to twice the minimum drag coefficient
CDo . The maximum lift-to-drag ratio can also be determined graphically by
drawing a line from the origin to a point tangent to the polar, irrespective of
the actual shape of the drag polar. Figure 8.14 shows examples of Eqs. (8.12)
and (8.13) drag polars.
One can do this mathematical exercise to derive Eq. (8.23) from
Eq. (8.12), but this time for Eq. (8.13). One gets
1
Em ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð8:24Þ
 2
2 K CDmin þ K CLo  K CLo
302 Operational Aircraft Performance and Flight Test Practices

1
CD = 0.025 + 0.1136 C2L
0.9
0.8
0.7
0.6
0.5
CL

CL
0.4 Em = = 9.382
CD
0.3
0.2
0.1
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
CD

1
CD = 0.025 + 0.08 (CL – 0.2)2
0.9
0.8
0.7
0.6
0.5 CL
CL

Em = = 15.87
0.4 CD

0.3
0.2
0.1
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
CD

Fig. 8.14 Finding Em graphically.

Going back to the basic model of Eq. (8.12), the airspeed at which the
maximum lift-to-drag ratio occurs (VEm ) can be found by inserting the
value of the lift coefficient for max L/D [Eq. (8.21)] into the lift equation
for level flight
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffi
1 2 2W 4 K
L ¼ rSL s VEm S CLEm ¼ W ! VEm ¼ ð8:25Þ
2 rSL s S CDo
 CHAPTER 8 Thrust and Drag Modeling 303

Returning to our original question, the minimum thrust required (or

minimum drag, Dmin ) to maintain a given flight condition (combination of
weight, configuration, airspeed, altitude, etc.) is
W
Dmin ¼ ð8:26Þ
Em
The minimum drag therefore occurs at a condition where the lift-to-drag
ratio is maximum, and it is directly proportional to the airplane’s weight.

Drag Variation with Airspeed, Drag Coefficients, and Weight

We now break down the drag equation into its main components, a para-
site drag component and an induced drag component, so we can better
observe the behavior of each with airspeed and airplane weight
1 1
D¼ rSL s V 2 S CDo þ rSL s V 2 S K CL2 ð8:27Þ
2 2
When combined with the lift equation, and we maintain the steady level
flight assumption (L ¼ W ), we get:
1 K W2
D¼ rSL s V 2 S CDo þ ð8:28Þ
2 1
r s V2 S
2 SL
We see that the parasite drag increases proportionally to the square of the
airspeed, and we therefore expect it to be predominant at high speed. We also
note that the induced drag decreases with increasing airspeed (proportional
to 1/V 2 ) and is therefore dominant at low airspeed. Figure 8.15 provides a

5000
Note: High speed (Mach) effects neglected for this example
4500
4000
3500
3000
Drag (lb)

2500
Total
2000
1500
Induced Parasite
1000
500
0
0 100 200 300 400 500 600 700 800
Airspeed (ft/s)

Fig. 8.15 Example of the variation of drag with airspeed.

304 Operational Aircraft Performance and Flight Test Practices

visual example of the expected behavior of each component and of the

total drag.
We can also observe in Fig. 8.15 that the minimum drag condition occurs
when each of its components are equal (i.e., parasite drag equals induced
drag).
The graph in Fig. 8.16 shows an example of how the lift-to-drag ratio of
an airplane typically varies with airspeed. This same graph also includes the
drag for this model airplane. Note, as expected, how the maximum value of
the L/D ratio (aerodynamic efficiency, E) occurs at the same airspeed as the
minimum drag condition.
As stated previously, to maintain a steady state airspeed at a constant alti-
tude, the thrust generated by the engines must be just enough to counteract
the airplane’s drag [Eq. (8.15)]. The thrust available from the engine(s), as we
will detail later, is a function of several factors, one of which is under the
control of the pilot via the throttle. For a given throttle position, the
engines will provide a given percentage of the maximum thrust available
for the flight conditions at hand.
Figure 8.17 shows the impact of varying the throttle position on the poss-
ible equilibrium airspeeds (where T = D). In this figure, we elected to assume
the thrust was independent of airspeed and that a given throttle setting pro-
vides an available thrust (horizontal line). Note that for each throttle position,
two possible equilibrium points exists. On the “front” side of the drag curve
(airspeeds higher than the minimum drag configuration), an increase in
thrust will lead to an increase in airspeed. On the “back” side of the drag

4000 16

3500 14
E
3000 12

2500 10
Drag (lb)

L/D

2000 8

1500 6

1000 D 4

500 2

0 0
0 100 200 300 400 500 600 700
Airspeed (ft/s)

Fig. 8.16 Evolution of drag and lift-to-drag with airspeed.

 CHAPTER 8 Thrust and Drag Modeling 305

2000
Throttle
1800 increase
1600
1400
Drag, thrust (lb)

1200
1000
800
600
400
200
Vmin1 Vmax1
0
0 100 200 300 400 500 600
Airspeed (ft/s)

Fig. 8.17 Equilibrium conditions vs throttle position.

curve (airspeeds lower than minimum drag airspeed), an increase in thrust is

required to maintain a lower airspeed. The front side of the drag curve is
considered the positive speed stability side of the curve where, when the air-
plane is disturbed/displaced from the equilibrium speed, it will naturally tend
to come back to that speed. The back side of the drag curve is the slow speed
domain; we will expand on this later in this chapter.
We can determine this airspeed limited by available thrust (Ta ) from
the Ta ¼ D relationship
1  
Ta ¼ rSL s V 2 S CDo þ K CL2 ð8:29Þ
2
Knowing that
W
CL ¼ ð8:30Þ
1
r s V2 S
2 SL
we expand Eq. (8.29) to
1 K W2
rSL s V 4 S CDo  Ta V 2 þ ¼0 ð8:31Þ
2 1
rSL s S
2
We can solve this for airspeed to find
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
" #
u
u T 4 K C
V ¼t
a D o
1+ ð8:32Þ
rSL s S CDo ðTa =W Þ2
306 Operational Aircraft Performance and Flight Test Practices

Note the + sign in Eq. (8.32). Using the positive sign will determine Vmax ,
the maximum airspeed limited by available thrust in level flight; using the
negative sign will determine Vmin , again a speed limited by available thrust.
This latest speed may be smaller than the airplane stall speed, a speed
limited by the capability of the airplane to generate lift.
The next item to be investigated for its impact on drag is the variation of
the parasite drag coefficient CDo . The baseline parasite drag coefficient of an
airplane is defined at the time of initial design of the airplane when the shape
and configurations (landing gear shape and retraction capability, flaps shapes
and deflection angles, etc.) are determined. Then, throughout the life of a
given design, the parasite drag coefficient tends to increase (often with the
addition of antennas or simply change in surface roughness such as peeling
paint, repairs, etc.). By inspection of Eqs. (8.27) and (8.32), we see that a
change in the parasite drag coefficient will have a dominant effect at high
speed. The effects of increasing CDo are shown in Fig. 8.18.
By observation of Fig. 8.18, we note the following:

• The minimum drag airspeed decreases with increasing CDo .

• The minimum drag increases with increasing CDo .
• As expected, the impact is more significant at high speed than low speed.

We now consider the impact of weight W and the induced drag factor K
on the drag behavior. (We consider both together here because they have
similar effects.) Note from Eq. (8.28) that the impact of the weight change
on drag will be proportional to the square of the weight, whereas a change

45,000
Note: High speed (Mach) effects neglected for this example
40,000

35,000

30,000
Increasing CDo
Drag (lb)

25,000

20,000

15,000

10,000

5000

0
0 100 200 300 400 500 600 700
Airspeed (ft/s)

Fig. 8.18 Impact of CDo variation on the drag curve.

 CHAPTER 8 Thrust and Drag Modeling 307

6000

5000

4000
Drag (lb)

3000

Increasing K
2000

1000

0
0 100 200 300 400 500 600 700
Airspeed (ft/s)

6000

5000

4000
Drag (lb)

3000 Increasing W

2000

1000

0
0 100 200 300 400 500 600 700
Airspeed (ft/s)

Fig. 8.19 Weight impact on drag vs airspeed.

in K has a proportional impact. To illustrate this point, we doubled the value

of K and of the weight from a reference drag curve in Fig. 8.19—the magni-
tude of the change can be clearly seen for each variable.
We also note the following:

• The minimum drag airspeed increases as the weight or K increases.

• The minimum drag value increases with weight and K.
308 Operational Aircraft Performance and Flight Test Practices

Furthermore, the change in drag at high speed is relatively low compared

to the change in drag at low speed with weight and K change.
One should expect that some change in airplane configuration will
impact both CDo and K. These could include:
• A flap and/or slat deflection.
• Deployment of wing-mounted speedbrakes.
• Change in engine power, especially for wing-mounted
propeller-driven airplanes.
• Even a wing-mounted landing gear can impact both coefficients.
A change in altitude will impact the parasite and induced drag differently.
Note that increasing altitude will decrease the parasitic drag (see Fig. 8.20)
while increasing the induced drag for a fixed weight and for a given true air-
speed. Note for Fig. 8.20 that the chart does not consider compressibility
effects at high speed; we will address this next. We also note:
• The speed for minimum drag tends to increase.
• The minimum drag remains unchanged.
We saw in Chapter 3 that as the airplane altitude increases, the Mach
number will increase for a given true airspeed (see Fig. 3.4). We also know
that as Mach drag rise is reached, the drag coefficient starts increasing
rapidly. The combination of both will induce a limit to the benefit of going

20,000

18,000

16,000
t
0f
,00

14,000
10

12,000
Drag (lb)

10,000 t
00f
8000 3 0,0

6000

4000

2000

0
0 100 200 300 400 500 600 700
Speed (KTAS)

Fig. 8.20 Altitude effect on drag.

 CHAPTER 8 Thrust and Drag Modeling 309

30,000

25,000

20,000
40,000 ft
20,000 ft
Drag (lb)

Sea level
15,000

10,000

5000 Dmin

Dmin impacted
0
0 100 200 300 400 500 600
KTAS

30,000

25,000

20,000 40,000 ft
20,000 ft
Drag (lb)

Sea level
15,000

10,000

5000 Dmin
Dmin impacted

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Fig. 8.21 Compressibility effects combined with altitude impact on drag.

to higher altitudes to reduce the drag at high speed. We created Fig. 8.21
using a drag model similar to that of Fig. 8.10 with an airplane having a
wing loading of 100 lb/ft2 to illustrate this combined impact of altitude
and compressibility.
We still observe the tendency of the drag to reduce at higher airspeed as
the altitude increases, but at one point, we do see an inflection in the drag
310 Operational Aircraft Performance and Flight Test Practices

away from the baseline model defined by constant CDo and K. We even note
that for the highest altitude shown in Fig. 8.21, the minimum drag starts
deviating from a constant value, and therefore there will be an overall loss
of aerodynamic efficiency for the airplane.

Power Required for Flight

It is sometimes more convenient to assess the airplane’s capability in
terms of power available vs power required instead of thrust vs drag. This
is particularly applicable for propeller or fan-driven airplanes (not turbofans)
where the driving engines have maximum power available Pa that is essen-
tially independent of airspeed. We define power required Pr for flight as
the airplane’s drag multiplied by true airspeed.
1 K W2
Pr ¼ D V ¼ rSL s V 3 S CDo þ ð8:33Þ
2 1
r sSV
2 SL
Observation of Eq. (8.33) reveals that the parasite power required (not
due to lift) increases proportional to the cube of the true airspeed, whereas
the power required due to lift decreases only as a function of 1/V. This
indicates that an airplane that has an engine designed to provide a constant
power with airspeed will rapidly run out of excess power (Pa – Pr ) and may
tend to favor a somewhat slower speed regime.
Because the power required parameter is based on drag, all factors
impacting drag that were discussion in the previous section also apply to
the power required.

Minimum Power Required Condition

Just like the drag curve, the power required curve will also have a
minimum condition. Another form of Eq. (8.33) is
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi v
u
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !
W u W u W 3 C
Pr ¼ D V ¼ u ¼u
D
ð8:34Þ
t1 t1 3=2
CL =CD C
r s S CL r sS L
2 SL 2 SL
 We see that the power required is then minimum when the ratio
3=2
CL =CD is maximum. Mathematically, we can find this condition by dif-
ferentiating Eq. (8.34) with respect to CL to get conditions of minimum
power MP. If the drag coefficient is of the form of Eq. (8.12), we find
 
K CL2 MP ¼ 3 CDo ð8:35Þ
pffiffiffi
CLMP ¼ 3 CLEm ð8:36Þ
CDMP ¼ 4 CDo ð8:37Þ
 CHAPTER 8 Thrust and Drag Modeling 311

rffiffiffi
3
EMP ¼ Em ð8:38Þ
4
All of these point to the condition of minimum power required occurring
at an airspeed lower than that for minimum drag, a flight condition con-
sidered to be in the slow flight speed domain.

Slow Flight Speed Domain

The minimum drag airspeed separates two flight regimes. Airspeeds
higher than VDmin represent the positive airspeed stability region where an
increase in thrust produces an increase in airspeed. That is the expected be-
havior for an airplane from a pilot point of view, and the increase in thrust is
provided by the pilot selecting a higher throttle setting.
For airspeeds slower than the minimum drag airspeed, a decrease in
airspeed requires an increase in thrust to maintain level flight, Fig. 8.22. To
slow down, the pilot must reduce thrust below minimum drag value, allow
the airspeed to decay, then re-increase the thrust to maintain the new
lower speed. This is a region of speed instability in that, should the airplane
be disturbed away from the condition of equilibrium with no follow-on pilot
action, the airplane will tend to move away from that condition naturally. If
the disturbance accelerates the airplane slightly, it will then tend to accelerate
to the point of positive speed stability (faster). If, on the other hand, the air-
plane is disturbed towards a slower speed, it will then tend to continue
slowing down until the airplane stalls (unless there is a pilot action to stop

25,000

20,000

15,000
Drag (lb)

10,000
VDmin
Vstall

5000

0
0 100 200 300 400 500
Airspeed (ft/s)

Fig. 8.22 Slow flight speed domain.

312 Operational Aircraft Performance and Flight Test Practices

3500
Flaps LANDING
Gear DOWN
3000
Flaps APROACH
Gear DOWN
2500 Vstall

2000 Dmin Flaps UP

Drag (lb)

Gear DOWN
Flaps UP
1500
Gear UP

1000

500

0
60 80 100 120 140 160 180 200
KCAS

Fig. 8.23 Evolution of slow speed regime on approach to land.

the condition), as was the case for the West Caribbean MD-82 in 2005 (see
the following accident box).
The slow flight airspeed domain can be counterintuitive for a pilot used to
flying in the positive stability airspeed region. In the slow flight domain, a
throttle movement (thrust change) will control altitude, and an elevator
input (change in AoA) will control airspeed. When selecting operational
speed for the crew, this flight regime should be avoided for safety reasons
because during a situation of high workload and little situational awareness,
the pilot may not notice the airplane is not in the expected flight condition
until something happens.
The slow speed flight domain changes with weight, altitude, temperature,
and airplane configuration. We offer the example shown in Fig. 8.23 of a
possible change as an airplane comes in to land from a cruise configuration
(flaps up, gear up). From a higher speed, 190 KCAS in Fig. 8.23, the pilot may
select gear down to help slow the airplane; this will increase parasite drag and
usually has little to no impact on stall speed. We see here that the condition
for minimum drag decreased by about 10 KCAS for this example. As the air-
plane slows down, the pilot selects approach flaps; this will impact both drag
coefficients (CDo and K) and reduce the stall speed; it moves the slow speed
flight domain to a lower speed bracket. As the airplane further reduces its air-
speed, the pilot selects the landing configuration (flaps landing, gear down),
and again the slow flight domain is pushed to even lower speeds. With a
proper set of conditions, the reference approach speed VREF in the landing
configuration will be just above minimum drag conditions (speed stability),
yet slow enough to minimize the landing distance.
 CHAPTER 8 Thrust and Drag Modeling 313

Ministerio del Poder Popular Para Transportey Comunicationes,

Expediente JIAAC-9-058-2005: West Caribbean Airways Flight
WCW-708, 16 Aug. 2005: The airplane departed controlled flight into a
deep stall from high altitude, the sequence of events were: While cruising at
FL310 and M0.75, the pilot requested climb to FL330. Autothrust and autop-
ilot were selected on (thrust channel controlling EPR and vertical channel
controlling Mach). The airplane climbed to FL323 with intermediate steps
(the airplane did not have enough thrust to climb direct to FL330), while
climbing, speed reduced. The pilot then selected autopilot “vertical speed”
mode for a fixed rate of climb and deactivated anti-ice (to produce more
thrust). The airplane finally reached FL330, but the speed had reduced to
M0.70 (minimum drag was estimated to be M0.73), and autopilot retrimmed
the stabilizer nose up to compensate. The anti-ice system was reselected on.
(This resulted in an available thrust reduction.) The airplane started a slow
deceleration. As the airplane reached M0.60, it started losing altitude
rapidly (stalled). By the time the airplane crossed FL317, the sink rate had
increased to over 2500 fpm. The stall warning activated and continued until
impact; the airplane was in a deep stall. The sink rate through 10,000 ft was
approximately 17,400 fpm.
Safety report: The pilots failed to recognize they were flying in a region of
increasing drag and the resulting loss of airspeed until the airplane departed
level flight.
Source: https://reports.aviation-safety.net/2005/20050816-0_MD82_HK-
4374X.pdf

Thrust Model
The force used to counter the airplane drag is the thrust produced by the
propulsion system. Many such systems have been developed over the years
with the dominant ones being the reciprocating engine and the gas turbine
(see Fig. 8.24). In more recent years, electric motors have started becoming

Airplane propulsion system

Reciprocating Electric motor Gas turbine Ramjet/Scramjet Rocket

Energy Energy conversion Thrust

1. Fuel/gasoline To mechanical work 1. Propeller/Open rotor
2. Electricity 1. Propulsion system 2. Fan/Jet
2. Hybrid system

Fig. 8.24 Propulsion system tree.

314 Operational Aircraft Performance and Flight Test Practices

in fashion with concerns for the environment. The basic principle of a pro-
pulsion system is to convert a source of energy into a forward thrust via
some mechanical work (see Fig. 8.24).
We will cover the basic principles of the three main propulsion systems
specified to see how their basic mode of operation could impact the thrust
generation and the consumption of energy to produce the thrust.

Reciprocating Engines
Reciprocating (or internal combustion) engines have been part of aviation
since the first flight of the Wright brothers in 1903. They use pistons in cylin-
ders to compress a gasoline–air mixture and ignite it to convert the energy
from aviation gasoline into work to drive a propeller (and in a few cases,
ducted fans). The reciprocating engine is still the most popular engine in
aviation, although now it is mostly found on the smaller general aviation
(GA) airplanes and smaller helicopters. Within this category, the four-stroke
engine is dominant, and we will use this type to create an engine power
model.
A four-stroke reciprocating engine has a noncontinuous cycle. There is a
“dead” engine revolution for a given piston where essentially no work is done
because the valves are open to either get air into the cylinder or push gases
out. It will therefore take two revolutions of the crankshaft to produce one
power cycle. A mathematical model that can be used while analyzing recipro-
cating engines is the Otto cycle (see Fig. 8.25).
The Otto cycle has the following phases:
1. From a to b, fresh air/gasoline intake, volume change at constant
pressure, no work
2. From b to c, valves close, gas mixture compressed isentropically, work done
to mixture (negative) by the piston
3. From c to d, ignition of the gas mixture at constant volume (transfer of
energy from the gasoline to the system), rise in pressure, no work

d Ideal otto cycle Real Cycle

Combustion
qin
ends
Ise Ex
nt pa
ro ns
pic ion
c Intake valve Ignition Exhaust valve
opens Com opens
Isen e pres
trop Exhaust sion
patm ic qout patm
a bf Intake valve
Intake closes

Min Max Exhaust valve Min Max

volume volume closes volume volume
A A
Compression ratio = ( max
/ min)

Fig. 8.25 Ideal Otto cycle vs “real cycle.”

 CHAPTER 8 Thrust and Drag Modeling 315

4. From d to e, expansion of the burnt gas mixture under isentropic

conditions, work done by the mixture (positive) onto the piston
5. From e to f, valves open, pressure goes back to outside air pressure,
no work
6. From f to a, piston pushes out gas mixture at constant pressure, no work
We know the cycle will not be ideal and may be more represented by the
“real cycle” on the right of Fig. 8.25. This real cycle will capture the actual effi-
ciency of the specific engine. We use the ideal cycle in the discussion that
follows to see what drives the ideal efficiency of a reciprocating engine.
The engine compression ratio r is defined as the ratio of the maximum
volume available at the bottom of the piston stroke to the ratio of the
minimum volume at the top of the stroke
8max
r¼ ð8:39Þ
8min
The temperature and pressure rise during the isentropic compression,
segment b to c, can be expressed as
Tc
¼ r ðgg 1Þ ð8:40Þ
Tb
pc
¼ r gg ð8:41Þ
pb
where gg is the ratio of specific heat for the gasoline–air mixture. Its value
depends on the relative ratio of each (also referred to as the fuel–air
ratio f ), but we can use an approximate value of 1.3 for the discussion of
the analysis of reciprocating engines. The next part of the cycle is the injec-
tion of energy at the time of the combustion of the gasoline–air mixture
(mixture ignited with a spark plug). Here we get the following change in
temperature and pressure (under constant volume conditions):
f LHV
Td ¼ Tc þ ð8:42Þ
cv
pd Td
¼ ð8:43Þ
pc Tc
Where the LHV is the low heating value of the gasoline; it represents the
usable specific energy of the gasoline. Reciprocating engines in aviation are
typically powered by aviation gasoline (100LL Avgas) with a specific energy
(LHV) of about 18,800 BTU/lb (43.71 MJ/kg). The coefficient cv in
Eq. (8.42) represents the specific heat of the gas–air mixture at constant
volume. Its value is approximately 0.16 BTU/(lbm 8R) or 0.655 kJ/(kg K).
The next part of the cycle where work is done is the power stroke where
the gas mixture is expanded. The change in temperature and pressure during
316 Operational Aircraft Performance and Flight Test Practices

this time takes on a similar form as during compression, namely

Te
¼ r ð1gg Þ ð8:44Þ
Td
pe
¼ r gg ð8:45Þ
pd
The net work W done by the cycle of compression–combustion–expan-
sion is the difference between the power stroke (expansion) and the com-
pression stroke, and takes the form
ð
 
W ¼ Wpower  Wcomp ¼ ppower  pcomp d8 ð8:46Þ

or,
W ¼ cv ½ðTd  Tc Þ  ðTe  Tb Þ m ð8:47Þ
where m is the mass of the fuel–air mixture in the cylinder. (The volume in
each cylinder and the intake pressure will dictate the mass of air and fuel
vapor that it can hold.) The net power produced by this cycle is equal to
the net work done per unit time, so reciprocating engines typically have a
design rotational speed [revolutions per minute (RPM)] point. For an
engine with N cylinders, using RPM as a reference of time (common units
for reciprocating engines), the indicated power (IP) for the engine is
expressed as follows (remember that only one in two revolutions produces
work, thus the 12 in the equation):
1 RPM
IP ¼ NW ð8:48Þ
2 60
The units commonly used for power in airplane performance are either
horsepower (HP) or kilowatts (kW). The conversion factors among the
various units are
1 HP ¼ 550 ft-lb=s ¼ 745:7 J=s ¼ 745:7 W ¼ 0:7067 BTU=s

Efficiency of a Reciprocating Engine

The efficiency of the engine is defined as the work produced vs the energy
(heat) input. The ideal efficiency of the Otto cycle, the maximum achievable
for a given engine design (compression ratio), is
work done Te  Tb 1
hi ¼ ¼1 ¼ 1  g 1 ð8:49Þ
heat input Td  Tc r g

We see that as compression ratio increases, so does the efficiency of the

cycle. Figure 8.26 shows the general trend of this efficiency.
From Fig. 8.26, one can note that at a compression ratio of 10, the ideal
efficiency would be 50% (that is for a specific heat ratio of 1.3 as used in our
 CHAPTER 8 Thrust and Drag Modeling 317

60%

50%

40%
Ideal efficiency

30%

20%

10%

0%
0 2 4 6 8 10 12 14 16
Compression Ratio

Fig. 8.26 Ideal efficiency of a reciprocating four-stroke engine vs compression ratio.

assumptions), meaning that only 50% of the energy injected into the system
(from gasoline) would be transformed into useful work. If one were to
increase the compression ratio further, an increase in efficiency would be
expected, but so would an increase in temperature Tc (temperature before
the fuel injection); if this temperature is too high, the mixture would auto-
ignite in the wrong location of the cycle. Avgas 100LL has an auto-ignition
temperature of 2468C (474.88F); thus, a peak compression temperature
some margin away from this value is usually the condition limiting the
maximum usable compression ratio.
The brake power [often referred to as brake horsepower (BHP)] is the
power delivered to the shaft (that is then connected to a propeller or a
fan). It is related to the IHP through a mechanical efficiency factor (hmech ).
That efficiency is essentially 100% for direct connection to a typical mid-
nineties when a gearbox is used to transfer the power from the engine to
the propeller or fan. The BHP can be expressed as

2p RPM t
BHP ¼ ¼ hmech IHP ð8:50Þ
33,000

where t is the torque provided to the propeller or fan. As one can see from
Eq. (8.50), the power is proportional to the RPM, so the design RPM must be
well matched to the propeller/fan optimum RPM to extract the most power
efficiently. We will cover propellers and fans later in this chapter.
318 Operational Aircraft Performance and Flight Test Practices

The “real” thermal efficiency of the engine takes on the following form:
BHP
hth ¼ ð8:51Þ
Ẇ f cv

The reciprocating engine efficiency is typically expressed in the form of

power specific fuel consumption, the ratio of the fuel flow to the shaft horse-
power produced.

Ẇ f
SFCp ¼ ð8:52Þ
BHP
This is not a constant value; the lowest SFC occurs at a design RPM.
Figure 8.27 is such an example, with the chart extracted from the user
manual of a Lycoming engine; the optimum efficiency here is near 2500 rpm.
A reciprocating engine requires a minimum RPM to remain on, the idle
setting. This condition is defined by the manufacturer as the RPM that is just
enough to support the power extraction for the accessories (e.g., generator)
and maintain a stable engine rotating speed. This means that while providing
no useful work to a propeller or fan for the purpose of keeping the airplane
flying, the engine is still consuming energy (gasoline) to keep operating.
One source of wasted energy in a reciprocating engine is the heat rejec-
tion into atmosphere—station e to f where the outflow valve is opened and
the cylinder is vented to the outside. The energy loss is proportional to
[cv (Te – Tf ) m].
We saw that the power generated by a reciprocating engine is pro-
portional to the engine RPM in Eq. (8.50). In theory, the RPM could just
be increased for a given engine design and extract more power out of it;
there are, however, mechanical limitations to this statement. First, the
engine’s moving parts have strength limits; the piston must go from max
volume to min volume and back to max volume again for each and every
revolution. This generates large accelerations and forces that must be
absorbed several times per second over the life of the engine (which has a
large impact on the parts’ fatigue life). Second, the reciprocating engine by
itself does not generate flight power; for that, the engine is typically con-
nected to a propeller. That propeller also has mechanical limits; in addition,
the RPM will be limited to conditions where the propeller tip does not exceed
approximately Mach 1.0. This will be covered in the propeller section of this
chapter. For this reason, typical maximum RPM, without the use of a
reduction gearbox, is typically limited to between 2500 and 3000 RPM for
reciprocating engines.
Equation (8.42) indicated that it may be possible to increase the work
done by the system, and therefore the power generated, by increasing the
fuel air ratio f in the system because it would increase the heat released
into the system for a given fuel heating value (LHV). Unfortunately, one
must deal with a system that has a very finite time to burn the fuel
 CHAPTER 8 Thrust and Drag Modeling 319

Curve no. 11259

Sea level power curve
lycoming model O-320-B,-D series
Compression ratio- 8.5:1
Spark timing- 25° BTC
Carburetor-marvel schebler-MA-4 SPA
Mixture setting- Full rich
Normal
Fuel grade- 91/96
rated
170 Operating conditions- STD. Sea level
power

160
Full throttle
horsepower
150

140
Brake horsepower

130

120

110

100
Propeller load
horsepower
90

80

70
Full throttle
specific fuel
consumption
55
Spec. fuel cons.
LB/BHP/HR

50
Propeller load
specific fuel
45
consumption
1800 2000 2200 2400 2600 2800
Engine speed – RPM

Fig. 8.27 Lycoming model O-320 power. Source: Lycoming user manual.

mixture, and one must also account for the amount of oxygen present in the
system to support the combustion. If the f is too high (mixture too rich) there
may be too much fuel present compared to the oxygen level, and the combus-
tion may not be complete. On the other hand, if f is too low (lean mixture),
less heat will be released into the system. The typical f to be expected with this
system varies from approximately 0.06 (lean) to 0.068 (stoichiometric com-
bustion for gasoline) to 0.10 (overrich).
320 Operational Aircraft Performance and Flight Test Practices

The mass of air introduced into the system, the charge per stroke, has a
direct impact on the power produced for a given f, as seen in Eq. (8.47).
This mass is proportional to the cylinder size (volume), the air density, and
the intake pressure [also called the manifold absolute pressure (MAP)].
Pressure and temperature have a direct impact on the air density, so one
needs to consider the density altitude as a primary impact on the power.
Other factors impacting the MAP include:
• The shape of the intake and the resulting pressure in the engine
compartment where the engine draws its air.
• The airspeed (impact pressure) may increase the intake pressure if intake is
properly aligned with airflow. (Some of the air must still flow around the
engine to provide cooling.)
• The use of a turbocharger to increase the intake pressure. Such a system
also increases the air temperature in the intake, thus impacting the air
density, and it is powered by energy from the system and thus will have an
impact on fuel consumption.
• A properly designed intake close to the propeller plane may actually take
advantage of the local pressure increase behind the plane with little
propeller efficiency loss.

Maximum Power Available Versus Altitude

We now define a standard model for the purpose of modeling the power
available vs atmospheric properties for the performance analysis. We have
seen that several variables impact the maximum power produced by a reci-
procating engine, and these could be modeled if a specific model/engine is
used. However, for general performance analysis we will reduce the model to:
• The power produced by the reciprocating engine will be independent of
the airspeed.
• For nonturbocharged engines, the maximum power available will be
proportional to the density altitude (combination of air pressure and air
temperature that dictates the mass of air in the cylinder).
P
¼s ð8:53Þ
PSL
• For turbocharged engines, one will assume the power is maintained
constant to a critical density altitude (Hcrit ) by the turbocharger, beyond
which the maximum power available decreases proportional to the density
altitude.
(
P 1:0 Hs  Hcrit
¼ s ð8:54Þ
PSL Hs . Hcrit
sscrit
 CHAPTER 8 Thrust and Drag Modeling 321

This author has used different models over the years and in previous pub-
lications, but has found that this simpler approach reminds students that the
power is a function of the mass of the fuel–air mixture in the system, and it is
easily understood. Under standard atmospheric conditions, the density alti-
tude and the pressure altitude are the same.

Reciprocating Engines’ Impact on the Environment and

Carbon Dioxide
The impact of burning fossil fuels to power airplanes has been steadily
growing over the years and has greatly accelerated since the mid-2010s. A
major measure of the impact of burning fuel is the production of carbon
dioxide (CO2 ). The amount of CO2 produced by a given fuel is determined
by the amount of carbon (C) present in it. A reciprocating engine burning
aviation gasoline (Avgas) will produce approximately 3.01 lb of CO2 for
every pound of gasoline burned.
The standard Avgas is the 100LL (low lead), meaning that general
aviation (GA) airplanes still burn gasoline with lead in it; at the time of
writing this chapter, Avgas was the only remaining lead-containing trans-
portation fuel (https://www.faa.gov/about/initiatives/avgas/). The lead in
the Avgas helps to prevent engine knock or detonation that can be dama-
ging and lead to failure. Several initiatives have been proposed to replace
Avgas with alternatives, but a very large number of GA airplanes (still
over 167,000 in the United States in early 2020) cannot be upgraded
easily to these new fuels, which is financially impacting the phasing out
of Avgas.
Some manufacturers are introducing diesel engines in GA in an effort to
offer an alternative to the disappearing Avgas. Diesel is similar to the most
commonly used aviation fuel, the Jet A type, although they do differ in
many ways. For one, Jet A fuel has a relatively high sulfur content whereas
diesel has a low sulfur content.
The diesel engine cycle differs from the Otto cycle in that the fuel–air
mixture is injected when the cylinder reaches top compression, and the com-
pression ratio auto-ignites the mixture. This allows the system to use a higher
compression ratio of around 15 to 20 with a resulting higher efficiency. Diesel
engines produce approximately 3.10 lb of CO2 for every pound of fuel used.

Gas Turbines
The next engine type of interest, and the one that is essentially the sole
source of power for transport category airplanes, is the gas turbine. The
main advantage of these engines over the reciprocating engines is their
use of continuous airflow, which allows for greater air mass flow through
the system and thus greater power generation compared to a similar size
322 Operational Aircraft Performance and Flight Test Practices

(weight) reciprocating engine. The basic principle of operation of a gas

turbine includes (see Fig. 8.28):
1. Ambient air is fed into a compressor (0 ! 2 ! 3) via an inlet.
(a) The compressor does work (Wc ) on the air to increase the pressure
with a resulting temperature increase.
(b) Conditions 2 and 3 represent total pressure and total temperature
conditions.
2. The air enters a combustion chamber where fuel is added and ignited
(qin ). The temperature increases under constant pressure conditions
(3 ! 4).
3. The combustion product is fed into a turbine (4 ! 5) where some energy
is extracted to feed the compressor (Wc ).
4. The remaining energy (5 ! 6) is available to provide propulsive power in
various ways (jet and/or shaft) that we will describe later in this chapter.
(a) Condition 5 is the total pressure and total temperature condition, and
condition 6 represents the static pressure (atmospheric) conditions.
The cycle shown in Fig. 8.28 is an ideal Brayton cycle. The cycle, as illus-
trated, assumes ideal efficiency for all segments of the cycle; from this we get
the theoretical ideal efficiency. The thermal efficiency hth of the cycle is
defined as the ratio of the net work produced to the heat energy injected
into the cycle, which can be approximated by the following equation (assum-
ing constant specific heat cp ):

net work cp ½ðT04  T6 Þ  ðT03  T0 Þ

hth ¼ ¼
heat input cp ðT04  T03 Þ
ðT6  T0 Þ
¼1 ð8:55Þ
ðT04  T03 Þ

The available energy for the airplane (5 ! 6) will increase if:

• The pressure ratio (0 ! 3) is increased for a given fuel energy input
(3 ! 4)
• More fuel energy is inserted into the system for a given pressure ratio
• A combination of both

4
Fuel (qin)
Combustion qin 5
Temperature

Energy
Inlet Compressor Turbine Exhaust Wc available
3
Wc Wshaft 6
ssure
ant pre
2 f const
0 Lines o
0 1 2 3 4 5 6
Free Entropy
stream

Fig. 8.28 Gas turbine cycle.

 CHAPTER 8 Thrust and Drag Modeling 323

80%

70%

60%
Ideal thermal efficiency

50%

40%

CFM56-3B (1982)

CFM56-5B (1992)

CFM LEAP (2014)

PT6A-67D (1990)

PW1500G (2011)
GE90-90B (1995)
JT8D-219 (1984)
30%

JT9D-3A (1969)
JT9D-70 (1973)
PT6A-6 (1963)

JT8D-1 (1963)

GE9X (2020)
20%

10%

0%
0 10 20 30 40 50 60
PR

Fig. 8.29 Thermal efficiency as a function of pressure ratio.

For an ideal cycle where the compression (0 ! 3) and the expansion

(4 ! 6) are isentropic and where the pressure remains constant in the com-
bustion chamber (3 ! 4), the thermal efficiency of the cycle can be written in
terms of the pressure ratio (PR) of the compression cycle (0 ! 3)
1
hth ¼ 1  ð8:56Þ
PRððg1Þ=gÞ
where g is the ratio of the specific heats for the working fluid. For gas tur-
bines, we start with air, but that changes as fuel is added into the combustion
chamber, of course. Using the reference value of 1.4 for air, Fig. 8.29 was
created. This chart shows that over the years, engine manufacturers have
been striving to find ways to improve the efficiency of gas turbines by building
engines with ever greater pressure ratios. We also see how flat the curve is
beyond a pressure ratio of about 20.
One of the impacts of increased pressure ratio is on the airplane’s effi-
ciency to carry a load. Increasing the pressure ratio results in higher core
engine pressures and temperatures. The engine casing must be made stron-
ger to resist the higher pressures, the engine must use more specialized
material to handle the higher temperature, and additional cooling of the
turbine blades must be introduced. The result is a heavier engine. We offer
Table 8.1 to show the evolution of a similar airframe (the Boeing 737,
three generations of about the same passenger load) and a similar engine
(same manufacturer). The engine efficiency (SFCT ) did improve over the
324 Operational Aircraft Performance and Flight Test Practices

Table 8.1 Evolution of an Engine on a Similar Airframe

Engine CFM56-3B-2 CFM56-7B26 CFM LEAP-1B28

Aircraft 737-400 (1988) 737-800 (1998) 737-8 MAX (2017)
MTOW 150,000 lb 174,200 lb 181,200 lb
Seating 159 all-economy 184 all-economy 189 all-economy
Max takeoff thrust 22,000 lb 26,300 lb 29,320 lb
Dry weight 4670 lb 5215 lb 6130 lb
Thrust/weight 4.7 5.0 4.8
Overall pressure ratio 28.8 32.7 40
Bypass ratio 5.9 5.1 9.0
Cruise SFC (est.) 0.67 lb/hr/lb 0.64 lb/hr/lb 0.55 lb/hr/lb

years from the Classic 737 (737-400) to the Next Generation (737-800) to the
current version (737-8), but the engines, as well as the airplanes, also grew in
weight to carry essentially the same passenger load. With all three airplanes
having similar maximum L/D (there were small increases from one model to
the next), the improvement in cruise efficiency (SAR, see Chapter 10) is not
fully proportional to the SFCT improvement.
The actual cycle is, of course, not ideal, so we use an axial flow compres-
sor to illustrate how some variables influence the efficiency and operation of
the gas turbine cycle. An axial flow compressor consists of a series of cascade
stages aligned along the longitudinal axis of the engine. Each stage is made up
of a rotor row (rotating part of the cascade stage) and a stator row (fixed); see
Fig. 8.30. Each rotor and stator row is made up of several airfoil-like blades
that control the flow. A mass flow . is fed into the compressor, which is pro-
portional to the inlet area Ai , the air density r, and an airspeed. As the air gets

Ai =  D2 Stators
4
rmean

Rotors
m· RPM
Q
 w  w  w

2
U U= RPM rmean U
60
Stalled Normal operation Stalled

Fig. 8.30 Axial flow compressor operating environment, rotor row.

 CHAPTER 8 Thrust and Drag Modeling 325

to the rotor, energy is added to compress the air in the form of a torque Q
applied to the rotor turning at a given RPM.
One can see from Fig. 8.30 that the combination of airflow coming into
the engine (axial velocity) and angular speed of the rotor U need to be well
matched to ensure the blades do not stall. Too slow of an RPM will result
in a negative stall; too fast of an RPM will lead to a positive stall. In
between these two conditions, the rotor will operate at varying efficiency
(like the L/D of the wing). Then the air mass flows in the stator row,
where it is redirected to best align with the next rotor row.
2p
U¼ RPM ð8:57Þ
60
The total temperature rise across a rotor row is proportional to the RPM
and the change in angular momentum Dcu of the air mass flow.
DT0 U Dcu
¼ ð8:58Þ
T01 cp T01
where T01 is the total temperature at the entrance to the rotor row. The effi-
ciency of the design will dictate the change in angular momentum. Because
the stator is fixed, no work is done in the flow for this part of the stage, so the
total gain in energy is the same for the rotor–stator pair. The torque Q fed to
the compressor comes from the turbine, and both systems must be optimized
to work together because they are tied by a common shaft and therefore have
the same RPM.
Final note on the gas turbine cycle: Note how much energy is wasted
(simply released into the air in the form of heat).
Wasted heat / cp ðT6  T0 Þ ð8:59Þ

Turbojet Cycle
A turbojet is a gas turbine in which all the air ingested flows through the
combustion chamber and the energy left at the turbine exit plane (see
Fig. 8.28) is then accelerated to produce thrust. That thrust is proportional
to the change of momentum of the mass flow going through the turbojet
plus the difference in pressure at the exhaust minus the atmospheric pressure
T ¼ ṁa ½ð1 þ f Þ ue  V1  þ ð pe  p1 Þ Ae ð8:60Þ
where ue represents the hot gas exhaust airspeed, pe is the exhaust plane
pressure, V1 is the airplane true airspeed, and p1 is the atmospheric pressure
at the flight altitude. For most subsonic airplanes, the gases are fully
expanded to atmospheric pressure by the time they reach the exhaust
plane, and Eq. (8.60) reduces to
T ¼ ṁa ½ð1 þ f Þ ue  V1  ð8:61Þ
326 Operational Aircraft Performance and Flight Test Practices

From this equation, one can expect some thrust reduction as the airplane’s
airspeed increases. (This is the thrust lapse rate with speed.) This is partially
offset by the increased total pressure. Here f is the fuel–air mass ratio
ṁf
f ¼ ð8:62Þ
ṁa
That fuel–air mass ratio is proportional to the temperature rise in the
combustion chamber and the fuel LHV
cp ðT04  T03 Þ
f ¼ ð8:63Þ
LHV
The low heating value (LHV) of Jet A fuel is 18,550 BTU/lb
(43.056 MJ/kg, 11.98 kWh/kg). This fuel type has a density of 6.76 lb/ft3
at 158C (see Fig. 5.17 in Chapter 5). The composition of this fuel is defined
by standards such as the American Society for Testing and Materials
(ASTM) D1655.
A limiting factor for T04 , the turbine inlet temperature, is the capacity of
the first turbine stage, rotating at high RPM, to absorb the heat. This drives
the type of material used, the cooling requirements, and the durability of the
stage. T04 is impacted by the combination of pressure ratio and fuel–air
mass ratio.
The exhaust flow airspeed can be computed by expanding the flow from
the total temperature and total pressure conditions after the turbine (05) to
the atmospheric pressure
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
u
u g  13
u 6 p6
ue ¼ u g 7
t2 cp T05 41  5 ð8:64Þ
p05

The engine’s efficiency at producing thrust is represented by the thrust

specific fuel consumption (SFCT ), which is the ratio of the fuel mass flow
to the thrust produced. With the previous equations, we find that the
SFCT for a turbojet can be written as
ṁf f
SFCT ¼ ¼ ð8:65Þ
T ð1 þ f Þ ue  V1
Although the equation seems simple, the combination of engine oper-
ation (RPM), altitude, temperature, and airspeed affect thrust production
and SFCT . The general behavior of thrust vs engine RPM at a given atmos-
pheric condition is similar to Fig. 8.31a. At low RPM, the engine is not
pumping air fast enough and is not very efficient. There will be a large
change in RPM with very little change in thrust. The idle RPM (minimum
RPM) is selected so that the engine remains stable. A constant RPM accelera-
tion from idle will not produce much thrust change in the low RPM regime
 CHAPTER 8 Thrust and Drag Modeling 327

a) b)
% max SFCT/ 
thrust

Mmax
RPM
max


Design
cruise
RPM
 idle
Idle
M=0

RPM T/

Fig. 8.31 Thrust and SFC variation vs RPM.

but will produce a much faster one at higher RPM. Figure 8.31b shows a
typical impact of RPM, temperature u, altitude d, and airspeed (Mach) on
the engine efficiency SFCT .
The thermal efficiency of the turbojet cycle can now be rewritten from
Eq. (8.55) to one that represents the rate of production of kinetic energy to
the heat input.
1
ð1 þ f Þ u2e  V1
2
hth ¼ 2 ð8:66Þ
f LHV
The propulsive efficiency hpr is defined as the ability for the engine to
develop flight power (T V1 ) divided by the rate of production of kinetic
energy.
T V1
h pr ¼ ð8:67Þ
1
ṁa ð1 þ f Þ u2e  V1
2
2
With the assumption that the fuel–air ratio is small (typically 0.05 to 0.2)
and combining Eqs. (8.67) and (8.61), one gets
 
2 Vu1e
h pr    ð8:68Þ
1 þ Vu1e

What we observe from Fig. 8.32 and Eq. (8.68) is that the propulsive effi-
ciency tends towards 100% as the speed ratio approaches 1.0. But in doing so,
Eq. (8.61) tells us that the thrust produced would also approach zero. As well,
the thermal efficiency of the system, from Eq. (8.66), would also approach
zero. One wants to maximize the total efficiency of the system; that efficiency
328 Operational Aircraft Performance and Flight Test Practices

0.9

0.8

0.7

0.6

0.5
pr

0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(V/ue)

Fig. 8.32 Propulsive efficiency vs speed ratio.

is defined as the product of the thermal and propulsive efficiency, or

T V1 ½ð1 þ f Þ ue  V1  V1
h ¼ hth h pr ¼ ¼ ð8:69Þ
ṁf LHV f LHV
We already saw that the thermal efficiency can be increased by increasing
the pressure ratio. One must then find a way to improve the propulsive effi-
ciency to increase the overall efficiency of the gas turbine.

Turbofan Cycle
A turbofan is a gas turbine system that has a core (flow going through the
combustion chamber, ṁhot , similar to a turbojet) and another propulsion
system (a ducted fan) connected in series. The fan is either in front of the
core (most common configuration) or aft. The aim of the configuration is
to increase the air mass flow through the system.
On a turbofan, part of the energy available (see Fig. 8.28 earlier in the
chapter) is provided to a ducted fan that will draw in additional air mass
flow that will not go into the combustion chamber, ṁcold . When the fan is
connected ahead of the core, it can also provide additional compression to
the core flow (see Fig. 8.33). The ratio of this extra air to that going
through the core is called the bypass ratio b.
ṁcold
b¼ ð8:70Þ
ṁhot
 CHAPTER 8 Thrust and Drag Modeling 329

Combustion
Fan chamber
4
Core flow
m·cold
qin 5 Bypass air
Wfan (1+f )m·hot
Temperature

Wc 6 m·a
3 Exhaust thrust
7
ssure
ant pre
2 f const
0 Lines o

Entropy High pressure Low pressure

compressor turbine
Low pressure High pressure
compressor turbine

Fig. 8.33 Turbofan cycle.

The advantage of the turbofan over the turbojet is that, while maintaining
a high core flow pressure ratio (high thermal efficiency), it can reduce the
average speed of the exhaust flow (increased propulsive efficiency) and
thus improve the overall fuel efficiency of the engine. Engine manufacturers
have progressively increased the bypass ratio of engines over the years for jet-
liners and business jets (see Fig. 8.34) from a bypass near 1.0 to over 12.2
today with the Pratt & Whitney Geared Turbofan (GTF). Engine manufac-
turers are looking at increasing this value even more (ultra-high bypass
ratio) in the future. The increased bypass ratio comes at the cost of increased
weight (heavier fan and heavier fan duct), which impacts the overall airplane

1 Early turbojets

Turbofans
0.9

0.8 Medium bypass

Medium bypass
high pressure ratio
0.7 Business jets
High bypass
Regional jets
0.6
Relative SFCT

Very high bypass Very high bypass

Geared fan
Commerical
0.5 jetliners
High pressure ratio

0.4 Pratt&Whitney GTF

0.3
Propfan and open rotor
0.2

0.1

0
1950 1960 1970 1980 1990 2000 2010 2020

Fig. 8.34 Improvement in SFC over the years.

330 Operational Aircraft Performance and Flight Test Practices

efficiency (airplane structure heavier to support engine, impacts airplane

drag). Unducted fan engines (compromise between turbofan and turboprop)
show promises of achieving the increased bypass goal without the weight
increase, but have not yet been perfected enough for operational use and
still have to address noise production.
The thrust equation for the turbofan can be written as
 
T ¼ ṁhot ð1 þ f Þ uehot  V1 þ ṁcold uecold  V1 ð8:71Þ
where
ṁf
f ¼ ð8:72Þ
ṁhot
and the thrust specific fuel consumption is
ṁf f
SFCT ¼ ¼   ð8:73Þ
T ð1 þ f Þ uehot  V1 þ b uecold  V1
We can see that the larger the bypass (everything else staying the same),
the lower the SFCT .

Turboprop Cycle
A turboprop is a gas turbine system that has a core (flow going through
the combustion chamber, similar to a turbojet) and a propeller connected in
series, often through a reduction gear box (see Fig. 8.35). The turboprop cycle
has a greater propulsive efficiency than the turbojet and turbofan cycles due
to the considerably larger bypass air (flow not going through the core).
The gas turbine system part of the turboprop is essentially a turbojet with
an extra turbine that drives a shaft that provides power to the propeller
system. The reduction gear box must be properly sized to allow the gas
turbine to turn at its optimal RPM while the propeller also gets to turn at
its much lower optimal RPM. In this way, the thermal efficiency of the
system, and the propulsive efficiency of the propeller, can be maximized.

Propeller

Reduction
gearbox Combustion
4 chamber
Shaft
qin 5
(1 + f )m·a
Temperature

Wshaft
Wc 6
Exhaust thrust
3
ure 7
nt press m·a
2
0 es of consta
Lin
Entropy Compressor Low pressure
High pressure turbine
turbine

Fig. 8.35 Turboprop cycle.

 CHAPTER 8 Thrust and Drag Modeling 331

2
Sea level ISA
1.8
10,000 ft
1.6

1.4
20,000 ft
SFCp (Ib/hr/SHP)

1.2

1
25,000 ft
0.8

0.6

0.4

0.2

0
0 100 200 300 400 500 600 700 800 900 1000
SHP

Fig. 8.36 Airspeed has little effect on SFC, throttle setting and altitude have more.

Turboprops are typically rated in terms of power generation capability (P,

typically shaft horsepower, SHP) rather than thrust generated because most
of the thrust generated by the system comes from the propeller, the gas
turbine being the power generator. The power P extracted from the turbo-
prop cycle and provided to the shaft driving the propeller is

P ¼ ṁa ð1 þ f Þ cp ðT05  T06 Þ ð8:74Þ

The exhaust flow from the gas turbine still contains some energy that can
produce a useful amount of jet thrust. For this reason, the power capability of
a turboprop engine is often quoted in terms of equivalent shaft horsepower
(ESHP)

T V1
ESHP ¼ SHP þ ð8:75Þ
hp

where hp is the propeller efficiency. This extra jet thrust tends to become a
greater percentage of the overall thrust production (propeller + jet thrust) as
the airplane goes up in altitude. The efficiency of the gas turbine part of the
turboprop system is provided by a specific fuel consumption based on power
(SFCP ) rather than thrust produced. Figure 8.36 shows a typical SFCP behav-
ior for a turboprop engine. Just like other gas turbine engines, the turboprop
332 Operational Aircraft Performance and Flight Test Practices

qin 5
Wshaft

Temperature
Wc
6 Negligible thrust
3
7
ure
t press
2 s of constan
0 Line

Entropy

Fig. 8.37 Turboshaft cycle.

tends to be less efficient at the lower power settings.

ṁf
SFCP ¼ ð8:76Þ
SHP
Turboprops are typically lighter weight than reciprocating engines for a
given amount of power generation, but their efficiency has yet to match reci-
procating engines for small Part 23 GA airplanes. Turboprops have replaced
reciprocating engines for all Part 25 airplanes.

Turboshaft
The last gas turbine cycle we will review is the turboshaft cycle. In this
configuration, the available energy leaving the core turbine is essentially com-
pletely removed, and all useful energy from flow is used to drive a shaft (see
Fig. 8.37). That shaft can be connected to ducted fans (not in series like tur-
bofans) or to electric generators.
An example of a turboshaft installation on modern jetliners are the auxili-
ary power units (APUs) typically located in the tail of the airplane. The APU
provides energy in various forms including powering electrical generators
and providing bleed air to the airplane (ECS, engine starts, etc.).

Maximum Thrust or Power Available Versus Altitude

The gas turbine cycles presented are a function of multiple variables, as
shown. For general airplane performance analysis, we will assume the
model behaves as follows:
• Maximum thrust or power is proportional to the air mass flow [Eqs. (8.61),
(8.71), and (8.74)] and therefore to the air density.
T P
¼s ¼s ð8:76Þ
TSL PSL
 CHAPTER 8 Thrust and Drag Modeling 333

• For the purpose of sensitivity analysis, again, the power generated by the
turboprop and the thrust generated by turbojets and turbofans would be
independent of airspeed.

Gas Turbine Engines’ Impact on the Environment and CO2

Gas turbines burn jet fuel. Burning jet fuel will produce approximately
3.15 lb of CO2 per pound of jet fuel. Considering that commercial aviation
by early 2020 was burning nearly 100 billion US gallons of fuel per year
and that each gallon contains approximately 6.7 lb of fuel, the yearly pro-
duction of CO2 attributed to commercial aviation was over 1 billion tons.
This represents only about 2% of the world’s CO2 generation but is expected
to continue to grow rapidly.
To help curb the growth of CO2 emissions, the use of biofuels was
approved in July 2011. Biofuels are generated from biomass, in contrast
with traditional fuels that are produced from fossil oil. It is a means to
bring the process of production–use–recycling (plants absorbing the CO2
produced) closer to a carbon-neutral (no CO2 growth) cycle. Biofuels were
still not readily available as of 2020; only a few airports actually have the
fuel on site.

Electric Motors
The final category of engines we will describe for this textbook are electric
motors. This type of motor started to power small airplanes in the late
2010s—the first FAA light sport airplane (LSA) was certified in 2018, the
Pipistrel Alpha Electro. This was followed by the first flight of the Harbour
Air Electric Beaver (a modified de Havilland DHC-2 Beaver) in Dec. 2019.
This author has had a chance to participate in the design of the Heart Aero-
space ES-19 since June 2019 and to have the first discussions with the Euro-
pean Aviation Safety Agency (EASA) in Nov. 2019 followed by formal Type
Certificate application in 2021. The ES-19 is an all-electric 19-seat regional
airplane (see Fig. 8.38).
What is particularly attractive for electric motors is that the energy stored
on the plane is used directly by a highly efficient electric motor (efficiency to

Pipistrel alpha electro Harbour air electric beaver Heart aerospace ES-19
FAA certified LSA 2018 first flight 10 December 2019 Type Certification application 2021

Fig. 8.38 Early electrically propelled airplanes.

334 Operational Aircraft Performance and Flight Test Practices

convert electric power to shaft power of the order of 90–97%; see Fig. 8.39).
These engines also operate at much lower temperatures (maximum around
1008C to 1508C) than reciprocating engines or gas turbines. These motors
will drive propellers, fans, and rotors depending on the application. ASTM
F3338 [3] provides design standards for electric propulsion.
Another advantage for electric motors is that they are typically much
lighter than equivalent power reciprocating or gas turbine engines (see
Fig. 8.40). They also are much less complicated systems with significantly
fewer parts, they produce less vibration, they have simpler installation
requirements, and they are expected to have reduced maintenance and acqui-
sition costs. (This last cost is vendor dependent.)

50,000 1200

Torque (blue/black dashed line) (Nm)

EMRAX 348 low voltage CC 1100
Power (blue/black solid line) (kw)

45,000
1000
40,000 Peak torque
900
35,000 Peak power
800
30,000 700
25,000 600
Continuous torque
20,000 500
400
15,000
Continuous power 300
10,000
200
5000 100
0 0
0 500 1000 1500 2000 2500 3000 3500 4000
Motor speed (rpm)

EMRAX 348 high voltage CC

efficiency map
Peak torque
1000
90–94%

800 94%
95%
Torque (Nm)

600
e
o rqu
us t 96%
nuo
400 Conti

200
86–90%
0
0 500 1000 1500 2000 2500 3000 3500 4000
Motor speed (rpm)

Fig. 8.39 Example of electric motor efficiency. Source: https://emrax.com/e-motors/

emrax-348/.
 CHAPTER 8 Thrust and Drag Modeling 335

Reciprocating Electric motor Gas turbine (turboprop)

Lycoming O-540-E4A5 EMRAX 348 low voltage Pratt&Whitney Canada PT6A-6

BHP = 260 HP (194 kW) SHP = 268 HP continuous, 536 HP peak (200–400 kW) SHP = 500 HP – 525 ESHP (373 kW)
Weight (dry) = 438 lb (199 kg) Weight (dry) = 88 lb (40 kg) Weight (dry) = 270 lb (122 kg)
Power to weight = 0.68 HP/lb (1.12 kW/kg) Power to weight = 3.05 to 6.09 HP/lb (5 to 10 kW/kg) Power to weight = 1.85 HP/lb (3.06 kW/kg)

Fig. 8.40 Comparing reciprocating, electric motor, and gas turbine engines.

Finally, with the proper motor controller, an electric motor can offer the
possibility to control not only RPM/power, but also spinning direction and
position vs time, allowing for synchronization of multiple systems. In some
cases, the electric motor can be used as a generator (e.g., using windmilling
force from the propeller).
From a modeling point of view, an electric motor converts electric power
Pin into a mechanical shaft power Pout . The power coming into the motor is a
product of the input voltage and input current
Pin ¼ IV ð8:77Þ
The mechanical shaft power produced is represented by a measure of
torque t and rotational speed v
Pout ¼ tv ð8:78Þ
Similar to a reciprocating engine, the rotational speed is often noted in
terms of revolutions per minute (RPM), which is transformed into the
rotational speed to be used in Eq. (8.78) as follows:
2p
v¼ RPM ð8:79Þ
60
The efficiency of an electric motor is defined as the ratio of the output
power to the input power
Pout
hm ¼ ð8:80Þ
Pin

Electric Motor Engines’ Impact on the Environment and CO2

When an electric motor is used in an all-electric configuration (batteries
containing the energy), it produces zero emissions in flight and no CO2 .
(This is not to say that the airplane is not polluting in some form if the
energy is, for example, generated by a coal plant.) This also made the
option very attractive as 2020 came along and the push for reduced aviation
emissions intensified (e.g., 2017 saw the start of the “flight shaming” or
flygskam movement in Sweden).
336 Operational Aircraft Performance and Flight Test Practices

What was not attractive about the pure electric system in the late 2010s
and early 2020s was the energy storage method, namely the batteries. Battery-
specific energy in early 2020 was of the order of 0.2 to 0.3 kWh/kg; thus, any
reasonable amount of energy stored on an airplane will result in a large
battery weight. This much larger weight will impact the drag and thus the
flight efficiency of the airplane. We will expand on the subject in Chapter 10.

Propeller Thrust: Actuator Disk Theory

In our discussion of thrust generation from the previously listed engines,
we have seen that all types need to transfer the converted fuel energy into a
propulsive force. The reciprocating engines and electric motor need to trans-
fer the power generated at a shaft to either a propeller, a rotor, or a ducted
fan, whereas the gas turbine could do the same if used as a turboshaft or tur-
boprop engine. We will develop models for propellers in this section and will
follow with ducted fans in the next section.
A simple model that can be used to give some insight into the ability of a
propeller to transfer energy from the shaft into the air stream is the actuator
disk theory (or momentum theory). The assumptions used to build this model
are:
• The propeller is replaced by an infinitely thin actuator disk of area A where
A ¼ 14p D 2 , D being the propeller diameter.
• The actuator has an infinite number of infinitely thin blades.
• The flow is inviscid, the disk offers no air resistance, and no energy is
dissipated by the disk.
• The rotation of the air going through the disk can be neglected.
• The velocity of the air through the disk is uniform across the entire surface
and equal to VP .
• The air mass going through the disk receives energy in the form of a
differential pressure Dp, also uniformly distributed across the area of
the disk.
• The air is assumed to be a perfectly incompressible fluid; the air density
remains constant.
As one can see, these assumptions show the basic limits of the model, but
the model that will be created offers a good insight into the behavior and
limitations of propellers. The mathematical model produced represents a
one-dimensional (1-D) flow in which all the air going through the disk is cap-
tured by a stream tube. A stream tube is a line in the flow where there is no
mass flow through the tube boundaries; therefore, the mass flow anywhere in
the stream tube is constant. The flow can be modeled as shown in Fig. 8.41.
Note that the air pressure far ahead of the disk p1 and far behind p4 are
equal to the atmospheric pressure at the flight altitude. As well, the airspeed
far ahead of the disk V1 is equal to the true airspeed of the airplane. The speed
of the airflow far behind the disk is called the slipstream VS .
 CHAPTER 8 Thrust and Drag Modeling 337

Stream tube

P1 P2 P3 P4

V1 V2 V3 V4

P3

P1 P4
P2

V4 = Vs
V1 = V V2 = V3 = Vp

Fig. 8.41 Actuator disk modeling.

The mass flow in the stream tube is constant and can be computed at the
disk where the area A is defined by the propeller diameter D. We get

p D2
ṁ ¼ r Vp A ¼ r Vp ð8:81Þ
4

As mentioned in the assumptions, the air density is constant and

equal to the air density at the flight altitude. The airspeed through the
disk, Vp , is a function of the energy imparted by the disk into the airflow.
Per our assumptions, the energy is inserted at the actuator disk in the form
of a differential pressure (Dp = p3 – p2 ). This Dp is applied over the entire
disk area to produce a forward force, the thrust T

T ¼ Dp A ð8:82Þ
338 Operational Aircraft Performance and Flight Test Practices

From Newton’s second law, this thrust applied to the flow generates an
airspeed increase DV in the form of
T ¼ ṁ DV ¼ ṁ ðVS  V1 Þ ¼ r Vp A ðVS  V1 Þ ð8:83Þ
With the assumption of incompressible flow, we can model the flow
ahead of the disk and behind the disk (but not through the disk) using
Bernoulli’s equation.
1 2 1
p1 þ r V1 ¼ p2 þ r Vp2 ð8:84Þ
2 2
1 1
p3 þ r Vp2 ¼ p1 þ r VS2 ð8:85Þ
2 2
or,
1  2 2
p3  p2 ¼ r Vs  V1 ð8:86Þ
2
Combining these equations yields an equation for the airspeed at the
actuator disk.
1
Vp ¼ ðV þ V1 Þ ð8:87Þ
2 S
These derivations and assumptions tell us that we should expect half of
the speed increase from far ahead of the propeller disk to the slipstream
far behind to occur by the time the air reaches the disk. Let’s define that
induced speed as v, which leads us to rewrite the previous equations as
Vp ¼ V1 þ v VS ¼ V1 þ 2 v ð8:88Þ
It also tells us the thrust is equal to
1  2
T¼ r A VS2  V1 ¼ r A ðV1 þ vÞ 2 v ¼ 2 ṁ v ð8:89Þ
2
From Eq. (8.89), one can see that the thrust increases as the speed
imparted to the airflow increases; the larger the difference in slipstream
speed to airplane airspeed, the larger the thrust. Figure 8.42 shows that the
required speed increase, v, becomes smaller as the flight airspeed gets larger.
Using these equations, one can write an equation for the induced speed v,
" sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#
1 2 þ2T
v ¼ V1 þ V1 ð8:90Þ
2 rA

Under steady state level flight, the flight power available is defined as the
thrust generated by the disk multiplied by the airplane airspeed, or
Pa ¼ T V1 ð8:91Þ
 CHAPTER 8 Thrust and Drag Modeling 339

30
D = 2.2 m
Standard sea level
25

V = 10 m/s
20
 (m/s)

15
V = 50 m/s
10
V = 100 m/s

5 V = 200 m/s

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10,000
T (N)

Fig. 8.42 Speed increase required, v, as a function of flight speed.

To produce this useful power, a certain amount of energy is inserted into

the flow at the actuator disk. This power is equal to

Pin ¼ T Vp ð8:92Þ

The ideal efficiency hi of the actuator disk is the ratio of the power avail-
able for flight to that at the disk.

Pa T V1 V1 1
hi ¼ ¼ ¼ ¼   ð8:93Þ
Pin T Vp V1 þ v 1 þ v
V1

The ideal efficiency can also be written in terms of the ratio of the slip
stream to the airspeed of the airplane as follows:

2
hi ¼   ð8:94Þ
VS
1þ V1

An interesting thing to note from Eq. (8.94) is that the ideal efficiency will
approach 100% as the speed imparted to the airflow approaches zero. But as
we have seen in Eq. (8.89), this also means the thrust will approach zero
under the same condition (similar to our review of the turbojet propulsive
efficiently earlier) (see Fig. 8.43). This is not a practical flight condition.
The ideal efficiency is the upper bound that one should expect to get from
the propeller. Once parasite and lift drags are added, that efficiency will
340 Operational Aircraft Performance and Flight Test Practices

reduce. A more detailed review of the propeller performance is then

necessary.
By observation, one sees that a propeller is composed of blades, which are
similar to wings, rotating about an axis and moving forward (see Fig. 8.44).

120%

V = 200 m/s
100%
V = 100 m/s
80% V = 50 m/s
Ideal efficiency

60%

40% V = 10 m/s
= 1.225 kg/m3
D = 2.2 m
20%

0%
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10,000
T (N)

100% 10,000

90% 9000

80% 8000
Thrust
70% 7000
Efficiency
Ideal efficiency

60% 6000
T (N)

50% 5000
= 1.225 kg/m3
40% 4000
D = 2.2 m
V = 10 m/s
30% 3000

20% 2000

10% 1000

0% 0
0 0.5 1 1.5 2 2.5 3
/V

Fig. 8.43 Efficiency of propeller defined by Fig. 8.42.

 CHAPTER 8 Thrust and Drag Modeling 341

2 n

Rotation axis

D 0.75 R Q 
Aero
 V force
Rotation plane 2 n r

Rotation axis


D  V

Rotation plane 2 n r

Fig. 8.44 Propeller geometry and forces.

One relates both motions through the nondimensional advance ratio J par-
ameter, a measure of the angle of attack in the propeller plane. Here, V is the
true airspeed of the airplane, D is the propeller diameter, and n is the revolu-
tions per second. This is a nondimensional parameter, so one should care-
fully select the units of the true airspeed and diameter.
V
J¼ ð8:95Þ
nD
Each blade is composed of a series of airfoils at varying pitch angles (u,
angle between the rotation plane and the local chord). The blade section at
a given radius r sees the rotation speed (2p n r) and the induced speed at
the propeller plane. The local angle of attack seen by the blade is a function
of these two speed components and the local blade pitch angle. The blade
twist is a function of the radius from the propeller hub, being small at the
blade tip and larger near the hub. The propeller pitch is typically defined
at the 75% radius because it is typically the most loaded part of the propeller
blade because the rotational speed is very low at the hub and increases pro-
portional to the radius, but like a wing, we know that the lift generated at the
tip is zero (see Fig. 8.44). We also observe that the aerodynamic force that was
expressed in terms of L/D for a wing can be broken down into a thrust T
component and a torque Q component for a propeller.
The propeller coefficients of thrust CT and torque CQ are defined in
terms of propeller geometry and rotation speed as
T
CT ¼ ð8:96Þ
r n2 D4
Q
CQ ¼ ð8:97Þ
r n2 D5
The capability of a propeller is typically presented in the form of a thrust
coefficient as a function of the advance ratio (see Fig. 8.45). From this chart,
342 Operational Aircraft Performance and Flight Test Practices

0.14
Stall-like
behavior
0.12
Decreasing air density
Increasing thrust
Decreasing RPM

0.10
45° Blade angle of 0.75 R
0.08 40°
CT

35°
0.06 30°
25°
0.04
20°
15°
0.02

0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
V/nD
Increasing forward speed
Increasing rotation speed Lower angle of attack
Increasing angle of attack

Fig. 8.45 Example of thrust coefficient as a function of advanced ratio.

Source: Adapted from [4].

we see that as the advance ratio increases (increasing airspeed or reducing

propeller RPM), the thrust coefficient tends to decrease in a more or less
linear behavior. As the advanced ratio reduces (reduced airspeed or increased
propeller RPM), the thrust coefficient increases but then flattens, similar to a
wing stall. We also note that to have a higher thrust coefficient at a higher
advance ratio, one needs to increase the propeller blade angle. One is essen-
tially adjusting the propeller angle of attack by changing the geometric pitch.
The power provided to the propeller by an engine (reciprocating, electric,
gas turbine, etc.) is a function of the torque and the RPM. Mathematically,

Pin ¼ 2p n Q ð8:98Þ

Pin
CP ¼ ð8:99Þ
r n3 D5
And we also note that

CP ¼ 2p CQ ð8:100Þ

Figure 8.46 shows a typical representation of the variation of the power

coefficient with advance ratio and adds lines of constant thrust coefficient
for reference. The figure also shows the same propeller at various pitch
angles. We note that there are conditions where a positive power coefficient
will lead to negative thrust coefficient—the propeller is actually producing
drag.
With this sample performance chart, we turn our attention back to the
propeller efficiency. We defined the efficiency as the ratio of the flight
power to the power being provided to the propeller. Using the coefficients
 CHAPTER 8 Thrust and Drag Modeling 343

0.18
0.16 5
.11 11
. 10
0.14 .
.09
0.12 .08
.07
0.10 .06 CT
.05
CP

0.08 .04
0.06 .03
.02
0.04 .01

0.02 0
15° 20° 25° 30° 35° Blade angle at 0.75 R
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8
V/nD

Fig. 8.46 Typical relationship between the power coefficient and the advance ratio. Source: [4].

defined earlier we get a “real” propeller coefficient hp

T
V
TV r n2 D4 C
hp ¼ ¼ ¼ T J ð8:101Þ
Pin Pin CP
nD
r n3 D5
The information contained in Figs. 8.45 and 8.46 can be remapped to give
the efficiency of the propeller vs advance ratio and blade angle (see Fig. 8.47).
Notice that the propeller efficiency can be maximized over a large advance
ratio range if one has the ability to modify the propeller pitch in flight. An
airplane equipped with a fixed-pitch propeller (many small general aviation
airplanes with limited speed range) would need to optimize the blade angle

1.0

0.8

0.6

0.4
Blade
0.2 angle
15° 20° 25° 30° 35° 40° 45° at 0.75 R

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
V
nD

Fig. 8.47 Propeller efficiency vs advance ratio. Source: [4].

344 Operational Aircraft Performance and Flight Test Practices

to meet the intended flight envelope of the airplane; one would either opti-
mize takeoff performance (finer pitch) and have a very limited speed envelope
or have a better speed envelope (coarser pitch) but reduced takeoff perform-
ance. A variable pitch propeller allows the blade pitch to change in flight,
either manually (pilot control in the cockpit) or automatically (computer
controlled, mechanical control). Some propellers are designated constant
speed props in that the propeller pitch is adjusted automatically to maintain
RPM. Such props need to be coupled with a constant speed engine (e.g., gas
turbine/turboprop) where the engine control changes the torque by varying
the fuel flow. Here the pilot controls the engine torque.
Equation (8.101) can also be written in terms of torque provided to the
propeller as follows:

1 CT
hp ¼ J ð8:102Þ
2p CQ

One can create a map of propeller efficiency as in Fig. 8.48, which relates
power and efficiency to advance ratio. This format helps to rapidly find the
best combination of propeller blade pitch and RPM for maximum efficiency.
The propeller thrust can be increased while maximizing the efficiency
by increasing the propeller diameter [increased mass flow through the disk
via larger disk area, Eq. (8.81), while minimizing impact on v]. There is,
however, a limit to this approach. Small airplanes will need propeller geo-
metric clearance with the ground, for example. Also, increasing the propeller
diameter leads to increased rotational speed at the propeller tip. For airplanes
with a higher cruise speed, this higher tip speed combined with higher flight

50°

Propeller pitch 45°

@ 0.75R p
20% 30% 40% 50% 60% 70%
40° 75%
80%
35°
Cp

85%
87%
30° 88%
89%
25°

20°
15°

Fig. 8.48 Propeller efficiency map.

 CHAPTER 8 Thrust and Drag Modeling 345

1.1

max
i 1.0
relative efficiency,

0.75R
0.9
Maximum

35°
40°
45°
0.8 50°
55°
60°
0.7
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Tip Mach number, Mt

Fig. 8.49 Effect of compressibility on relative maximum efficiency. Source: [5].

speed may lead to Mach issues. The propeller tip Mach number is defined as
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V 2 þ ½ðD=2Þ 2p n2
Mtip ¼ ð8:103Þ
a
This is not necessarily the Mach number with the induced speed at the
propeller plane, but rather a freestream reference; however, it is a reasonable
value to use because we have seen that the induced speed does reduce with
increasing airspeed to produce a given thrust, and this reference is not engine
power dependent. What is observed for propellers that can maintain essen-
tially constant propeller efficiency at lower Mach numbers is a reduction in
the peak efficiency when the propeller tip gets close to Mach 1 and shock
waves form on the blades. NACA performed supersonic propeller research
in the 1950s and found that propellers tapering from about 5% thickness
ratio at the base to 2% at the tip with either zero or very small cambers
yielded efficiencies of 75–80% at a forward Mach number of 0.9. At Mach
1, peak efficiencies as high as 0.75 were obtained (see Fig. 8.49). The infor-
mation was used by some manufacturers in the 1950s on some jets (see
Fig. 8.50). To minimize the impact of shock waves on the tip, the diameter
of the propeller was minimized, and the blade chords were increased to
absorb the power provided to the propeller; the propeller was designed for
a high advance ratio.
The activity factor of a propeller blade represents the ability of the blade
to absorb power. The local dynamic pressure on the blade is proportional to
the square of true airspeed and rotational speed (n r).
1  
qr ¼ r V 2 þ ðn r Þ2 ð8:104Þ
2
346 Operational Aircraft Performance and Flight Test Practices

XF-84H – 1955 XF-88B – 1956 GE UDF – 1980s

Fig. 8.50 High-speed propellers.

The activity factor is the integration of the blade chord multiplied by its
radial position cube, as follows:

ð
1:0R
105
AF ¼ 5 c r 3 dr ð8:105Þ
D
0:15R

The ability of a propeller to absorb power is then proportional to the

number of blades (B) times the activity factor of the blades (AF).
5
D
P  4 r p3 n3 cd B AF ð8:106Þ
10

The larger the activity factor, the more blade area there is near the pro-
peller tip where most of the aerodynamic forces are developed (see
Fig. 8.44 earlier in the chapter).
The best example of the use of activity factor this author knows is the
story of the evolution of the C-130 Hercules (see Fig. 8.51). The airplane
was designed in the early 1950s with wide-chord, four-bladed model
54H60 constant speed propellers to absorb the power of the four Allison
T56-A-15 turboprop engines generating nearly 4590 SHP (3420 kW) each.

AF ≈ 80 AF ≈ 110 AF ≈ 162

C130H upgrade C130J C130H

NP2000 R391 54H60-91

Fig. 8.51 Change in propellers on C-130 Hercules over time.

 CHAPTER 8 Thrust and Drag Modeling 347

The last of the early models produced was the C-130H variant, last produced
in the mid-1990s. This was followed by the C-130J model in the 1990s, which
has six-bladed model R391 propellers fitted to a Rolls-Royce AE 2100D3 tur-
boprop generating 4637 SHP (3458 kW), essentially the same power as pre-
vious engines. In the 2010s, some older C-130H’s were being equipped with
new eight-bladed model NP2000 propellers to improve field performance. All
propellers have the same diameter, and all absorb about the same power.
The aim of creating a thrust model for propeller-equipped airplanes is to
create one that best integrates the airplane excess thrust (T – D) to match the
performance expected and demonstrated. This may require the analyst to
either fit the model over a smaller speed range or add terms to the thrust
model to best fit the data collected.

Adaptation of Actuator Disk Theory to Ducted Fan

A ducted fan operates in a similar way to the fan on a turbofan, but in this
case, the power provided to the fan comes from an engine not necessarily in
series with the fan. The fan (typically with more blades than a propeller) is
contained in a duct. Power is fed by either an indirect shaft from a turboshaft
engine or a reciprocating engine, or even from an electric motor into the hub
of the fan. The ducted fan may also have a stator behind it to recover some of
the swirl.
We set up the analysis like we did the actuator disk for the propeller. Here
again, the energy is added in the form of a pressure rise DP. For the following
analysis, we set the exhaust area Ae to be equal to the fan disk area Af to allow
comparison to the propeller analysis (see Fig. 8.52). We also neglect internal
friction drag generated by the flow.
p
A e ¼ A f ¼ D2 ð8:107Þ
4
In this scenario, the pressure in the exhaust plane is equal to the atmos-
pheric pressure; thus, the airspeed from just behind the disk to the exhaust
plane is constant and equal to the fan airspeed Vf . The mass flow through
the system is
ṁ ¼ r Af Vf ð8:108Þ
The thrust produced is
T ¼ Dp Af ð8:109Þ
and
   
T ¼ ṁ DV ¼ ṁ Vf  V1 ¼ r Af Vf Vf  V1 ð8:110Þ

Per the propeller analysis, we now define the induced speed v as

v ¼ Vf  V1 ð8:111Þ
348 Operational Aircraft Performance and Flight Test Practices

which leads to

T ¼ ṁ v ð8:112Þ

Solving for the induced speed,

" sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#
1 2 þ 4T
v ¼ V1 þ V1 ð8:113Þ
2 r Af

Note that for a given fan disk and given thrust, the induced flow from the
ducted fan will be larger than that of a propeller of the same diameter produ-
cing the same thrust. Figure 8.53 shows the ratio of the ducted fan induced
speed from Eq. (8.113) to the induced speed of the propeller from Eq. (8.90).

Stream tube
P1 P2 P3 P4

V1 V2 V3 V4

P4
P1 P

V4 = Vs
V1 = V V2 = V3 = Vf

Fig. 8.52 Actuator disk applied to ducted fan.

 CHAPTER 8 Thrust and Drag Modeling 349

2.2
Diameter = 5 ft
ISA SL conditions
Induce speed ratio (ducted fan/propeller)

2
V = 400 ft/s
V = 200 ft/s
1.8

V = 100 ft/s
1.6

V = 0 ft/s
1.4

1.2

1
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Thrust (lb)

Fig. 8.53 Induced speed ratio, ducted fan to propeller of same diameter.

Advantages of the ducted fan include:

• Reduced fan tip losses with the presence of the duct compared to propeller
tip losses.
• Ducted fans are quieter than propellers because the duct provides some
sound attenuation.
• For the same input power and thrust, the ducted fan is smaller than
the propeller.
• With proper sizing of the duct, the ducted fan can be used to higher Mach
numbers (e.g., on turbofans).
The disadvantages of ducted fans include:
• The duct installation is much heavier than an equivalent thrust propeller.
• The ducted fan operates at a higher RPM than a propeller for a
given thrust.
• At a high angle of attack, part of the duct can have flow separation.

Thrust Summary
For performance analysis, one must understand the impact of the thrust
lapse rate (change with airspeed) of the propulsion system used. We pre-
sented two extreme conditions—constant thrust with airspeed and constant
350 Operational Aircraft Performance and Flight Test Practices

power with airspeed (see Fig. 8.54). Most propulsion systems will fall some-
where in between these two extremes. It is important for the performance
engineer to fully understand the expected behavior of the system installed
on the airplane.
All air-breathing propulsion systems will be greatly impacted by air
density, a combination of pressure altitude and temperature at altitude.
We have used this as part of the basic modeling of those propulsion systems.

Thrust-Drag Bookkeeping
We have created both drag and thrust models for the purpose of analyz-
ing the performance of a plane. On their own, the models for each force com-
ponent provide reasonable trends and expected behavior. The production of
thrust, however, which involves imparting energy to the airflow around the
airplane, will have some impact, sometimes significant, on the aerodynamic
forces acting on that airplane and thus on the drag force (see Fig. 8.55).
For example, for underwing-mounted turbofans, some of the exhaust flow
coming out of the engine will rub on the engine pylon. Is this extra drag or
less thrust than a comparable engine without the presence of the pylon?
More air may be channeled between the wing and the engine nacelle with
increased thrust, which may impact the pressure distribution on the wing
and thus its drag. Is this change in force attributed to the engine (reduced
thrust) or the airframe (increased drag)? On rear fuselage–mounted
engines, the change of airflow around the airplane tends to be less in
forward flight. That configuration may behave closer to the drag and thrust
models with no interference.

Constant thrust assumption (turbojet)

Thrust, drag

Low bypass turbofan

Drag
High bypass turbofan

Turboprop
Constant power assumption
(piston-prop)

VSR VDmin

Airspeed

Fig. 8.54 Establishing thrust and drag models.

 CHAPTER 8 Thrust and Drag Modeling 351

Away from
tail

Little aero
impact
Flow entrainment

Extra drag due to increased

airspeed by engine exhaust

Fig. 8.55 Engine thrust production on the airplane drag.

Reciprocating engines typically have a cooling flap that helps draw more
air into the engine compartment to control the engine temperature. That
cooling flap opens when the engine gets hotter (manually or automatically)
and impacts the drag of the plane. Do you attribute this extra drag to the air-
frame or treat it as a reduced thrust?
It is important to establish the guidelines on how to treat each component
of drag and thrust and account for each one only once in the total equation;
this is the basis of thrust-drag bookkeeping. Once the methodology is estab-
lished, care must be taken to follow it during the performance analysis for
an airplane.
Consider the concept of excess thrust when performing the analysis of
the thrust and drag of an airplane. Excess thrust is simply thrust minus
drag (T – D). If the difference is positive, the airplane will accelerate and/
or climb. If the difference is negative (excess thrust deficit), the airplane
will slow down and/or descend.

Boundary Layer Ingestion (BLI)

In the late 2010s, NASA proposed a concept to integrate an engine into
the wake of the fuselage of an airliner as a way to improve (reduce) fuel con-
sumption. The general idea is that the engine would ingest the boundary layer
from the fuselage (slower moving air) and would thus require less power to
generate a given amount of thrust. As well, the engine would pull on the
slower moving boundary layer airflow at the back of the airplane and help
reduce the drag of the airplane. One can easily see how this highly integrated
airframe-engine would need careful thrust-drag bookkeeping to create an
overall airplane performance model.
The initial research from NASA, the STARC-ABL concept, is meant to
bridge the gap between current jet fuel–powered aircraft and future all-
electric vehicles (see Fig. 8.56). It is a 150-passenger commercial transport
with a traditional “tube-and-wing” shape. The plane relies on turboelectric
propulsion, meaning that it uses electric motors powered by onboard gas
turbines.
352 Operational Aircraft Performance and Flight Test Practices

Fig. 8.56 NASA’s STARC-ABL. Source: NASA: https://sacd.larc.nasa.gov/asab/

asab-projects-2/starc-abl/.

Excess Thrust Manager (ETM)

This author applied for a patent on the Excess Thrust Manager (ETM) in
2020 while working at Bombardier Aviation. The ETM is meant to be a
hybrid-electric system with multiple electric propulsors in the tail of the air-
plane fed by either the airplane’s APU or the main engines, depending on the
flight phase. The ETM is meant to supplement thrust when needed but also
to control flight path by managing the airflow around the airplane. It can
produce drag (reverse the flow of the propulsors to help the airplane
descend instead of using wing spoilers that also kill lift), reduce drag
(control the thick boundary layer of the aft fuselage), and increase forward
thrust (produce more forward force than required to bring the boundary
layer air back to flight speed). The ETM can also produce reverse thrust
on the ground. In other words, the ETM can manage the airplane’s excess
thrust. It is meant to be controlled by the pilot via the traditional controls,
modulated by an onboard computer.

No propeller
Propellers on

Rotation Rotation

Fig. 8.57 Effect of propeller wash on spanwise lift distribution.

 CHAPTER 8 Thrust and Drag Modeling 353

CL Advance ratio CL J2
J2 L/D
J2< J1
J1 J1
Idle Idle

 CD

Fig. 8.58 Impact of propeller flow (various propeller advance ratios) on airplane aerodynamics.

We will address excess thrust testing further in Chapters 11 (cruise), 12

(climb performance), and 13 (OEI climb performance).

Propeller Installation and Aerodynamic Impact on the Airplane

Airplane designers strive to make the airplane symmetric to ease handling
qualities and flight control use. Contrary to jet engines, where the impact of
producing thrust on the aerodynamics can be minimized (not completely
removed), propellers will often make the flow asymmetric, and propeller-
driven airplanes with a large propeller disk with respect to the airplane size
will have their aerodynamics significantly impacted (see Fig. 8.57). Two
major impacts are:
1. Increased dynamic pressure in the airflow behind the propeller.
• But also increased suction in front of the propeller that will impact the
flow ahead of the propeller.
• The increased dynamic pressure [proportional to the induced speed,
Eq. (8.90)] on the wing depends on the distance of the propeller plane to
the wing—the closer it is, the more interference.
2. Change in the angle of attack due to the propeller disk imparting energy to
the airflow and due to the propeller swirl. With the latter, the upgoing
propeller will increase the angle of attack behind it while the downgoing
propeller will reduce the angle of attack behind it.
These impacts will be larger at slower airspeeds and higher angles of
attack (see Fig. 8.58).

References
[1] Hoerner, S. F., Fluid–Dynamic Drag, self-published, 1965.
[2] Roskam, J., Airplane Design, Part I: Preliminary Sizing of Airplanes, 2nd ed., self-
published, 1989.
354 Operational Aircraft Performance and Flight Test Practices

[3] ASTM International, “Standard Specification for Design of Electric Propulsion Units
for General Aviation Aircraft,” ASTM F3338-18, 2018.
[4] Hartman, E. P., and Biermann, D., “The Aerodynamic Characteristics of Full-Scale
Propellers Having 2, 3, and 4 Blades of Clark Y and R.A.F. 6 Airfoil Sections,”
NACA TR640, 1938.
[5] Stack, J., Draley, E. C., Delano, J. B., and Feldman, L., “Investigation of the NACA
4-(3)(08)-03 and NACA 4-(3)(08)-45 Two-Blade Propellers at Forward Mach
Numbers to 0.725 to Determine the Effects of Compressibility and Solidity Perform-
ance,” NACA-R-999, 1950.
Chapter 9 Flight Envelope

Chapter Objective
An airplane’s flight envelope is the boundaries in the sky within which the air-
plane has been shown to operate safely. These limits are defined by minimum
and maximum airspeed at a given altitude. The flight envelope can also be
defined by temperature limits. The maximum/minimum airspeeds and alti-
tudes an airplane can reach based on available excess thrust may not be the
maximum/minimum authorized. Indeed, as we will see, the flight envelope
of Part 25 airplanes is heavily regulated.

A Regulation-Based Limit

T
ransport category airplanes (FAA Part 25) have been shown to be
able to safely operate within a certain set of boundaries in the
skies. This is best exemplified by the very low accident rate of com-
mercial airplanes while operating more than a 100,000 flights per day world-
wide. The 2017 FAA report “Air Traffic by the Numbers” [4] shows that the
national airspace (NAS) contains more than 2.5 million passengers daily on
scheduled flights (see Fig. 9.1).
The Part 25 airplane flight envelope, the boundaries, is dependent on
many variables, including:
• Airplane configuration
• Gear position, if retractable
• Flaps/slats positions
• Weight and center of gravity location
• Speed brake position
• Throttle position
• Atmospheric conditions
• Temperature
• Altitude (pressure)
• Humidity
• Icing and/or rain
• The manufacturer’s demonstrated certification limits

355
 356
Operational Aircraft Performance and Flight Test Practices
IFR flights handled by the ATO VFR Flights*
16,800,000 46,000 11,600,000 32,000
16,600,000 45,500 11,400,000 31,500
31,000

Average daily flights

45,000

Average daily flights

16,400,000 11,200,000
44,500 30,500
Yearly flight

Yearly flight
16,200,000 11,000,000
44,000 30,000
16,000,000 10,800,000 29,500
43,500
15,800,000 10,600,000 29,000
43,000
15,600,000 10,400,000 28,500
42,500
28,000
15,400,000 42,000 10,200,000 27,500
15,200,000 41,500 10,000,000 27,000
15,000,000 41,000 9,800,000 26,500
09

10

11

12

13

14

15

16

15
09

10

16
11

12

14
13
FY

FY

FY

FY

FY

FY

FY

FY

FY
FY

FY

FY
FY

FY

FY
FY
OPSNET reports VFR activity as total operations (arrivals + departures). Total VFR flights are approximated by dividing
total operations by 2.
Trends in flights and available seat miles (ASMs)
11,000 1700
Thousands

Billions
1650
10,500 Passenger statistics
1600
10,000 1550 FY 2015 FY 2016

ASMs
1500 Yearly passengers 912,485,113 946,846,490
Flights

9500
1450 Average daily passengers 2,499,959 2,587,012
9000 1400
Revenue passenger miles (trillions) 1.30 1.37
1350
8500 Available seat miles (trillions) 1.58 1.67
1300
8000 1250 Passenger load factor (%) 82.46% 82.18%
10

11

12

13

14

15

16
09

FY

FY

FY

FY

FY

FY

FY
FY

Flights Available seat miles

Fig. 9.1 NAS flights, 2017 FAA report.

 CHAPTER 9 Flight Envelope 357

The performance-related limits of interest for this book and chapter are
defined by the following regulations:
• §25.103 Stall Speed
• §25.107 Takeoff Speeds
• §25.125 Landing
• §25.143 General (Controllability and Maneuverability)
• §25.149 Minimum Control Speed
• §25.161 Trim
• §25.201 Stall Demonstration
• §25.203 Stall Characteristics
• §25.207 Stall Warning
• §25.233 Directional Stability and Control (Ground)
• §25.237 Wind Velocities
• §25.251 Vibration and Buffeting
• §25.253 High-Speed Characteristics
• §25.255 Out-of-Trim
• §25.335 Design Airspeeds
• §25.345 High Lift Devices
• §25.629 Aeroelastic Stability Requirements (Flutter)
• §25.1505 Maximum Operating Limit Speed
• §25.1511 Flap Extended Speed
• §25.1515 Landing Gear Speeds
• §25.1517 Rough air speed, VRA
• §25.1527 Ambient Air Temperature and Operating Altitude
• §25.1563 Airspeed Placard
• §25.1583 Operating Limitations

These are the more direct limitations specified in the regulations; other
regulations may impose an indirect limit on speed, altitude, or temperature
because a criterion is not met. We will address the field performance
speeds specifically in Chapters 15–20.
From this, we can define a flight envelope. We will address that envelope
from a low-speed and high-speed point of view first.

Low-Speed Limit
In Chapter 7, we discussed the minimum speed the airplane can maintain
in level flight, the stall speed Vstall . That speed was defined in several ways,
from the minimum speed where 1-g flight could be maintained (VS1g ) to
the reference stall speed VSR of §25.103 at Amendment 108 and beyond to
the pilot-identified stall speed Vs of earlier amendments. With this speed
defined, one could extract an equivalent maximum usable lift coefficient
358 Operational Aircraft Performance and Flight Test Practices

CLmax that could be used for performance

analysis.
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
CLmax

u W
Vstall ¼u
t1 (9:1)
r s S CLmax
2 SL

Mach We did not include the load factor in Eq.

(9.1) because the CLmax computed contains
the effect of the load factor for Eq. (9.1)
Fig. 9.2 CLmax vs Mach. (either defined by the 1 g stall condition or
minimum speed stall condition).
The minimum operational speeds for
takeoff and landing, to be covered in Chapters 15 to 20, are typically
ratioed from this stall speed. The ratio having been historically proven to
provide acceptable safety margins.
In Chapter 7, we also saw that the CLmax may also vary as a function of the
airplane weight and altitude (W/d). This effect can be represented as CLmax
as a function of Mach number.
We can rewrite Eq. (9.1) in terms of Mach number to use the information
from Fig. 9.2.
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u ðW =dÞ
Mstall ¼u
t1 (9:2)
g pSL S CLmax
2

Combining this information onto a single chart (Fig. 9.3) shows that
the minimum Mach number that should be expected at a given
altitude will increase with increasing altitude. Of course, we remember that
these stall speeds were demonstrated per §25.201 and were found to be
acceptable per §25.203. The speeds in the aircraft flight manual (AFM) also
represent the speeds for the most adverse center of gravity (CG); therefore,
the airplane may be capable of slightly
W1 < W2 < W3 slower speeds than those provided in the
manual.
Operationally, the airplane is not expected
Altitude

to reach the published stall speed unless it

deviates significantly from designated oper-
ational speeds. FAA 14 CFR §25.207 requires
that the pilot be provided with a stall
warning (SW) at a speed sufficiently high
Mach
(VSW ) to ensure the airplane can be recovered
prior to reaching stall. This stall warning can
Fig. 9.3 Variation in stall speed come in the form of natural airframe buffet
with altitude and weight. or by artificial means like a stick shaker.
 CHAPTER 9 Flight Envelope 359

40,000
Note: the speeds on this chart VSR
35,000 assume the same aircraft
weigh at all altitudes. Vbuff
30,000

25,000
Altitude (ft)

Vmin VSW
20,000

15,000

10,000
5000

0
0 50 100 150 200 250 300 350
Airspeed (KTAS)

Fig. 9.4 Vmin vs low speed buffet/stall warning/VSR as a function of altitude.

The minimum acceptable margin, per §25.207(d), between stall warning

and the reference stall speed VSR , as defined by §25.103, is 3 kt or 3%, which-
ever is highest. Stall warning also needs to be 5 kt or 5% of the pilot-identified
stall speed VSid per §25.201(d).

VSW  1:03 VSR VSW  1:05 VSid (9:3)

These speeds are based on the airplane’s lift generation capability. The
minimum flight speed can also be thrust limited. As we saw in Chapter 8,
the airplane drag coefficient at low speed is proportional to the lift coefficient
squared; the drag rapidly increases as you slow down. This condition where
the airplane generates enough lift but not enough thrust to maintain speed in
level flight typically does not occur at low altitudes where excess thrust is
usually sufficient, but at high altitudes there could be a corner of the envelope
(e.g., see Fig. 9.4) where the condition would occur. When the condition is
reached, the airplane does not immediately fall out of the sky; rather, the air-
plane will start a slow deceleration that will accelerate if no preventative
measures are applied. The airplane could then slow down to stall, such as
what occurred for West Caribbean Airways Flight WCW708 in 2005.

Venezuela Report JIAAC-9-058-2005: On 16 Aug. 2005, a McDonnell

Douglas MD-82, Flight WCW708, departed Panama City (Panama) bound
for the Caribbean island of Martinique. It departed controlled flight after
reaching cruise altitude and a slow speed decay that led to stall.
360 Operational Aircraft Performance and Flight Test Practices

The sequence of events that led to the accident included a cruise at FL310
followed by a request to climb to FL330 on autopilot and autothrust. The air-
plane climbed to FL323 with intermediate steps (airplane did not have enough
thrust to climb directly to FL330). While climbing, the speed reduced. The
pilots then selected autopilot “vertical speed” mode for a fixed rate of climb
and deactivated anti-ice (more thrust). The airplane finally reached FL330,
but the speed had reduced to M0.70 (minimum drag estimated at M0.73),
and the autopilot retrimmed the stabilizer nose up to compensate. Anti-ice
then reselected on (thrust reduction), and the airplane started a slow decelera-
tion. By the time the airplane reached M0.60, it started losing altitude rapidly.
By the time the airplane crossed FL317, the sink rate was 2500 fpm. The stall
warning activated and continued until impact, with the airplane in a deep stall.

8000 1.0

NTSB simulation
Boeing stability & control required thrust
Boeing performance required thrust
7000 FDR Mach no.
0.8
Thrust per engine, Fn (pounds)

6000

Mach
0.6

5000

0.4
4000

3000 0.2
06:42:00 06:44:00 06:46:00 06:48:00 06:50:00 06:52:00 06:54:00 06:56:00 06:58:00
Time (UTC)

Boeing noted that anti-ice bleed extraction had a large EPR (thrust) debit
on the JT8D, as much as a 3000-ft impact on cruise altitude capability.
Turning on the anti-ice system lowered the thrust below that required to
maintain level flight.

Maneuver Margin
The requirement to demonstrate acceptable maneuver margin at
minimum operational speeds was added to §25.143(g) at Amendment 108
of Part 25 (later moved to §25.143(h) in Amendment 121, in 2007). The
requirement was written in terms of demonstrating a minimum bank angle
capability for a critical takeoff or landing configuration and most adverse
weight and CG combination expected in service. These angles are shown
in Table 9.1.
 CHAPTER 9 Flight Envelope 361

Table 9.1 Maneuver Margins per §25.143

Maneuvering Bank in
Configuration Speed a Coordinated Turn Thrust Power Setting
Takeoff V2 30 deg Asymmetric WAT-limited
Takeoff V2 þ XX 40 deg All-engine operating climb
En route VFTO 40 deg Asymmetric WAT-limited
Landing VREF 40 deg Symmetric for –3-deg flight path

The notice of proposed rulemaking (NPRM) for Amendment 108 speci-

fied the following (underline added by author):

The FAA proposes to require a minimum bank angle capability in a coordi-

nated turn without encountering stall warning or any other characteristic
that might interfere with normal maneuvering. This requirement would be
added to Sec. §25.143 as a new paragraph (g). The proposed minimum
bank angles were derived by adding a 15-degree allowance for wind gusts
and inadvertent overshoot to a maneuvering capability the FAA considers
necessary for the specific cases identified in the proposed new paragraph.
These proposed maneuver margin requirements are intended to ensure
that the level of safety in maneuvering flight is not reduced by the proposed
change to the reference stall speed and the reduction in the multiplying
factors used to determine the minimum operating speeds.

When the FAA proposed this, it fell short of specifying what “any other
characteristic that might interfere with normal maneuvering” meant,
instead leaving the manufacturer to define what it meant for them on a
case-by-case basis. Some manufacturers have used conservative definitions
for this limit (especially when defining the limit while using an envelope pro-
tection from a fly-by-wire system).
The FAA did provide some additional clarification as to the tolerance of a
system providing such protection (underline added by author):
Because the proposed requirements already provide the capability to over-
shoot the intended bank angle by 15 degrees, the small differences in the
speed at which the stall warning system operates due to system tolerances
are not as critical. Therefore, the FAA intends for the minimum bank
angles in the proposed Sec. §25.143(g) to apply at the designed nominal
setting of the stall warning system. To ensure that large production toler-
ances do not adversely impact the airplane’s maneuvering capability free of
stall warning, the bank angle capability specified in the proposed Sec.
§25.143(g) should not be reduced by more than two degrees with the stall
warning system operating at its most critical tolerance. Applicants would
be expected to demonstrate this capability either by flight test with the
system set to its critical tolerance, or by analytically adjusting flight test
data obtained at some other setting.

In the end, the intent of the requirement is to ensure that the minimum oper-
ational speeds for takeoff and landing have an adequate margin away from
362 Operational Aircraft Performance and Flight Test Practices

stall warning to prevent stalling the airplane. We will cover maneuver margin
in more details in Chapter 14.
Amendment 121 to Part 25 required that the margin also apply under
icing conditions. This amendment came following some icing stall and loss
of control events. One such event was the crash of an Embraer EMB-120 Bra-
silia (NTSB report A-98-094) where the pilot lost control of the plane while
maneuvering in normal, nonice conditions.

NTSB A-98-094: On 9 Jan. 1997, an Empresa Brasileira de Aeronautica, S/A

(Embraer) EMB-120RT, operated by COMAIR Airlines, crashed during a
rapid descent after an uncommanded roll excursion near Monroe, Michigan.
The flight departed Covington with 2 flightcrew, 1 flight attendant, and 26
passengers on board. There were no survivors. The airplane was destroyed
by ground impact forces and a postaccident fire. IMC prevailed at the time
of the accident, and the flight was operating on an IFR flight plan.
The probable cause of this accident was the FAA’s failure to establish ade-
quate aircraft certification standards for flight in icing conditions. NTSB rec-
ommendation to the FAA: Require manufacturers of all turbine-engine–
driven airplanes (including the EMB-120) to provide minimum maneuvering
airspeed information for all airplane configurations, phases, and conditions of
flight (icing and nonicing conditions); minimum airspeeds also should take
into consideration the effects of various types, amounts, and locations of ice
accumulation, including thin amounts of very rough ice, ice accumulated in
supercooled large droplet icing conditions, and tailpane icing.

Buffet Boundary: Low Speed

Buffeting is airframe vibration induced by flow separation on the aircraft
(mostly from the wing) with increasing angle of attack and/or Mach. At the
lower airspeeds, this phenomenon is tied to the stall and high angles of attack.
It can be coincident with the stall or
Low-speed start at a slightly higher airspeed. At
buffet Typical buffet the higher altitudes, low speed buf-
limit feting may actually start before artifi-
cial stall warning (stick shaker).
nz The buffet limit for an airplane is
CLbuff

1-g High-speed often represented by the maximum

flig buffet
ht lift capability CLbuff as a function of
Mach number. The shape of the
buffet curve on Part 25 transport cat-
Mach
egory airplanes will typically be fairly
flat at the low Mach number side of
Fig. 9.5 Typical Part 25 transport category the curve (see Fig. 9.5). That section
airplane buffet envelope. is limited by flow separation at high
 CHAPTER 9 Flight Envelope 363

angles of attack. On the high Mach number side, the maximum lift coefficient
free of buffet will rapidly reduce because shock wave flow separation will
trigger buffeting even at low angles of attack.
We have added the curve for the required lift coefficient for 1-g flight
[Eq. (9.4)] to Fig. 9.5 and note that this curve intersects the buffet curve in
two locations: a low-speed condition and a high-speed condition. The low-
speed condition is the stall, which we covered extensively in Chapter 7. We
will expand the high-speed side later in this chapter.
nz ðW =dÞ
CL ¼ 1 (9:4)
2 g pSL M2 S
The airplane maneuvering capability without encountering buffet at a
given Mach number will then be the ratio of the buffet lift coefficient to
that of the airplane in level flight at the same Mach number.
CLbuff
nz ¼ (9:5)
CL1g

Takeoff and Landing Speed Restrictions

Although we will cover the takeoff and landing speeds extensively in the
takeoff and landing chapters (Chapters 15–20), we introduce them here to
discuss some of the restrictions on those speeds that may not be covered
in future chapters. The applicant (airplane manufacturer) typically wants
the lowest takeoff and landing speeds possible for a given flap configuration
because the takeoff and landing distances are approximately proportional to
the square of the airspeed.
Over the years, safety margins on those speeds have been established and
are reflected in the certification regulations. For takeoff, the speeds are
defined by §25.107:
§25.107 Takeoff speeds.
(a) V1 must be established in relation to VEF as follows:
(1) VEF is the calibrated airspeed at which the critical engine is assumed
to fail. VEF must be selected by the applicant, but may not be less
than VMCG determined under §25.149(e).
(2) V1, in terms of calibrated airspeed, is selected by the applicant;
however, V1 may not be less than VEF plus the speed gained with
critical engine inoperative during the time interval between the
instant at which the critical engine is failed, and the instant at which
the pilot recognizes and reacts to the engine failure, as indicated by
the pilot’s initiation of the first action (e.g., applying brakes, reducing
thrust, deploying speed brakes) to stop the airplane during
accelerate-stop tests.
(b) V2 MIN, in terms of calibrated airspeed, may not be less than—
(1) 1.13 VSR for—
(i) Two-engine and three-engine turbopropeller and reciprocating
engine powered airplanes; and
364 Operational Aircraft Performance and Flight Test Practices

(ii) Turbojet powered airplanes without provisions for obtaining a

significant reduction in the one-engine-inoperative power-on
stall speed;
(2) 1.08 VSR for—
(i) Turbopropeller and reciprocating engine powered airplanes with
more than three engines; and
(ii) Turbojet powered airplanes with provisions for obtaining a
significant reduction in the one-engine-inoperative power-on
stall speed; and
(3) 1.10 times VMC established under §25.149.
(c) V2, in terms of calibrated airspeed, must be selected by the applicant to
provide at least the gradient of climb required by §25.121(b) but may not
be less than—
(1) V2 MIN;
(2) VR plus the speed increment attained (in accordance with
§25.111(c)(2)) before reaching a height of 35 feet above the takeoff
surface; and
(3) A speed that provides the maneuvering capability specified in
§25.143(h).
(d) VMU is the calibrated airspeed at and above which the airplane can safely
lift off the ground, and continue the takeoff. VMU speeds must be
selected by the applicant throughout the range of thrust-to-weight ratios
to be certificated. These speeds may be established from free air data if
these data are verified by ground takeoff tests.
(e) VR, in terms of calibrated airspeed, must be selected in accordance with
the conditions of paragraphs (e)(1) through (4) of this section:
(1) VR may not be less than—
(i) V1;
(ii) 105 percent of VMC;
(iii) The speed (determined in accordance with §25.111(c)(2)) that
allows reaching V2 before reaching a height of 35 feet above
the takeoff surface; or
(iv) A speed that, if the airplane is rotated at its maximum
practicable rate, will result in a VLOF of not less than —
(A) 110 percent of VMU in the all-engines-operating condition,
and 105 percent of VMU determined at the
thrust-to-weight ratio corresponding to the
one-engine-inoperative condition; or
(B) If the VMU attitude is limited by the geometry of the
airplane (i.e., tail contact with the runway), 108 percent of
VMU in the all-engines-operating condition, and 104
percent of VMU determined at the thrust-to-weight ratio
corresponding to the one-engine-inoperative condition.
(2) For any given set of conditions (such as weight, configuration, and
temperature), a single value of VR, obtained in accordance with this
paragraph, must be used to show compliance with both the
one-engine-inoperative and the all-engines-operating takeoff
provisions.
(3) It must be shown that the one-engine-inoperative takeoff distance,
using a rotation speed of 5 knots less than VR established in
accordance with paragraphs (e)(1) and (2) of this section, does not
exceed the corresponding one-engine-inoperative takeoff distance
using the established VR. The takeoff distances must be determined
in accordance with §25.113(a)(1).
(4) Reasonably expected variations in service from the established
takeoff procedures for the operation of the airplane (such as
 CHAPTER 9 Flight Envelope 365

over-rotation of the airplane and out-of-trim conditions) may not

result in unsafe flight characteristics or in marked increases in the
scheduled takeoff distances established in accordance with
§25.113(a).
( f) VLOF is the calibrated airspeed at which the airplane first becomes
airborne.
(g) VFTO, in terms of calibrated airspeed, must be selected by the applicant
to provide at least the gradient of climb required by §25.121(c), but may
not be less than—
(1) 1.18 VSR; and
(2) A speed that provides the maneuvering capability specified in
§25.143(h).
(h) In determining the takeoff speeds V1, VR, and V2 for flight in icing
conditions, the values of VMCG, VMC, and VMU determined for
non-icing conditions may be used.

As can be seen, the takeoff speeds are heavily regulated with margins and
restrictions. Figure 9.6 provides an easy navigation guide for §25.107 for a jet-
driven airplane. One must remember that in the end, it is the applicant’s
responsibility to select the right speeds for a given airplane’s takeoff
configuration.
The landing speed VREF is defined by §25.125(b)(2) as shown here. Note
how the latest amendment levels request the applicant to consider icing
(starting at Amendment 121) in the selection of the landing speed.

(i) In non-icing conditions, VREF may not be less than:

(A) 1.23 VSR0;
(B) VMCL established under §25.149(f); and
(C) A speed that provides the maneuvering capability specified in
§25.143(h).

Selected by
25.107(a)(1) applicant
Critical engine
VEFmin VEF
failure speed

“Decision speed”
25.107(a)(2) V1
≥VEF + ΔVpilot V1min Highest speed for start
of breaking – RTO
25.107(e)(1)(i)
≥V1 Rotation speed
25.107(e)(1)(ii) VRmin VR
≥1.05 VMC

≥1.10 VMU,AEO 25.107(e)(1)(iv) Lift-off speed

VLOFmin VLOF
≥1.05 VMU,OEI
25.107(b)(1)
Take-off
≥1.13 VSR V2min V2 safety
25.107(b)(3) speed
≥1.10 VMCA
25.107(c)
25.107(c)(3) ≤35 ft AGL
25.143(h) Maneuvering capability 25.107(c)(2) 25.121(b)
Climb gradient

Fig. 9.6 Takeoff speed restrictions.

366 Operational Aircraft Performance and Flight Test Practices

(ii) In icing conditions, VREF may not be less than:

(A) The speed determined in paragraph (b)(2)(i) of this section;
(B) 1.23 VSR0 with the most critical of the landing ice accretion(s)
defined in Appendices C and O of this part, as applicable, in
accordance with §25.21(g), if that speed exceeds VREF selected for
non-icing conditions by more than 5 knots CAS; and
(C) A speed that provides the maneuvering capability specified in
§25.143(h) with the most critical of the landing ice accretion(s)
defined in Appendices C and O of this part, as applicable, in
accordance with §25.21(g).

Minimum Control Speed

Multiengine transport category airplanes should still be capable of flying
safely after the loss of one engine while the other(s) are at the maximum
expected thrust in service, including continuing a takeoff, provided they
remain above their minimum control airspeed VMC . The concept of a
minimum control speed for flight has been around since early civil aviation
regulations. Indeed, Civil Aviation Regulation (CAR) 4b from 1953 stated:
§4b.133 Minimum control speed, VMC. (a) A minimum speed shall be deter-
mined under the conditions specified in this paragraph, so that when the
critical engine is suddenly made inoperative at that speed it shall be possible
to recover control of the airplane with the engine still inoperative, and main-
tain it in straight flight at that speed, either with zero yaw or, at the option of
the applicant, with an angle of bank not in excess of 58.

Back then, the emphasis of the regulation was on propeller-driven airplanes

(see Fig. 9.7), as exemplified by Interpretation No. 1 of CAR §4b.133 by the
Civil Aeronautics Board:

(1) The Board interprets and construes subparagraph (8) of §4b.133(a) as

requiring the Administrator to accept for the purposes of §4b.133 a value
for the one-engine-inoperative minimum control speed which has been
established in accordance with the provisions of that section with the
propeller of the inoperative engine feathered: Provided, that the airplane
involved is equipped with an automatic feathering device acceptable to
the Administrator under §4b.10 for demonstrating compliance with the
take-off path and climb requirement of §§4b.116 and 4b.120 (a) and (b).

The CAR 4b VMC rule was integrated into the FAA Federal Aviation Regu-
lations when they were created in 1965 under paragraph §25.149. We will
cover how the current regulations define VMC in the following subsections.

Minimum Control Speed: Air VMCA

Much of the original CAR 4b VMC rule was rolled into the new Part 25
regulations in 1965, specifically in paragraph §25.149. The text is generally
assigned to the minimum control speed in the air VMCA for the plane and
addresses the speeds for takeoff. Some modifications and additions were
introduced over the years. In its current form, VMCA (still referred to
mostly as VMC ) is:
 CHAPTER 9 Flight Envelope 367

Rolling moment

Critical engine failure

– Loss of lift outboard
– Rolling moment generated
– Loss of thrust outboard
– Yawing moment generated

Fig. 9.7 Impact of critical engine failure.

§25.149 Minimum control speed.

(a) In establishing the minimum control speeds required by this section, the
method used to simulate critical engine failure must represent the most
critical mode of powerplant failure with respect to controllability
expected in service.
(b) VMC is the calibrated airspeed at which, when the critical engine is
suddenly made inoperative, it is possible to maintain control of the
airplane with that engine still inoperative and maintain straight flight
with an angle of bank of not more than 5 degrees.
(c) VMC may not exceed 1.13 VSR with—
(1) Maximum available takeoff power or thrust on the engines;
(2) The most unfavorable center of gravity;
(3) The airplane trimmed for takeoff;
(4) The maximum sea level takeoff weight (or any lesser weight
necessary to show VMC );
(5) The airplane in the most critical takeoff configuration existing along
the flight path after the airplane becomes airborne, except with the
landing gear retracted;
(6) The airplane airborne and the ground effect negligible; and
(7) If applicable, the propeller of the inoperative engine—
(i) Windmilling;
(ii) In the most probable position for the specific design of the
propeller control; or
(iii) Feathered, if the airplane has an automatic feathering device
acceptable for showing compliance with the climb
requirements of §25.121.
(d) The rudder forces required to maintain control at VMC may not exceed
150 pounds nor may it be necessary to reduce power or thrust of the
operative engines. During recovery, the airplane may not assume any
dangerous attitude or require exceptional piloting skill, alertness, or
strength to prevent a heading change of more than 20 degrees.

Figure 9.8 shows the resulting asymmetric forces generated with resulting
moments when one engine becomes inoperative on a twin jet airplane in
368 Operational Aircraft Performance and Flight Test Practices

Bank angle to
counter rudder
side force

Ailerons deflection to
counter rudder rolling moment

DW
Rudder deflection
to counter Change in pitch trim
asymmetric thrust

Fig. 9.8 Impact of asymmetric thrust on flight control displacement.

flight. VMCA is then used to limit the minimum rotation speed VR and takeoff
safety speed V2 of Part 25 airplanes.
Among the changes introduced over the years, the maximum pedal force
to be used in the demonstration of VMC was reduced from 180 lb to 150 lb in
1978 by Amendment 42. We will cover VMCA more in the takeoff perform-
ance chapters (Chapters 15 and 16).

Dynamic VMCA
A sudden engine thrust loss for multiengine airplanes can cause a rapid
attitude change in all three axes, as can be seen in Fig. 9.7. Depending on
how suddenly the thrust is lost, the pilot may be overwhelmed by the devel-
oping situation, even while flying at or slightly above VMCA determined by
more static test conditions.
We provide the following description of the events in the critical con-
dition for dynamic VMCA , noting that the pilot will need to contend with
changing forces and moments, requiring rapid flight control inputs in all
three axes almost simultaneously. Refer to Fig. 9.9 for the discussion.
• The critical thrust-to-weight condition is maximum takeoff thrust at
minimum flight weight. This results in a high pitch attitude at engine
failure:
W Pitch attitude is equal to angle of attack plus flight path angle:

u ¼ a þ g.
W Angle of attack for most Part 25 airplanes is approximately 7 to 12 deg

at the low speeds associated with VMCA .

 CHAPTER 9 Flight Envelope 369

W Flight path angle is proportional to the specific excess thrust:

g ¼ sin1 ½ðT  DÞ=W .
W Most Part 25 airplanes have a thrust-to-weight ratio of 0.30 to 0.45 at

light weight, resulting in an excess thrust capability of more than 0.20

at light weight.
W The combination easily results in a pitch attitude over 20 deg at speeds

near VMCA ; therefore, at the time of engine failure, the pilot will need
to rapidly react to the engine thrust spooling down and pitch the nose
down in order not to stall.
• The rapidly developing thrust asymmetry will generate a yaw rate ċ before
it is arrested by the opposing rudder input side force.
W That yaw rate will result in the retreating wing producing less lift than

the advancing wing, which in turn will generate a roll rate and possibly
a wing tip stall.
W Local airspeed variation along the span at station r: Dv ¼ r ċ.
• In addition to the retreating wing loss of lift, the engine thrust loss may also
result in lower lift on the retreating wing, especially for propeller-driven
airplanes where part of the wing lift comes from the propeller slipstream
(see Fig. 9.10), further increasing the roll rate and risk of wing tip stall.
W The dropping wing increases the local (radius r from center of
 
r ḟ
rotation) angle of attack: Da  V .

·


Retreating

V wing

Advancing
 wing

·


Fig. 9.9 Dynamic VMCA additional forces due to roll, pitch, and yaw.
370 Operational Aircraft Performance and Flight Test Practices

Lift distribution over Hercules wing, power-off

Lift distribution over Hercules wing, engine 1 inoperative
350

300

250

200


150

100

50

0
–1 –0.8 –0.6 –0.4 –0.2 0 0.4 0.6 0.8
2yw/Qw 0.2 1

Fig. 9.10 Example of propeller wash effects over wing with left engine inoperative.
Source: Adapted from [5].

W Aileron deflection to stop the rolling motion will change the local wing
lift curve, which, like a flap, will result in lower local astall .
W A bank angle f will lead to an induced rate of turn ẋ that will
contribute to the heading change: ẋ ¼ ½g tanðfÞ=V .
W The heading change will be proportional to the time the bank angle
is present.

The guidance material of FAA AC25-7D recommends (underline added

by author):
If the dynamic tests result in a VMCA greater than the static value, the incre-
ment between the static and dynamic VMCA at the same altitude should be
added to the sea level extrapolated value. If the dynamic value is less than the
static value, the static VMCA should be used for the AFM data expansion.

Of course, as the modeling capability of manufacturers has improved, the

industry has started moving from flight test demonstration to more analy-
sis/simulation verification for dynamic VMCA . With the introduction of
roll spoilers to enhance roll capability at low speeds, most manufacturers
have not seen any effects of dynamic VMCA on the static VMCA speed dem-
onstration, and analysis of dynamic VMCA is now an acceptable means of
compliance for certification.
 CHAPTER 9 Flight Envelope 371

Safe, Intentional One-Engine-Inoperative Speed VSSE

It is recognized that for some Part 23 twin-engine propeller airplanes,
flying at exactly VMCA may prove challenging and has resulted in some acci-
dents during dedicated training for one-engine-inoperative handling. The
FAA introduced a minimum speed to intentionally render the critical
engine inoperative VSSE in Amendment 45 of Part 23 in 1993:
§23.149(d) A minimum speed to intentionally render the critical engine
inoperative must be established and designated as the safe, intentional,
one-engine-inoperative speed, VSSE .

On some airplanes, depending on the weight at the time of the engine failure,
the airplane could be rudder travel limited (directional control) or aileron
deflection limited (roll control) (see Fig. 9.10). This should be taken into
account when selecting VSSE . For this example (Fig. 9.11), the manufacturer
may recommend a VSSE speed of 140 KCAS (approximately the minimum
control speed for wings level and maximum asymmetric thrust) because it
would be a good roll attitude to start the training into airplane characteristics
with high thrust asymmetry.
It is important to remember that VSSE is not established as a limitation, but
rather as a recommended operational procedure, as identified by §23.1585:
§23.1585 Operating procedures.
(a) For each airplane, information concerning normal, abnormal, and
emergency procedures and other pertinent information necessary for
safe operation and the achievement of the scheduled performance must
be identified and segregated, including . . .
(c) For multiengine airplanes, the information must include—
(6) The VSSE determined in Sec. 23.149.

170
160 Into live engine
Minimum control speed (KCAS)

150
d

140
ite
m
r li

130
de

Ai
d

le
Ru

120 ro
n
110 lim
ite
d
100
90
80
–10 –8 –6 –4 –2 0 2 4
Bank angle (deg)

Fig. 9.11 Minimum control speed vs bank angle for maximum asymmetric thrust.
372 Operational Aircraft Performance and Flight Test Practices

Of interest for this subject, a search of the Internet for Part 23 airplane pilot
operating handbooks (POHs) has highlighted some of the issues related to
VMCA on small Part 23 twins; for example, a given light twin has a VMCA
of 80 KIAS and a VSSE of 92 KIAS. By themselves, they are interesting to
know, but when one looks at the specified target speed at 50 ft (V50 ) for a
lightweight takeoff for the same plane, one finds that it is 81 KIAS, below
the VSSE speed. That same airplane has a note in the POH that states
“Although the airplane is controllable at the air minimum control speed,
the airplane performance is so far below optimal that continued flight near
the ground is improbable. A more suitable recommended safe single-engine
speed is 92 KIAS.”
With the major rewrite of Part 23 in Amendment 64, issued in 2017, the
VSSE is not stated implicitly anymore but rather rolled into §23.2135,
Controllability.
§23.2135 Controllability

(c) VMC is the calibrated airspeed at which, following the sudden critical loss
of thrust, it is possible to maintain control of the airplane. For multien-
gine airplanes, the applicant must determine VMC , if applicable, for the
most critical configurations used in takeoff and landing operations.

At the option of the applicant, VSSE can and should still be presented in the
documentation of a small multiengine airplane; it still represents a speed that
is judged acceptable for a pilot of average skill to not be overwhelmed during
a training exercise that is required to maintain pilot proficiency.

Minimum Control Speed: Landing VMCL

Two new minimum control speeds for Part 25 were introduced in
Amendment 42 (1978), VMCL and VMCL22 . The speeds are associated with
the minimum control speed in the landing configuration for an engine
failure during landing approaches that are initiated with all engines operating
(AEO) and with one engine inoperative (OEI). The all-engine-operating con-
dition [§25.149(f)] is intended to determine a minimum control speed for the
situation where an engine fails after power or thrust has been increased to
make a go-around.
§25.149 (f) VMCL, the minimum control speed during approach and landing
with all engines operating, is the calibrated airspeed at which, when the critical
engine is suddenly made inoperative, it is possible to maintain control of
the airplane with that engine still inoperative, and maintain straight flight with
an angle of bank of not more than 5 degrees. VMCL must be established with—
(1) The airplane in the most critical configuration (or, at the option of the
applicant, each configuration) for approach and landing with all engines
operating;
(2) The most unfavorable center of gravity;
(3) The airplane trimmed for approach with all engines operating;
(4) The most unfavorable weight, or, at the option of the applicant, as a
function of weight;
 CHAPTER 9 Flight Envelope 373

(5) For propeller airplanes, the propeller of the inoperative engine in the
position it achieves without pilot action, assuming the engine fails while
at the power or thrust necessary to maintain a three degree approach
path angle; and
(6) Go-around power or thrust setting on the operating engine(s).

The second landing minimum control speed introduced in Amendment 42 of

Part 25, VMCL22 , was done as part of the creation of §25.149(g), which
requires the determination of a minimum control speed at which the airplane
can be safely controlled when an approach is initiated with one engine
inoperative and another engine fails during the approach (see Fig. 9.12).
The regulation states:
§25.149 (g) For airplanes with three or more engines, VMCL-2, the minimum
control speed during approach and landing with one critical engine inopera-
tive, is the calibrated airspeed at which, when a second critical engine is sud-
denly made inoperative, it is possible to maintain control of the airplane with
both engines still inoperative, and maintain straight flight with an angle of
bank of not more than 5 degrees. VMCL-2 must be established with—

(1) The airplane in the most critical configuration (or, at the option of the
applicant, each configuration) for approach and landing with one critical
engine inoperative;
(2) The most unfavorable center of gravity;
(3) The airplane trimmed for approach with one critical engine inoperative;
(4) The most unfavorable weight, or, at the option of the applicant, as a
function of weight;
(5) For propeller airplanes, the propeller of the more critical inoperative
engine in the position it achieves without pilot action, assuming the
engine fails while at the power or thrust necessary to maintain a three
degree approach path angle, and the propeller of the other inoperative
engine feathered;

DW Second critical engine fail

DW
First critical engine fail

Fig. 9.12 VMCL22 condition.

374 Operational Aircraft Performance and Flight Test Practices

(6) The power or thrust on the operating engine(s) necessary to maintain an

approach path angle of three degrees when one critical engine is
inoperative; and
(7) The power or thrust on the operating engine(s) rapidly changed,
immediately after the second critical engine is made inoperative, from
the power or thrust prescribed in paragraph (g)(6) of this section to—
(i) Minimum power or thrust; and
(ii) Go-around power or thrust setting.

For both the AEO and the OEI conditions, in addition to the steady heading
requirements and dynamic case, an additional roll rate demonstration
requirement was added to the landing configuration at the minimum
control speed, §25.149(h)(3). This additional requirement can be viewed as
follows: unlike the takeoff speed V2 , which is a minimum margin of 1.10
away from VMCA [§25.107(b)(3)], the requirements of §25.125 allow VREF to
be no lower than VMCL with no additional margin. Flying at the minimum
control speed for the configuration would require larger lateral control
inputs with less excess roll capability. This can be especially true for
propeller-driven airplanes (see Fig. 9.13); as well, swept wing airplanes tend
to have a large adverse yaw that tends to reduce roll capability. The require-
ment of §25.149(h)(3) thus define what becomes an acceptable minimum.
§25.149 (h) In demonstrations of VMCL and VMCL-2—
(1) The rudder force may not exceed 150 pounds;
(2) The airplane may not exhibit hazardous flight characteristics or require
exceptional piloting skill, alertness, or strength;
(3) Lateral control must be sufficient to roll the airplane, from an initial
condition of steady flight, through an angle of 20 degrees in the direction
necessary to initiate a turn away from the inoperative engine(s), in not
more than 5 seconds; and
(4) For propeller airplanes, hazardous flight characteristics must not be
exhibited due to any propeller position achieved when the engine fails or
during any likely subsequent movements of the engine or
propeller controls.

Minimum Control Speed: Ground VMCG

The minimum control speed on the ground VMCG , as defined by
§25.149(e), is a mathematical definition of the minimum V1 [§25.107(a)]
speed that can be used to compute the
takeoff field length of an airplane; VMCG is
20 deg in 5 s a performance speed.
§25.107(a) V1 must be established in
relation to VEF as follows:
(1) VEF is the calibrated airspeed at
which the critical engine is
assumed to fail. VEF must be
selected by the applicant, but may
not be less than VMCG
Fig. 9.13 VMCL roll requirements. determined under §25.149(e).
 CHAPTER 9 Flight Envelope 375

(2) V1, in terms of calibrated airspeed, is selected by the applicant; however,

V1 may not be less than VEF plus the speed gained with critical engine
inoperative during the time interval between the instant at which the
critical engine is failed, and the instant at which the pilot recognizes and
reacts to the engine failure, as indicated by the pilot’s initiation of the first
action (e.g., applying brakes, reducing thrust, deploying speed brakes) to
stop the airplane during accelerate-stop tests.

Per §25.149(e), VMCG , the minimum control speed on the ground, is the
calibrated airspeed during the takeoff run at which, when the critical
engine is suddenly made inoperative, it is possible to maintain control of
the airplane using rudder control alone (without the use of nosewheel
steering), as limited by 150 lb of force, and the lateral control to the
extent of keeping the wings level to enable the takeoff to be safely contin-
ued using normal piloting skill. In the determination of VMCG , assuming
that the path of the airplane accelerating with all engines operating is
along the centerline of the runway, its path from the point at which the
critical engine is made inoperative to the point at which recovery to a
direction parallel to the centerline is completed may not deviate more
than 30 ft laterally from the centerline at any point. VMCG must be estab-
lished with
• The airplane in each takeoff configuration or, at the option of the applicant,
in the most critical takeoff configuration
• Maximum available takeoff power or thrust on the operating engines
• The most unfavorable center of gravity
• The airplane trimmed for takeoff
• The most unfavorable weight in the range of takeoff weights

What essentially happens during an engine failure event at takeoff is one

engine will rapidly lose thrust due to a failure, the most adverse cause often
being fuel starvation. Following that failure, with rapidly decreasing thrust on
one engine, the airplane will start yawing away from the runway centerline
(Fig. 9.14). Once the failure is detected, the pilot will apply rudder, and

Gear side
force
Original track/heading
ing
Head
30 ft
d track
Groun

Rudder side
force

Fig. 9.14 VMCG event and forces.

376 Operational Aircraft Performance and Flight Test Practices

often nosewheel steering, input to counter this adverse yawing moment to

regain directional control of the plane and bring it back toward a runway
heading. The lateral force produced is, in part, countered by the main
landing gear wheels, but there will be some lateral skidding. This can be
seen during VMCG testing as a deviation between the airplane heading and
the ground track of the center of gravity.
Notice of Proposed Rulemaking (NPRM) 75-25 (1975) stated:

“V1” is now defined as the critical-engine-failure speed, and the accelerate-

stop distance is established as the distance required to accelerate the airplane
to VC1 and stop. Section §25.101(h) requires allowance for any time delays in
the execution of procedures that may reasonably be expected in service.

The notice also added:

Section §25.107(a) presently contains a ground controllability requirement

in the determination of V1. A proposal for Sec. §25.107(a) makes it desir-
able for clarity to transfer the ground control provisions to Sec. §25.149.
The proposal for Sec. §25.149(e) would require the determination of
VMCG, the minimum control speed on the ground for the sudden
failure of the critical engine during the takeoff roll. During this demon-
stration, the permissible lateral deviation of the path of the airplane
would be limited to 30 feet.

Section 25.107(a) prior to Amendment 42 contained statements about the

need for acceptable controllability for a continued takeoff:

V1 must be selected by the applicant and must be at least the minimum cali-
brated airspeed at which controllability by primary aerodynamic controls
alone is shown (during the takeoff run) to be adequate to safely continue
the takeoff, using normal piloting skill, when the critical engine is suddenly
made inoperative.

There was no VMCG prior to Amendment 42, just an understanding

that during a continued takeoff, the lateral deviation should not be greater
than 25 ft (FAA flight test guide of the time). After industry and agency
review of NPRM 75-25, Amendment 42 to Part 25 was adopted in 1978
and introduced §25.149(e) VMCG . It also introduced a critical-engine
failure speed VEF under §25.107(a), and it redefined V1 as shown at the
beginning of this section; §25.107(a) and §25.149(e) took on their current
form in 1978.
To further expand on the statement that VMCG is a performance speed,
the guidance material of FAA advisory circular (AC) 25-7D allows the speed
to be achieved in zero crosswind, and it states that VMCG testing should be
conducted at the most critical weight in the range where VMCG may
impact AFM V1 speeds.
We will expand on VMCG in the chapter on rejected takeoff testing
(Chapter 18).
 CHAPTER 9 Flight Envelope 377

NTSB AAR-95/06: Uncontrolled collision with terrain, DC-8-63, Kansas

City MO, 16 Feb. 1995: The flight crew of the accident airplane attempted to
start engine no. 1 on the day of the accident but were unable to do so. Air
Transport International (ATI) management authorized a ferry flight (Part
91 operation) with a three-engine takeoff with a specific procedure from
Kansas City to Westover AFB where the engine would be replaced/repaired.
The three-engine takeoff procedure called for partial application of reduced
thrust on the opposite engine to the inoperative one until the aircraft exceeded
ground minimum control speed.
Events relating to this accident include:
• The crew computed the wrong ground minimum control speed (9 KCAS
too low).
• Application of maximum thrust on the outboard engine before VMCG .
• The pilot lost control during the takeoff roll.
• The pilot decided to continue takeoff with an early rotation below the
computed rotation speed, resulting in a premature liftoff at a speed below
VMCA.
• Further loss of control after liftoff followed by collision with the terrain.
378 Operational Aircraft Performance and Flight Test Practices

Trim
The airplane must be capable of being trimmed within most of its flight
envelope. For an airplane to be trimmed, there needs to be a steady state (air-
speed, altitude and/or flight path, and attitude) with no pilot input through
the flight controls. FAA §25.161 states:

§25.161(a) Each airplane must meet the trim requirements of this section
after being trimmed, and without further pressure upon, or movement of,
either the primary controls or their corresponding trim controls by the
pilot or the automatic pilot.

The reason behind this is simple: one should not expect the pilot to have to
maintain a flight control input force in any axis for more than a relatively
short duration. The forces expected should not exceed those defined by
§25.143(d):

Force, in pounds, applied to the control wheel

or rudder pedals Pitch Roll Yaw
For short term application for pitch and roll—two 75 50
hands available for control
For short term application for pitch and roll—one 50 25
hand available for control
For short term application for yaw control 150
For long term application 10 5 20

Short-term and long-term forces are described by AC 25-7D as follows:

Short-term forces are the initial stabilized control forces that result from
maintaining the intended flight path following configuration changes and
normal transitions from one flight condition to another, or from regaining
control following a failure. It is assumed that the pilot will take immediate
action to reduce or eliminate such forces by re-trimming or changing con-
figuration or flight conditions, and consequently short-term forces are not
considered to exist for any significant duration. They do not include transi-
ent force peaks that may occur during the configuration change, change of
flight conditions, or recovery of control following a failure.
Long-term forces are those control forces that result from normal or
failure conditions that cannot readily be trimmed out or eliminated.

The reason for this is simple: one cannot expect a pilot to be able to main-
tain forces throughout the flight without getting fatigued, which would
impact the pilot’s ability to safely control the plane. The first condition to
consider is the capability to trim laterally and directionally, as specified by
§25.161(b):

§25.161 (b) Lateral and directional trim. The airplane must maintain lateral
and directional trim with the most adverse lateral displacement of the center
 CHAPTER 9 Flight Envelope 379

of gravity within the relevant operating limitations, during normally expected

conditions of operation (including operation at any speed from 1.3 VSR1 to
VMO /MMO ).

This requirement typically is achieved by building airplanes that are aerody-

namically symmetric with respect to the vertical plane (see Fig. 9.15). Most
Part 23 and Part 25 airplanes are built this way, which also minimizes the
lateral center of gravity (CG) excursion from the plane of symmetry; but
there are exceptions. The Rutan Boomerang is one such example where
the designer had to ensure the airplane would be trimmable over its designed
speed envelope with careful sizing of all control surfaces.
The next condition is the longitudinal trim capability. This condition
requires careful sizing of the horizontal tail trim system because the airplane
pitching moment will be impacted by airplane configuration (flaps/slats and
gear positions), airspeed and altitude, center of gravity location, and thrust
setting. Section 25.161(c) specifies the minimum requirements for longitudi-
nal trim capability.
§25.161 (c) Longitudinal trim. The airplane must maintain longitudinal trim
during—

(1) A climb with maximum continuous power at a speed not more than
1.3 VSR1 , with the landing gear retracted, and the flaps (i) retracted and
(ii) in the takeoff position;
(2) Either a glide with power off at a speed not more than 1.3 VSR1 , or an
approach within the normal range of approach speeds appropriate to the
weight and configuration with power settings corresponding to a 3
degree glidepath, whichever is the most severe, with the landing gear
extended, the wing flaps (i) retracted and (ii) extended, and with the most
unfavorable combination of center of gravity position and weight
approved for landing; and
(3) Level flight at any speed from 1.3 VSR1 , to VMO /MMO , with the landing
gear and flaps retracted, and from 1.3 VSR1 to VLE with the landing
gear extended.

Fig. 9.15 Most Part 23 and 25 airplanes are symmetric.

380 Operational Aircraft Performance and Flight Test Practices

Nose-down trim capability

Aft CG
Trim setting

Fwd CG
Fwd CG in icing

Nose-up trim capability

High speed Low speed CL

Fig. 9.16 Longitudinal trim capability chart example.

A useful way to display an airplane’s longitudinal trim capability for a given

configuration over the operating speed range and thrust settings is to plot the
trim system position as a function of the airplane’s lift coefficient (see
Fig. 9.16). One can then usually clearly see if any marginal conditions exist.
The last condition that requires attention from a trim capability point of
view is the condition for one engine inoperative (OEI). This requirement is
divided into two subparagraphs of §25.161, subparagraphs (d) and (e).
§25.161 (d) Longitudinal, directional, and lateral trim. The airplane must
maintain longitudinal, directional, and lateral trim (and for the lateral trim,
the angle of bank may not exceed five degrees) at 1.3 VSR1 during climbing
flight with—
(1) The critical engine inoperative;
(2) The remaining engines at maximum continuous power; and
(3) The landing gear and flaps retracted.

For airplanes with four or more engines:

§25.161 (e) Airplanes with four or more engines. Each airplane with four or
more engines must also maintain trim in rectilinear flight with the most
unfavorable center of gravity and at the climb speed, configuration, and
power required by §25.123(a) for the purpose of establishing the en route
flight paths with two engines inoperative.

Maneuvering Speed
The maneuvering speed VA is the speed at which the airplane will stall
while flying at its design maximum load factor; we will cover the equations
for this condition in Chapter 14. From a Part 25 certification point of view,
§25.1583 requires that VA be provided in the airplane AFM:
 CHAPTER 9 Flight Envelope 381

§25.1583(3) The maneuvering speed VA and a statement that full appli-

cation of rudder and aileron controls, as well as maneuvers that involve
angles of attack near the stall, should be confined to speeds below this value.

This requirement states that it should be possible to apply full flight controls
deflection below VA without a risk of damaging the plane. At speeds above
VA , care must be taken because the airplane will reach structural limits
prior to stalling. Typical variation in VA vs weight and altitude is shown in
Fig. 9.17. From a certification point of view, §25.335 defines VA as:
§25.335 Design Airspeeds

(c) Design maneuvering speed VA . ForpVA , the following apply:

(1) VA may not be less than VS1 n where—
(i) n is the limit positive maneuvering load factor at VC ; and
(ii) VS1 is the stalling speed with flaps retracted.
(2) VA and VS must be evaluated at the design weight and altitude
under consideration.
(3) VA need not be more than VC or the speed at which the
positive CN max curve intersects the positive maneuver load factor
line, whichever is less.

The accident of American Airlines Flight 587 in 2001 revealed that VA may
not always protect for all full deflections. In this case, an Airbus A300 lost its
vertical tail following the rapid application of full rudder reversal.
The positive load factor for large transport airplanes is 2.5 g for most jet-
liners. The certification requirements recognize that lighter weight airplanes
with their lower pitch inertia may generate larger values during rapid man-
euvering, and they ask that the airplanes be certified to higher values per
14 CFR §25.337 (see Fig. 9.18).

45,000
40,000

35,000
MMO
30,000
Altitude (ft)

25,000
VA OWE
20,000
VA MTOW VMO
15,000

10,000
5000

0
100 150 200 250 300 350
Airspeed (KCAS)

Fig. 9.17 Maneuvering speed example.

382 Operational Aircraft Performance and Flight Test Practices

4.5

4 Not more than 3.8

3.5
Load factor (g)

2.5
Not less than 2.5

1.5

1
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000
MTOW (lb)

Fig. 9.18 Design maneuver load factor per §25.337.

§25.337 Limit maneuvering load factors

(b) The positive limit maneuvering load factor n for any speed up to Vn may
not be less than 2.1 þ 24,000/(W þ 10,000) except that n may not be less
than 2.5 and need not be greater than 3.8—where W is the design
maximum takeoff weight.

That positive limit load factor can be reduced to 2.0 g for flaps extended, per
§25.345.
§25.345 High lift devices.

(a) If wing flaps are to be used during takeoff, approach, or landing, at the
design flap speeds established for these stages of flight under
§25.335(e) and with the wing flaps in the corresponding positions, the
airplane is assumed to be subjected to symmetrical maneuvers and
gusts. The resulting limit loads must correspond to the conditions deter-
mined as follows:
(1) Maneuvering to a positive limit load factor of 2.0

NTSB AAR-04/04: On 12 Nov. 2001, about 0916:15 Eastern Standard Time,

American Airlines flight 587, an Airbus Industrie A300-605R, N14053,
crashed into a residential area of Belle Harbor, New York, shortly after
takeoff from John F. Kennedy International Airport, Jamaica, New York.
Soon after takeoff, as the airplane was climbing through 2500 ft at an air-
speed of about 251 KIAS, the airplane experienced some turbulence. The
copilot responded by applying large aileron input and maximum rudder
deflection immediately followed by a full reversal. The vertical tail departed
the airplane, and the crew lost control and crashed.
 CHAPTER 9 Flight Envelope 383

The NTSB determined that the probable cause of this accident was the
in-flight separation of the vertical stabilizer as a result of the loads beyond ulti-
mate design that were created by the first officer’s unnecessary and excessive
rudder pedal inputs. Contributing to these rudder pedal inputs were charac-
teristics of the Airbus A300-600 rudder system design and elements of the
American Airlines Advanced Aircraft Maneuvering Program.
In a postaccident interview, American Airlines’ managing director of flight
operations technical stated that the rudder should be able to be fully displaced
and stay within its structural limit as long as the rudder travel limiter was
working properly and the airplane was traveling below VA . Also, he thought
that the rudder travel limiter would protect the airplane with a full deflection
of the rudder followed by a deflection in the opposite direction as long as the
airplane was traveling below VA . He further stated that most of the company
pilots believed that, if the pilot made right, left, and right rudder inputs, the
airplane would be protected as long as it was traveling below VA .
The American Airlines A300 Operating manual at the time of the accident
stated that VA (AFM) was 270 kt/0.78 Mach, whichever is lower.

Crosswinds
Crosswind information for takeoff and landing provided in flight manuals
of Part 23 and Part 25 airplanes is often not a limiting value, but rather a value
that was demonstrated to be safe, most of the time, by flight testing. We
specify most of the time because there is a push by the industry to expand
the crosswind values provided in the flight manual by performing more
analysis, simulation, and pilot in the loop simulation.
The minimum requirement in terms of crosswind demonstration for
Part 23 prior to Amendment 64 was:
§23.233 Directional stability and control. (amendment 50, 1996)

(a) A 90 degree cross-component of wind velocity, demonstrated to be safe

for taxiing, takeoff, and landing must be established and must be not less
than 0.2 VS0 .
(b) The airplane must be satisfactorily controllable in power-off landings at
normal landing speed, without using brakes or engine power to maintain
a straight path until the speed has decreased to at least 50 percent of the
speed at touchdown.
(c) The airplane must have adequate directional control during taxiing.

This was changed to the following in Amendment 64:

§23.2155 Ground and water handling characteristics.

For airplanes intended for operation on land or water, the airplane must have
controllable longitudinal and directional handling characteristics during taxi,
takeoff, and landing operations.
384 Operational Aircraft Performance and Flight Test Practices

For Part 25 transport category airplanes, the crosswind requirements fall into
§25.237:
§25.237 Wind Velocities
(a) For land planes and amphibians, the following applies:
(1) A 90-degree cross component of wind velocity, demonstrated to be
safe for takeoff and landing, must be established for dry runways and
must be at least 20 knots or 0.2 VSR0, whichever is greater, except
that it need not exceed 25 knots.
(2) The crosswind component for takeoff established without ice
accretions is valid in icing conditions.
(3) The landing crosswind component must be established for:
(i) Non-icing conditions, and
(ii) Icing conditions with the most critical of the landing ice
accretion(s) defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g).

Tailwinds
The typical wind limit for a transport category airplane is 10 kt. This is
achieved for “free” during initial certification based on the probable airplane
exposure to normal flight test environment over the 2000 to 4000 flight test
hours required to demonstrate compliance with requirements and by the
basic design of the plane (as shown historically and by analysis). This is recog-
nized in the flight test guide FAA AC 25-7D, which states:
Wind Velocities of 10 Knots or Less. Approval may be given for performance,
controllability, and engine operating characteristics for operations in
reported tailwind velocities up to 10 knots without conducting additional
flight tests at specific wind speeds.

If the airplane manufacturer wants greater capability (typically over time

they do because 10-kt tailwind clearance does not allow a landing with 7
kt of tailwind and 4 kt of gust, for example), then a demonstration of the air-
plane’s capability is required. Initially, the guidance material requested that
the airplane be tested to 150% of the desired tailwind certification; this
meant that if 15 kt was required, the minimum demonstration (takeoff and
landing) needed to be at least 22.5 kt (average winds, not peak winds). Of
course, a manufacturer wanting to demonstrate at least this value would
end up testing much higher (overdemonstration), something definitely
not required.
The reasoning provided by FAA guidance material for such excess testing
was that takeoff, rejected takeoff, and landing distances, measured in tailwind
conditions greater than 10 kt, are unreliable for determining airplane per-
formance. Wind conditions of such magnitude are generally not sufficiently
consistent over the length of the runway or over the time period required to
perform the test maneuver.
This last part, where unknowns of measured and reported winds are a
reason to increase the test margins, seems contrary to the normal
 CHAPTER 9 Flight Envelope 385

certification steps in which a simple demonstration with supporting analysis

and simulation would suffice, such as in the case of crosswinds. The elements
of tailwind takeoff or landing can easily be broken down into components to
be analyzed. Figure 9.19 shows some of the impact of tailwind on a landing as
an example. Each of the components can be analyzed individually and in
combinations (sensitivity analysis) to see the impact any change would
have on the airplane landing flight characteristics and handling by either
pure simulation or by pilot in the loop simulation. It is true that the random-
ness of the wind makes any two landings (or takeoffs) nonrepeatable, but ana-
lyzing each individual variable impact on the maneuver will provide the
manufacturer with valuable clues on the capability of the airplane to
handle such randomness.
Engines (§25.903) will still have to be tested in large tailwind conditions to
determine how the integration with the airframe under varying wind con-
ditions will impact their operating characteristics; however, this can be
done on the ground under static conditions. This is a more adverse tailwind
than an airplane traveling along the runway. The exception can be thrust
reversers where the engine thrust schedule is programmed with airspeed;
in this case, some level of testing needs to be done. Impact on the landing
gear loads must also be investigated if more than 10 kt of tailwind is
desired for certification, per §25.479(a):
§25.479 Level landing conditions
(a) In the level attitude, the airplane is assumed to contact the ground at
forward velocity components, ranging from VL1 to 1.25 VL2 parallel to
the ground under the conditions prescribed in §25.473 with—
(1) VL1 equal to VS0 (TAS) at the appropriate landing weight and in
standard sea level conditions; and
(2) VL2 equal to VS0 (TAS) at the appropriate landing weight and
altitudes in a hot day temperature of 41 degrees F. above standard.
(3) The effects of increased contact speed must be investigated if
approval of downwind landings exceeding 10 knots is requested.

 = (T – D)/W
• Thrust: min idle, dynamic response
• Drag: config (flap, gear, etc.)
Tailwind (TW) landing • Airspeed
• Altitude
• Pilot FOV (pitch =  + )
g ILS
3 de
deg Sink rateTW = KTAS sin( > 3 deg)
>3
KTAS cos( > 3 deg) Wind
• Pilot reaction time reduced
• Flare requirements increased
Ground speedTW
• More elevator
3 deg ILS = tan–1(Sink rateTW/Ground speedTW) • More load factor
• Higher flare altitude…

• Consideration for higher tire speed vs limits

• MLG VL2 consideration

Fig. 9.19 Factors affecting tailwind landing.

386 Operational Aircraft Performance and Flight Test Practices

The guidance material of AC 25-7D states:

The test wind condition of 150 percent of the proposed tailwind limit should
be an averaged or smoothed wind speed, not a peak wind speed. Airplane
control characteristics should be evaluated under the following conditions
with the CG at the aft limit and the test mean tailwind velocity equal to
the proposed limit tailwind factored by 150 percent:

(1) Takeoff. Both all-engines-operating and one-engine-inoperative (i.e.,

with a simulated failure of the critical engine at the engine failure speed)
takeoffs should be evaluated at a light weight with maximum approved
takeoff flap deflection.
(2) Landing. Approach and landing at light weight with maximum approved
landing flap deflection.
(3) Determination of the increased ground speed effect on gear vibration or
shimmy, and flight director, or autopilot instrument landing system (ILS)
approaches, terrain awareness warning system (TAWS) sink rate
modes, etc.
(4) If engine idle power or thrust is increased to account for the increased
tailwind velocity, ensure that deviations above the glideslope are
recoverable.

A report published by the Flight Test Harmonization Working Group

(FTHWG) in April 2017 did recognize that this position was too conserva-
tive; it took a first step towards addressing this by changing the demon-
stration from at least 150% of the desired tail wind to 5 kt more, both still
being the average wind during the takeoff or landing. Although this seems
like a significant reduction, it should be looked at in terms of what would
really be authorized, a maximum of 15 kt tailwind. So, the reduction pro-
posed by the FTHWG is really reducing the average demonstrated wind
from 22.5 kt to 20 kt, not a big reduction. This author strongly recommends
that analysis should be used in support of testing with some limited expan-
sion beyond the demonstrated value being allowed and that the manufacturer
not be forced to only have a “pass by flight test” result. Winds, just like other
weather-related certification items, are hard to achieve (in the case of tail-
winds, the manufacturer will most likely ask to land in a direction opposite
to the active runway, thereby requiring the airport authorities to insert
their testing into the traffic pattern) and time consuming, which drives up
the cost of certification.
We will cover more about this topic in Chapters 15–20 including what
can be done during the analysis to support the results and publication of
the performance under tailwinds. In the end, the manufacturer should
expect an AFM limitation of a 10-kt tailwind with no dedicated testing or
15 kt with testing (and with exceptional cases, higher).

High-Speed Limit
On the high-speed side of the envelope, as we have seen in Chapter 8, one
would expect the airplane to be limited by the maximum thrust available to a
 CHAPTER 9 Flight Envelope 387

maximum airspeed in level flight. From a pure performance analysis point of

view, that would be the natural limit. The maximum speed a Part 25 airplane
is authorized to achieve in “normal” flight may be restricted to a much lower
value, as will be seen.

Maximum Operating Airspeed/Mach

The maximum operating airspeed/Mach (VMO /MMO ) for an airplane is
selected by the applicant and can vary with altitude. Under Part 25, Sub-Part
G, Operating Limitation—Airplane Flight Manual, VMO /MMO is defined by
§25.1583 as follows:
§25.1583 Operating limitations (a) Airspeed limitations. The following air-
speed limitations and any other airspeed limitations necessary for safe oper-
ation must be furnished:
(1) The maximum operating limit speed VMO /MMO and a statement that
this speed limit may not be deliberately exceeded in any regime of flight
(climb, cruise, or descent) unless a higher speed is authorized for flight
test or pilot training.

This maximum operating speed is validated to be safe by the execution of

flight test maneuvers beyond VMO /MMO per other requirements. This
limit is often a lot less than the thrust limited speed Vmax at low altitude
and beyond the maximum speed at near maximum altitude, as shown in
Fig. 9.20. Some of the certification requirement that validate VMO /MMO are:
• §25.161 Trim
• §25.251 Vibration and Buffet
• §25.253 High Speed Characteristics
• §25.255 Out-of-Trim

40,000
35,000
MMO
30,000
Altitude (ft)

25,000
20,000
VMO
15,000
Vmax
10,000
5000

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9
Mach

Fig. 9.20 VMO /MMO speeds vs maximum thrust limited speed.

388 Operational Aircraft Performance and Flight Test Practices

• §25.335 Design Airspeeds

• §25.629 Aeroelastic Stability Requirements
• §25.1505 Maximum Operating Limit Speed
We will briefly touch on each of these to show what it means to the
maximum speed that can be used on a Part 25 certified airplane. First, trim:
§25.161(c)(3) Trim: The airplane must maintain longitudinal trim during
Level flight at any speed from 1.3 VSR1 , to VMO /MMO , with the landing
gear and flaps retracted, and from 1.3 VSR1 to VLE with the landing gear
extended.

As we discussed earlier in this chapter, a trimmed airplane is one that can

maintain the specified flight condition with aerodynamic surfaces only
without any input from the pilot or an autopilot. This is not to say that a com-
ponent of the autopilot cannot be used to achieve trim, but rather that once
the flight condition is reached, no additional aerodynamic surface movement
is commanded. For a fly-by-wire airplane, this would simply mean that the
airplane flight condition can be maintained because, in this case, the pilot
does not have direct control of the flight controls. As a note, for most Part
25 airplanes, the critical configuration (least margins) for trim at high
speed is the lightweight aft center of gravity location, as can be observed in
Fig. 9.15 earlier in the chapter.

High-Speed Buffet
We briefly discussed low-speed buffeting earlier in this chapter and how
it was tied to high angles of attack and stall. At high Mach numbers, as low as
Mach 0.60 for some airplanes, part of the wing will first reach sonic speed.
Beyond that Mach number, shock waves will start to form. As the shock
waves grow stronger, the airflow going through them may separate, leading
to buffeting of the airplane (see Fig. 9.21). At a high enough Mach number
for a given airplane configuration, this buffeting may start in level flight
conditions.

Typical buffet
Shock limit
CLbuff

1-g High-speed
Shock-induced flig
separation ht buffet

Mach

Fig. 9.21 High-speed buffet.

 CHAPTER 9 Flight Envelope 389

Early jetliners were designed following CAR 4b regulation and the initial
version of Part 25. As such, they were required to not have any “severe” buf-
feting that could interfere with the crew’s ability to safely operate the air-
plane. Part 25, Amendment 23 (1970) introduced the requirement that the
high-speed buffet condition should not occur in level flight. The Notice of
Proposed Rule Making (NPRM) specified:
Section §25.251(c) requires that there be no buffeting in normal flight severe
enough to interfere with the control of the airplane, to cause excessive fatigue
to the crew, or to cause structural damage; however, the present regulations
do not require the determination of the combinations of weights, altitudes,
speeds, and load factors at which incipient low and high speed buffeting
occurs. Service experience has shown that these data, commonly referred
to as the Operation V-N envelope, are a valuable aid to the pilot since it pro-
vides the information he needs to avoid the buffet range at low speeds when
the airplane is in a holding pattern and at high altitudes where the speed
range between the occurrence of low and high speed buffeting is small.
The proposal would require the determination of buffet boundary envelopes
at which the onset of perceptible buffeting occurs for the cruise configuration
and that the data be included in the Airplane Flight Manual.

The NPRM also added:

Buffeting is considered to be a warning to the pilot that the airplane is
approaching an undesirable and eventually dangerous flight regime, i.e.,
stall buffeting, high speed buffeting, or maneuvering (load factor) buffeting.
Therefore, in straight flight, such buffet warning should not occur at any
normal operating speed up to VMO /MMO . Stall warning buffeting would,
of course, continue to be permitted to enable compliance with Section
§25.207.

With this, the FAA introduced §25.251(d):

§25.251(d) Buffeting: There may be no perceptible buffeting condition in the
cruise configuration in straight flight at any speed up to VMO /MMO , except
that stall warning buffeting is allowable.

This is a subjective evaluation by the crew that the level of vibration, as

perceived by them, is small enough to be considered not perceptible. FAA
AC 25-7 stipulates that there is “no established criterion for buffet level at
the pilot station. A normal acceleration of +0.05g has been proposed;
however, that will vary from airplane to airplane and may also be affected
by the dynamic response of the accelerometer.” The AC goes on to say,
“This boundary should be established by pilot event.” This often remains a
very subjective evaluation, and the perception level varies from pilot to
pilot. It is easier to define when there is a sudden onset (step change) in
the vibration level than when the vibration gradually increases with air-
speed/Mach or load factor. We will elaborate more on buffeting later in
this chapter.
It is interesting to see some airplanes evolve with changing requirements.
Consider the Boeing 737, which first flew in 1967 in the 737-100 version,
390 Operational Aircraft Performance and Flight Test Practices

followed by the 737-200 version. These versions of the 737 had a certification
.

based on Part 25, Amendments 25-0 to 25-15 per FAA TCDS A16WE. The
next evolution of the 737 series is now referred to as the classics (the 300, 400,
and 500 versions). These were introduced in the 1980s with new high bypass
turbofans and retained the same certification basis of §25.251; thus, they did
not have to adhere to §25.251(d). Both the original models and the classics
had a long-range cruise speed on the order of Mach 0.74, and some crews
have reported they had a noticeable level of buffet below MMO . The
follow-on evolution of the 737, the Next Generation (NG-600, -700, -800,
and -900 versions) came in the 1990s as a response to competition from
Airbus with the A320 with its higher design cruise speed of M0.78. The
NG had a new wing that improved cruise Mach (now a long-range cruise
of M0.78; see Fig. 9.22) and delayed buffeting. Boeing adopted regulations
from §25.251 up to Amendment 25-77. Finally, the Max version came
along in the 2010s containing new engines and more aerodynamics tweaking
to counter new airliners (Bombardier CSeries and Aribus A320NEO). Boeing
did adopt Part 25, Amendments 25-0 to 25-137 for most regulations, but
retained Amendment 25-77 for §25.251.
In Part 25, Amendment 23, as stated previously, the manufacturer was
also mandated, per §25.251(e), to include in the AFM the envelopes of load
factor, speed, altitude, and weight that provide a sufficient range of speeds
and load factors for normal operations; specifically:

§25.251(e) For an airplane with MD greater than 0.6 or with a maximum

operating altitude greater than 25,000 feet, the positive maneuvering load
factors at which the onset of perceptible buffeting occurs must be deter-
mined with the airplane in the cruise configuration for the ranges of airspeed
or Mach number, weight, and altitude for which the airplane is to be

CD
0.0350 CL = 0.4

0.0300
737-300 (Classics)

0.0250 737-200
MMO = 0.82
LRC = 0.78
(Classics)
LRC = 0.74

737-800 (NG)
(NG)

0.0200
0.50 0.60 0.70 0.80 M

Fig. 9.22 Boeing 737 evolution of airplane drag and cruise speed.
 CHAPTER 9 Flight Envelope 391

Buffet boundary
Associated conditions Example
Flaps..................................Up Bank angle....................................30°
Landing gear....................Retracted Corresponding load factor.....1.15g
Weight.............................................550,000 LBS
Pressure altitude.......................43,000 FT
Low speed buffet onset mach......0.779
High speed buffet onset mach...0.908

T
LB 600,000

–F
–
ht

de
eig 550,000

ltitu
W

FT
FT

FT
el

FT
T

T
re a

lev

0F

0F
000
000

000

000
500,000

,00
500
S ea
ssu

10,
15,

20,

25,

30

T
0F
Pre
450,000

,00
35
400,000
FT
350,000 ,0 00
40
T
0F
,00
43 0F
T
,00
45
Bank angle – deg
0 15 30 45 50 55 60

1.0 1.5 2.0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Load factor – g Buffer onset Mach number

Fig. 9.23 Example of buffet boundary envelope for AFM.

certificated. The envelopes of load factor, speed, altitude, and weight must
provide a sufficient range of speeds and load factors for normal operations.
Probable inadvertent excursions beyond the boundaries of the buffet onset
envelopes may not result in unsafe conditions.

The data collected during the flight test program to generate the curve of
Fig. 9.20 (CLbuff vs Mach) can be converted into a curve of (nz W/d vs
Mach). That curve can then be converted into a chart similar to that of
Fig. 9.23.

Example 9.1
Using the data presented in Fig. 9.22, create a buffet envelope (min and max
speed) from 25,000 ft to 45,000 ft for an airplane weighing 550,000 lb flying in
steady level flight (1.0 g). The MMO of the plane is 0.90. If the chart has no data
or goes beyond MMO on the high-speed side, stop the plot at MMO . Also add
the limits for 30-deg bank (1.15 g) and 40-deg bank (1.3 g) maneuvering
capability.

(Continued)
392 Operational Aircraft Performance and Flight Test Practices

Example 9.1 (Continued)

Solution: We get the following plot:

50,000

Max altitude
45,000

40,000
Altitude (ft)

1.1 g
1.0

5g
MMO

g
35,000

1.3
30,000

25,000
0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Note how the low-speed side of the envelope and the high-speed side of
the envelope come closer and closer together as the altitude increases. This
narrowing of the envelope is called the coffin corner, where the airplane’s
ability to fly without encountering buffeting becomes less and less likely.
The European Union Aviation Safety Agency (EASA) proposes an acceptable
means of compliance for §25.251(e) in CS25 Book 2 as follows:
An acceptable means of compliance with the requirement is to estab-
lish the maximum altitude at which it is possible to achieve a positive
normal acceleration increment of 0.3 g without exceeding the buffet
onset boundary.

For the previous example, that altitude would have been approximately
43,300 ft.

Design Dive Speed

The certification regulations for Part 23 and Part 25 airplanes add speed
margins above VMO /MMO where the airplane needs to be shown to be safe
should an inadvertent speed excursion occur. One such margin for Part 25 is
the design dive speed (VD /MD ) defined by §25.335(b) (see Fig. 9.24).
§25.335(b) Design Dive Speed, VD . VD must be selected so that VC /MC is not
greater than 0.8 VD /MD , or so that the minimum speed margin between
VC /MC and VD /MD is the greater of the following values:
 CHAPTER 9 Flight Envelope 393

Altitude

20 s

7.5 deg

Recovery

Airspeed
VMO
mach
MMO

Fig. 9.24 Maneuver per §25.335(b)(1).

(1) From an initial condition of stabilized flight at VC /MC , the airplane is

upset, flown for 20 seconds along a flight path 7.58 below the initial path,
and then pulled up at a load factor of 1.5g (0.5g acceleration increment).
The speed increase occurring in this maneuver may be calculated if
reliable or conservative aerodynamic data is used. Power as specified in
§25.175(b)(1)(iv) is assumed until the pullup is initiated, at which time
power reduction and the use of pilot controlled drag devices may be
assumed;
(2) The minimum speed margin must be enough to provide for atmospheric
variations (such as horizontal gusts, and penetration of jet streams and
cold fronts) and for instrument errors and airframe production
variations. These factors may be considered on a probability basis. The
margin at altitude where MC is limited by compressibility effects must
not less than 0.07M unless a lower margin is determined using a rational
analysis that includes the effects of any automatic systems. In any case,
the margin may not be reduced to less than 0.05M.

This version of the regulation was made final in 1997.

VD /MD Versus VDF /MDF

There have been some discussions between authorities and applicants
over the interpretation of VDF /MDF and VD /MD over the years. The follow-
ing statement in FAA AC 25-7C (and in AC 25-7D) has been at the heart of
this discussion (underline added by author):
Section 25.1505 states that the speed margin between VMO /MMO , and
VD /MD or VDF /MDF , as applicable, “may not be less than that determined
under § 25.335(b) or found necessary during the flight tests conducted under
§ 25.253.” Note that one speed margin must be established that complies
with both § 25.335(b) and § 25.253. Therefore, if the applicant chooses a
VDF/MDF that is less than VD/MD, then VMO/MMO must be reduced by
the same amount (i.e., compared to what it could be if VDF /MDF
were equal to VD /MD ) in order to provide the required speed margin to
VDF /MDF .
394 Operational Aircraft Performance and Flight Test Practices

The heart of the discussion focuses on the underlined part of the AC para-
graph. Whereas some manufacturers have defined MD and MDF
using §25.335(b), some have not; they have instead elected to use an MD of
0.07 M regardless of the value obtained by §25.335(b) analysis. That speed
is used for flutter and structural analysis prior to the flight test program.
These same manufacturers have then used an MDF (flight demonstration)
per §25.335. If the discussion moves away from the letters MD and MDF
and instead focuses on the actual Mach number that must be demonstrated,
the issues go away, and the airplane design meets the intent of §25.1505 (see
underlined part of rule):
§25.1505 The maximum operating limit speed (VMO /MMO , airspeed or
Mach Number, whichever is critical at a particular altitude) is a speed that
may not be deliberately exceeded in any regime of flight (climb, cruise, or
descent), unless a higher speed is authorized for flight test or pilot training
operations. VMO/MMO must be established so that it is not greater than
the design cruising speed VC and so that it is sufficiently below VD/MD
or VDF/MDF, to make it highly improbable that the latter speeds will be inad-
vertently exceeded in operations. The speed margin between VMO/MMO
and VD/MD or VDF/MDF may not be less than that determined under
§25.335(b) or found necessary during the flight tests conducted under
§25.253.

High-Speed Characteristics
With the desired VMO /MMO established, the applicant must then verify
that the airplane will not inadvertently be exposed to speeds greater than the
design dive speed [VD /MD per §25.335(b)]. The minimum prescribed verifi-
cation for a Part 25 airplane is defined by §25.253. This will include inten-
tional pitch and roll upsets, and verification that the airplane can be easily
recovered to a speed below VMO /MMO with no exceptional piloting skills
and without being exposed to excessive vibration and buffeting.
§25.253 High-speed characteristics.
(a) Speed increase and recovery characteristics. The following speed
increase and recovery characteristics must be met:
(1) Operating conditions and characteristics likely to cause inadvertent
speed increases (including upsets in pitch and roll) must be
simulated with the airplane trimmed at any likely cruise speed
up to VMO /MMO . These conditions and characteristics include
gust upsets, inadvertent control movements, low stick force
gradient in relation to control friction, passenger movement,
leveling off from climb, and descent from Mach to airspeed
limit altitudes.
(2) Allowing for pilot reaction time after effective inherent or artificial
speed warning occurs, it must be shown that the airplane can be
recovered to a normal attitude and its speed reduced to VMO /MMO ,
without—
(i) Exceptional piloting strength or skill;
(ii) Exceeding VD /MD , VDF /MDF , or the structural limitations; and
 CHAPTER 9 Flight Envelope 395

(iii) Buffeting that would impair the pilot’s ability to read the
instruments or control the airplane for recovery.
(3) With the airplane trimmed at any speed up to VMO /MMO , there
must be no reversal of the response to control input about any axis
at any speed up to VDF /MDF . Any tendency to pitch, roll, or yaw
must be mild and readily controllable, using normal piloting
techniques. When the airplane is trimmed at VMO /MMO , the slope
of the elevator control force versus speed curve need not be stable
at speeds greater than VFC /MFC , but there must be a push force at
all speeds up to VDF /MDF and there must be no sudden or excessive
reduction of elevator control force as VDF /MDF is reached.
(4) Adequate roll capability to assure a prompt recovery from a lateral
upset condition must be available at any speed up to VDF /MDF .
(5) With the airplane trimmed at VMO /MMO , extension of the
speedbrakes over the available range of movements of the pilot’s
control, at all speeds above VMO /MMO , but not so high that VDF /
MDF would be exceeded during the maneuver, must not result in:
(i) An excessive positive load factor when the pilot does not take
action to counteract the effects of extension;
(ii) Buffeting that would impair the pilot’s ability to read the
instruments or control the airplane for recovery; or
(iii) A nose down pitching moment, unless it is small.
(b) Maximum speed for stability characteristics, VFC /MFC. VFC /MFC is the
maximum speed at which the requirements of §§25.143(g), 25.147(f),
25.175(b)(1), 25.177(a) through (c), and 25.181 must be met with flaps
and landing gear retracted. Except as noted in §25.253(c), VFC /MFC may
not be less than a speed midway between VMO /MMO and VDF /MDF ,
except that, for altitudes where Mach number is the limiting factor, MFC
need not exceed the Mach number at which effective speed
warning occurs.
(c) Maximum speed for stability characteristics in icing conditions. The
maximum speed for stability characteristics with the most critical of
the ice accretions defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g), at which the requirements of
§§25.143(g), 25.147(f), 25.175(b)(1), 25.177(a) through (c), and 25.181
must be met, is the lower of:
(1) 300 knots CAS;
(2) VFC ; or
(3) A speed at which it is demonstrated that the airframe will be free of
ice accretion due to the effects of increased dynamic pressure.

Out-of-Trim Characteristics
A new requirement that came along in Part 25, Amendment 42 (1978)
was to demonstrate that an airplane could be recovered from a mistrim con-
dition at high speed. The NPRM introducing the requirement stated: “Service
experience indicates that out-of-trim conditions can occur in flight for
various reasons, and that the control and maneuvering characteristics of
the airplane may be critical in recovering from upsets.”
The amount of mistrim that needs to be demonstrated by flight test or
analysis is specified in §25.255; the rule remains unchanged since its
introduction:
396 Operational Aircraft Performance and Flight Test Practices

§25.255 Out-of-trim characteristics

(a) From an initial condition with the airplane trimmed at cruise speeds
up to VMO /MMO , the airplane must have satisfactory maneuvering
stability and controllability with the degree of out-of-trim in both the
airplane nose-up and nose-down directions, which results from the
greater of—
(1) A three-second movement of the longitudinal trim system at its
normal rate for the particular flight condition with no aerodynamic
load (or an equivalent degree of trim for airplanes that do not have a
power-operated trim system), except as limited by stops in the trim
system, including those required by §25.655(b) for adjustable
stabilizers; or
(2) The maximum mistrim that can be sustained by the autopilot while
maintaining level flight in the high speed cruising condition.
(b) In the out-of-trim condition specified in paragraph (a) of this section,
when the normal acceleration is varied from þ1 g to the positive and
negative values specified in paragraph (c) of this section—
(1) The stick force vs. g curve must have a positive slope at any speed up
to and including VFC /MFC ; and
(2) At speeds between VFC /MFC and VDF /MDF the direction of the
primary longitudinal control force may not reverse.
(c) Except as provided in paragraphs (d) and (e) of this section, compliance
with the provisions of paragraph (a) of this section must be demonstrated
in flight over the acceleration range—
(1) 21 g to þ2.5 g; or
(2) 0 g to 2.0 g, and extrapolating by an acceptable method to 21 g and
þ2.5 g.
(d) If the procedure set forth in paragraph (c)(2) of this section is used to
demonstrate compliance and marginal conditions exist during flight
test with regard to reversal of primary longitudinal control force, flight
tests must be accomplished from the normal acceleration at which a
marginal condition is found to exist to the applicable limit specified in
paragraph (b)(1) of this section.
(e) During flight tests required by paragraph (a) of this section, the limit
maneuvering load factors prescribed in §§25.333(b) and 25.337, and
the maneuvering load factors associated with probable inadvertent
excursions beyond the boundaries of the buffet onset envelopes
determined under §25.251(e), need not be exceeded. In addition, the
entry speeds for flight test demonstrations at normal acceleration values
less than 1 g must be limited to the extent necessary to accomplish a
recovery without exceeding VDF /MDF .
( f) In the out-of-trim condition specified in paragraph (a) of this section, it
must be possible from an overspeed condition at VDF /MDF to produce at
least 1.5 g for recovery by applying not more than 125 pounds of
longitudinal control force using either the primary longitudinal control
alone or the primary longitudinal control and the longitudinal trim
system. If the longitudinal trim is used to assist in producing the required
load factor, it must be shown at VDF /MDF that the longitudinal trim can
be actuated in the airplane nose-up direction with the primary surface
loaded to correspond to the least of the following airplane nose-up
control forces:
(1) The maximum control forces expected in service as specified in
§§25.301 and 25.397.
(2) The control force required to produce 1.5 g.
(3) The control force corresponding to buffeting or other phenomena of
such intensity that it is a strong deterrent to further application of
primary longitudinal control force.
 CHAPTER 9 Flight Envelope 397

Flutter Speed and Reversal Speed

Airplane structures are flexible and must absorb the aerodynamic energy
of the airplane flying through the atmosphere. Flutter occurs when there is an
imbalance among the aerodynamic forces, the stiffness of the airframe, and
the inertial forces of the structure. When the structural damping of the air-
plane, most of the time the wing (with some exceptions like the accident of
the Grob SPn, discussed in the box), is insufficient to damp out the airframe
motion, structural failure can occur.

On 29 Nov. 2006, the second prototype of the light business jet Grob SPn
crashed while performing a flight demonstration for a group of visitors. The
airplane took off from runway 33 at Mindelheim-Mattsies Airport in
Germany and performed a righthand circuit. As the airplane was lining up
for a fly-by, part of the stabilizer broke off; the pilot then lost control and
crashed. The probable airspeed at the time of the accident was between 240
and 270 kt, well below the maximum speed allowed by flutter test, 297 kt.
Probable cause: The manufacturer changed the design of the control sur-
faces on this second prototype and did not clear the new structure. The tail-
plane separated due to flutter, leading to an uncontrollable airplane.

Prior to 1992, the requirements of §25.629 defined a minimum margin

for flutter of 20% above the dive speed. This verification could be done by
analysis only. Amendment 77 changed Part 25 (1992), §25.629, to its
present form where the margin between VD /MD and the worst flutter
speed VF was reduced to 15%, but validation via flight testing [§25.629(e)]
was added to verify that the structural damping was not reducing as the air-
plane approached VDF /MDF .
§25.629 Aeroelastic stability requirements
(a) General. The aeroelastic stability evaluations required under this section
include flutter, divergence, control reversal and any undue loss of
stability and control as a result of structural deformation. The aeroelastic
evaluation must include whirl modes associated with any propeller or
rotating device that contributes significant dynamic forces. Compliance
with this section must be shown by analyses, wind tunnel tests, ground
vibration tests, flight tests, or other means found necessary by the
Administrator.
(b) Aeroelastic stability envelopes. The airplane must be designed to be free
from aeroelastic instability for all configurations and design conditions
within the aeroelastic stability envelopes as follows:
(1) For normal conditions without failures, malfunctions, or adverse
conditions, all combinations of altitudes and speeds encompassed by
the VD /MD versus altitude envelope enlarged at all points by an
increase of 15 percent in equivalent airspeed at both constant Mach
number and constant altitude. In addition, a proper margin of
stability must exist at all speeds up to VD /MD and, there must be no
large and rapid reduction in stability as VD /MD is approached. The
398 Operational Aircraft Performance and Flight Test Practices

enlarged envelope may be limited to Mach 1.0 when MD is less than

1.0 at all design altitudes, and
(2) For the conditions described in §25.629(d) below, for all approved
altitudes, any airspeed up to the greater airspeed defined by;
(i) The VD /MD envelope determined by §25.335(b); or,
(ii) An altitude-airspeed envelope defined by a 15 percent increase in
equivalent airspeed above VC at constant altitude, from sea level
to the altitude of the intersection of 1.15 VC with the extension
of the constant cruise Mach number line, MC , then a linear
variation in equivalent airspeed to MC þ .05 at the altitude of the
lowest VC /MC intersection; then, at higher altitudes, up to the
maximum flight altitude, the boundary defined by a .05 Mach
increase in MC at constant altitude.
§25.629 (e) Flight flutter testing. Full scale flight flutter tests at speeds up to
VDF /MDF must be conducted for new type designs and for modifications to a
type design unless the modifications have been shown to have an insignifi-
cant effect on the aeroelastic stability. These tests must demonstrate that
the airplane has a proper margin of damping at all speeds up to VDF /
MDF , and that there is no large and rapid reduction in damping as VDF /
MDF , is approached. If a failure, malfunction, or adverse condition is simu-
lated during flight test in showing compliance with paragraph (d) of this
section, the maximum speed investigated need not exceed VFC /MFC if it
is shown, by correlation of the flight test data with other test data or analyses,
that the airplane is free from any aeroelastic instability at all speeds within
the altitude-airspeed envelope described in paragraph (b)(2) of this section.

Summary of the High-Speed Requirements

The regulatory limits that are used to validate the maximum operating
speed (VMO /MMO ) are summarized by Fig. 9.25.

Maximum Altitude
The maximum altitude an airplane will be authorized to fly to is impacted
by several factors, some of which are summarized in Fig. 9.26.
The maximum certified altitude rests with the applicant and is dependent
on what is shown to be compliant by analysis and/or test.
FAA 14 CFR §25.1527, Ambient Air Temperature and Operating Alti-
tude, specifies:
The extremes of the ambient air temperature and operating altitude for
which operation is allowed, as limited by flight, structural, powerplant, func-
tional, or equipment characteristics, must be established.

Some other considerations the manufacturer must account for include:

• Pressurized fuselage, DP maintained to not exceed an equivalent altitude
inside (not more than 8000 ft per §25.841, with more and more airplanes
now going towards lower 6000-ft cabin altitude). The Sino-Swearingen
SJ30-2 business jet had sea level pressure in the cabin at 41,000 ft, which
 CHAPTER 9 Flight Envelope 399

25.1505
Hp 25.335(b)
& 25.629(b)(2)

Can be less than M1.0 if MD<1.0

0.07 M
0.05 M min
25.629(b)(1)
MMO
MDmin
15%

MDF
MD

25.253
VMO

VF
VDF
Flutter speed, analysis
VD of flight test results
VDmin

Fig. 9.25 Summary of Part 25 high-speed margins.

45,000

40,000 Vbuff
Vbuff
35,000
MMO
30,000
Altitude (ft)

25,000

20,000
Vmin
15,000
VSR
10,000
VSW VMO
5000
Vmax
0
0 100 200 300 400 500 600
Airspeed (KCAS)

Fig. 9.26 Speed-altitude limits.

400 Operational Aircraft Performance and Flight Test Practices

dictates the differential pressure. This, in turn, dictates loop forces

experienced by the fuselage. Safety margins are then applied to those forces
per §25.365:

§25.365 Pressurized compartment loads:

(a) The airplane structure must be strong enough to withstand the flight
loads combined with pressure differential loads from zero up to the
maximum relief valve setting.
...
(d) The airplane structure must be designed to be able to withstand the
pressure differential loads corresponding to the maximum relief valve
setting multiplied by a factor of 1.33 for airplanes to be approved for oper-
ation to 45,000 feet or by a factor of 1.67 for airplanes to be approved
for operation above 45,000 feet, omitting other loads.

• Emergency descent capability (ability to descend fast enough) per

§25.841(a)(1) and (2):

(1) If certification for operation above 25,000 feet is required, the airplane
must be designed so that occupants will not be exposed to cabin pressure
altitudes in excess of 15,000 feet after any probable failure condition in
the pressurization system.
(2) The airplane must be designed so that occupants will not be exposed to a
cabin pressure altitude that exceeds the following after decompression
from any failure condition not shown to be extremely improbable:
(i) Twenty-five thousand (25,000) feet for more than 2 minutes; or
(ii) Forty thousand (40,000) feet for any duration.

Temperature Envelope
Other limits that must be observed by the air crew are the maximum and
minimum temperatures for operation of the aircraft. These are limits within
which the manufacturer has demonstrated that the aircraft can operate
reliably and safely. FAA requirement §25.1527 states:
§25.1527 Ambient air temperature and operating altitude. The extremes of
the ambient air temperature and operating altitude for which operation is
allowed, as limited by flight, structural, powerplant, functional, or equipment
characteristics, must be established.

Some manufacturers may elect to show the temperature limitations in terms

of static air temperature (see Fig. 9.27 for an example); others will do so in
total air temperature. Some temperature limits may be specific to a certain
system operation, such as:

• Do not turn bleed wing anti-icing system on above a total temperature of

þ108C.
• Do not inflate wind deicing boots below a static air temperature of
–408C.
 CHAPTER 9 Flight Envelope 401

50,000

40,000

Pressure altitude (ft) 30,000

ISA

20,000

10,000

–10,000
–100 –80 –60 –40 –20 0 20 40 60
Temperature (°C)

Fig. 9.27 Typical temperature envelope.

Flaps Out and Gear Down Envelopes

The previous discussion focused on the flaps up, gear up configuration.
Additional limitations may be imposed by different configurations; the
most common configurations are flaps down (various deflection angles)
and gear down or in transition.

Flaps Down Airspeeds and Altitude

Extending the flaps to improve (reduce) the low airspeed limits of the air-
plane translates into shorter takeoff and landing distances. It will also impose
new maximum airspeed limits (much lower) on the airplane due to the
additional air loads on the “extended” structure (see Fig. 9.28). The low
and high airspeed limits must allow the airplane to meet the same flying
characteristics as the flaps up conditions.
For the high-speed limit, FAA 14 CFR §25.1511 specifies, “The estab-
lished flap extended speed VFE must be established so that it does not
exceed the design flap speed VF chosen under §§25.335(e) and 25.345, for
the corresponding flap positions and engine powers.”
FAA 14 CFR §25.335 Design Airspeeds, (e) Design Flap Speed VF :
(1) The design flap speed for each flap position (established in accordance
with §25.697(a)) must be sufficiently greater than the operating speed
recommended for the corresponding stage of flight (including balked
landings) to allow for probable variations in control of airspeed and for
transition from one flap position to another.
402 Operational Aircraft Performance and Flight Test Practices

(2) If an automatic flap positioning or load limiting device is used, the speeds
and corresponding flap positions programmed or allowed by the device
may be used.
(3) VF may not be less than—
(i) 1.6 VS1 with the flaps in takeoff position at maximum takeoff weight;
(ii) 1.8 VS1 with the flaps in approach position at maximum landing
weight, and
(iii) 1.8 VS0 with the flaps in landing position at maximum landing
weight.

Most manufacturers also specify a maximum operating limit for flaps out.
Above a certain maximum altitude, typically between 15,000 ft and 20,000 ft,
there is no practical application for flaps out conditions (no airports). As
the altitude increases, so does the Mach number for a given calibrated air-
speed, and additional work/flight testing may be required to validate the air-
plane handling, stall speeds, and flap loads; if the envelope were not limited,
the manufacturer would be required to provide this extra level of effort for no
obvious benefit.
So, as one can see, although it is a simple concept to select a maximum
operating speed for flaps, a lot of work goes into validating that speed.
A similar story unfolds when we look at the maximum airspeed for
landing gear. Airplanes that have the capability to retract/extend the
landing gear will have maximum speeds at which these systems where
shown to be fully compliant with Part 25 (see Fig. 9.29). That limit speed
may be set by the maximum design load on the landing gear itself or on
the landing gear doors (e.g., maximum lateral load at maximum expected
sideslip angle in service at the proposed maximum landing gear speed).

45,000
40,000
35,000
30,000
Altitude (ft)

25,000 Flaps up
20,000 envelope

15,000

10,000
VSR VSW Flaps VFE
5000 out
0
0 50 100 150 200 250 300 350
Airspeed (KCAS)

Fig. 9.28 Flaps out flight envelope.

 CHAPTER 9 Flight Envelope 403

45,000

40,000

35,000

30,000
Altitude (ft)

25,000

20,000
15,000
VLO VLE
10,000

5000

0
0 50 100 150 200 250 300 350
Airspeed (KCAS)

Fig. 9.29 Landing gear extended and operating limit speeds.

The certification regulation for Part 25 allows for providing two (or three)
design landing gear speeds via §25.1515:
§25.1515 Landing gear speeds

[(a) The established landing gear operating speed or speeds, VLO , may not
exceed the speed at which it is safe both to extend and to retract the
landing gear, as determined under Sec. 25.729 or by flight characteristics.
If the extension speed is not the same as the retraction speed, the two
speeds must be designated as VLOðEXTÞ and VLOðRETÞ , respectively.]
(b) The established landing gear extended speed VLE may not exceed the
speed at which it is safe to fly with the landing gear secured in the fully
extended position, and that determined under Sec. 25.729.

Some manufacturers will limit the gear down condition to a maximum alti-
tude, say 20,000 ft, beyond which they do not have to show compliance.
Others will not limit the altitude but may impose a combination of airspeed
and Mach number (lowest of the two, similar to VMO /MMO ). On some of
these airplanes with no altitude restrictions, the landing gear may actually
be used in an emergency descent (if authorized by the manufacturer) to
help the airplane come down faster.
Other configurations may exist that would have similar limitations. A
search of the FAA TCDS database for various manufacturers has also
shown the following:
• Maximum landing light extended
• No spoiler deployment with flaps down in flight
• Maximum speed to deploy arresting parachute
404 Operational Aircraft Performance and Flight Test Practices

Placard Speeds
To help the flight crew remember all the different limit speeds that can
result from a change in configuration, the FAA requires that a placard
showing such limits be clearly visible near the actuation point/handle for
the flaps.
§25.1563 Airspeed placard. A placard showing the maximum airspeeds for
flap extension for the takeoff, approach, and landing positions must be
installed in clear view of each pilot.

This requirement is also carried over to other systems. A search of the Inter-
net for “cockpit placard” will reveal several different ways manufacturers have
implemented this requirement.

Reduced Vertical Separation Minimum (RVSM)

Another heavily regulated part of the flight envelope is the altitude band
between flight level 290 (29,000 ft) and 410 (41,000 ft). Starting in 1997, the
reduced vertical separation minimum (RVSM) envelope, the flight levels
between airplanes going in opposite directions, were progressively reduced
to only 1000-ft separation in various parts of the world from the
original 2000-ft separation (see Fig. 9.30). This was done to offer better
routing options for airliners, most of which are designed to fly in those
flight levels.

Cruising flight levels

Prior to RVSM With RVSM

FL410

FL400
FL390

FL380

FL370
FL360

FL350

FL340
FL330

FL320
FL310
FL300

FL290

Fig. 9.30 Evolution of authorized flight levels, RVSM.

 CHAPTER 9 Flight Envelope 405

To better understand why the airways were not made available sooner, it
is worth looking at a brief history of air navigation, starting with prejet airli-
ners. The following is extracted from [3]:
In the 1940’s, the airliners were few, propeller driven, with limited altitude
capability (typically below 20,000 ft). Vertical separation was 1,000 ft with
some exception where 500 ft is authorized when:
(1) The airplanes are being flown in conditions of flight visibility of less than
3 miles but not less than 1 mile; and
(2) The airplanes are holding above a well-defined top of cloud during hours
of darkness with forward visibility of not less than 1 mile.

Airplanes are flying airways where flight paths cross at or near reporting
points. There was no required separation for enroute traffic during daylight
hours if the airplanes were above well-defined top of clouds with visibility of
at least 3 miles. ‘Holding on top’ placed the responsibility of separation on
the flight crew. The mid-air collision of a TWA Constellation and a United
Airlines DC-7 over the Grand Canyon in 1956 showed the limitation of
this approach.
The 1950’s introduced jet powered airliners with real capabilities to fly in
the low Stratosphere. The International Civil Aviation Organization (ICAO)
formed the Vertical Separation Panel in 1954 to identify the factors that
have the greatest impact on the loss of separation. The panel proposed steps
to reduce or eliminate those factors. The ICAO 1958 RAC/SAR divisional
meeting specified “The vertical separation minimum between IFR traffic
shall be a nominal 300 meters (1000 ft) below an altitude of 8850 meters
(29000 ft) or flight level 290 and a nominal 600 meters (2000 ft) at or above
this level, except where, on the basis of regional air navigation agreements, a
lower level is prescribed for the change to a nominal 600 meters (2000 ft) ver-
tical separation minimum.

The altitude of 29,000 ft was chosen as a practical limit for the air data system
of the time. As an illustration of the precision required by the air data system,
consider that the change in pressure per 100 ft of altitude varies per Fig. 9.31.
The resolution of the air data system at 41,000 ft needs to be almost half that
at 29,000 ft, and as we saw in Chapter 4, this includes not only the capability
of the sensor, but also that of the installation of the probe, the precision of the
fuselage contour where the probe is installed, and so forth. The tight require-
ments of air data system precision do not stop at the design/manufacturing
stage, where the manufacturer is responsible for the build, but also apply to
the operators of the airplane and their maintenance program to maintain
such accuracy in the air data system.

Civil Aeronautics Board Accident Investigation Report, File no. 1-0090,

17 April 1957: On 30 June 1956, a Trans World Airlines (TWA) Lockheed
1049A, N6902C, collided with a United Air Lines Douglas DC-7, N6324C,
at approximately 21,000 ft over the Grand Canyon National Park. Both air-
planes fell into the canyon; there were no survivors. The collision took
406 Operational Aircraft Performance and Flight Test Practices

place in uncontrolled airspace, where it was the pilots’ own responsibility to

maintain separation.
TWA Flight 2 was a scheduled flight from Los Angeles to Kansas City with
a planned flight altitude of 19,000 ft. At 9h21, through company radio com-
munication, Flight 2 reported approaching Daggett and requested a change
in flight plan altitude to 21,000 ft. ARTC (Los Angeles Center) advised they
were unable to approve the requested altitude because of traffic (United
Flight 718). Flight 2 requested clearance of 1000 ft on top (still IFR rules,
but at the time this allowed for temporary lifting of separation restrictions
normally applied by ATC; it placed the responsibility for maintaining safe sep-
aration on the crew). ARTC cleared the flight. At 0959, Flight 2 reported its
position, having passed Lake Mohave at 0955, 1000 ft on top at 21,000 ft
and estimating it would reach Painted Desert at 1031. This was the
last communication.
United Flight 718 was a scheduled flight from Los Angeles to Chicago. It
left LA 3 minutes after Flight 2. It had a flight plan altitude of 21,000 ft. It
reported its position at approximately 0958 stating it would be over Painted
Desert at approximately 1031. At 1031 an unidentified radio transmission
was heard by Aeronautical Radio communicators at Salt Lake City and
San Francisco. The message “Salt Lake, United 718. .. ah . . . we’re going in.”
The investigation revealed that the two airplanes, flying at the same altitude
and nearly the same speed, collided at a closing angle of about 25 deg while the
pilots were maneuvering around towering cumulus clouds. Postcrash analysis
showed that United Flight 718 was banked right wing down and pitching down-
ward, indicating the pilot might have tried to avoid the collision.

The 1988 ICAO Review of the General Concept of Separation Panel

(RGCSP) concluded that RVSM was technically feasible with the emerging
technology without placing undue burden on the airplane manufacturers.
The panel produced draft RVSM guidance material that was approved in
1990. RVSM was then gradually implemented from 1997 to 2011 (see Fig. 9.32).
As mentioned previously, RVSM approval is both airplane (certification)
and operators (ops procedures and maintenance program). Operational
regulations for RVSM are found under FAA 14 CFR, Part 91, Appendix G,
which specifies [1]:

Within RVSM airspace, air traffic control (ATC) separates aircraft by a

minimum of 1,000 feet vertically between flight level (FL) 290 and FL 410
inclusive. RVSM airspace is special qualification airspace; the operator and
the aircraft used by the operator must be approved by the Administrator.
Air-traffic control notifies operators of RVSM by providing route planning
information.

The guidance material for operators falls under FAA AC 91-85 [2], Author-
ization of Aircraft and Operators for Flight in Reduced Vertical Separation
Minimum Airspace.
 CHAPTER 9 Flight Envelope 407

0.00

RVSM
–0.01
Delta pressure per 100 ft (psi)

–0.02

–0.03

–0.04

–0.05

–0.06
0
00

00

00

00

00

00

00

00

00

00
50

,0

,0

,0

,0

,0

,0

,0

,0

,0
10

15

20

25

30

35

40

45

50
Pressure altitude (ft)

Fig. 9.31 Required air data pressure resolution for altitude measurement.

RVSM Wordlwide implemented as of November 2011

Canada North
Canada South Europe 1/02 Japan/Korea
4/02
1/05 9/05
Caucasus
Area
Domestic US 3/05
IRAQ FIR
1/05 NAT 3/97
3/11 EURASIA
China Pacific
11/11 2/00
11/07
Mid East
Pacific EUR/SAM 11/03 Western Pacific
2/00 WATRS 11/01 Corridor 1/02
Asia/Europe South China Sea
South of Himalayas 11/03 2/02
Africa
9/08
EURASIA
CAR/SAM Australia
1/05 Russian 11/01
Federation
Afghanistan
Kazakhstan
Kyrgyzstan
Mongolia
Implemented Tajikistan
Uzbekistan

Fig. 9.32 RVSM airspace as of 2011.

408 Operational Aircraft Performance and Flight Test Practices

RVSM Certification: Airplane

At a minimum, for an airplane to be RVSM approved, it must have the
following equipment:
• Two operational independent altitude measurement systems
• At least one automatic altitude control system (such as an autopilot) that
controls the altitude:
• Within a tolerance band of +65 ft about an acquired altitude when the
airplane is operated in straight and level flight under nonturbulent,
nongust conditions
• Within a tolerance band of +130 ft under nonturbulent, nongust
conditions for airplanes for which application for type certification
occurred on or before 9 April 1997 that are equipped with an
automatic altitude control system with flight management/
performance system inputs
• The airplane must be equipped with an altitude alert system that signals an
alert when the altitude displayed to the flight crew deviates from the
selected altitude by more than:
• +300 ft for airplanes for which application for type certification was
made on or before 9 April 1997
• +200 ft for airplanes for which application for type certification is
made after 9 April 1997.
• A Traffic Alert and Collision Avoidance System (TCAS) compatible with
RVSM operations
The requirements for the altitude measurement system are also more
stringent than the baseline requirements of §§25.1323 and 25.1325 FAA
Part 91 Appendix G specifies (underline added by author):

W/

Min drag Basic

W/ envelope
MMO
1.3 g to
Normal

Thrust limit
cruise

buffet 4M
0.0
+
LR
C
VMO
MMO Min w/ at FL290

Mach
W/
Normal
cruise

C M
LR 04 Thrust limit
0. Full
+ envelope
C
LR MMO
Normal
cruise

VMO 4M
Min w/ at FL290 0.0
+
C
LR
Mach Min w/ at FL290
VMO

Mach

Fig. 9.33 Example of RVSM flight envelopes.

 CHAPTER 9 Flight Envelope 409

Altimetry system error containment: Group aircraft for which application for
type certification is made after April 9, 1997. To approve group aircraft for
which application for type certification is made after April 9, 1997, the
Administrator must find that the altimetry system error (ASE) is contained
as follows:
(1) At the point in the full RVSM flight envelope where mean ASE
reaches its largest absolute value, the absolute value may not exceed
80 feet.
(2) At the point in the full RVSM flight envelope where mean ASE plus three
standard deviations reaches its largest absolute value, the absolute value
may not exceed 200 feet.
Altimetry system error containment: Nongroup aircraft. To approve a non-
group aircraft, the Administrator must find that the altimetry system error
(ASE) is contained as follows:
(1) For each condition in the basic RVSM flight envelope, the largest
combined absolute value for residual static source error plus the avionics
error may not exceed 160 feet.
(2) For each condition in the full RVSM flight envelope, the largest
combined absolute value for residual static source error plus the avionics
error may not exceed 200 feet.

An airplane group approval implies that the airplanes falling under this cat-
egory satisfy the following:
• They have been manufactured to the same design and have been approved
under the same type certificate.
• The static system for each airplane is installed in a manner and position
that is the same as those of the other airplanes in the group, and the same
static source error correction (SSEC) has been applied to each airplane in
the group.
• The avionics units installed in each airplane meet the minimum RVSM
equipment requirements.
An airplane approved under a nongroup approval is simply one that was
individually approved.

RVSM Flight Envelope

FAA Part 91, Appendix G defines an airplane’s RVSM flight envelope as
including the range of Mach number, W/d, and altitudes over which that air-
plane is approved to be operated in cruising flight within an RVSM airspace.
Appendix G also specifies two possible envelopes (see Fig. 9.33):

(a) The full RVSM flight envelope is bounded as follows:

(1) The altitude flight envelope extends from FL 290 upward to the
lowest altitude of the following:
(i) FL 410 (the RVSM altitude limit);
(ii) The maximum certificated altitude for the aircraft; or
(iii) The altitude limited by cruise thrust, buffet, or other flight
limitations.
410 Operational Aircraft Performance and Flight Test Practices

(2) The airspeed flight envelope extends:

(i) From the airspeed of the slats/flaps-up maximum endurance
(holding) airspeed, or the maneuvering airspeed, whichever is
lower;
(ii) To the maximum operating airspeed (VMO /MMO ), or airspeed
limited by cruise thrust buffet, or other flight limitations,
whichever is lower.
(3) All permissible gross weights within the flight envelopes defined in
paragraphs (1) and (2) of this definition.
(b) The basic RVSM flight envelope is the same as the full RVSM flight
envelope except that the airspeed flight envelope extends:
(1) From the airspeed of the slats/flaps-up maximum endurance
(holding) airspeed, or the maneuver airspeed, whichever is lower;
(2) To the upper Mach/airspeed boundary defined for the full RVSM
flight envelope, or a specified lower value not less than the
long-range cruise Mach number plus .04 Mach, unless further
limited by available cruise thrust, buffet, or other flight limitations.

Conclusion
This chapter provided a quick overview of the various limits that exist on
the airplane’s flight envelope for a Part 25 transport category airplane. As was
seen, the flight envelope is heavily regulated based on years of industry and
agencies’ experience.
Although most regulations may not be discussed further during our
analysis of the performance of airplanes, it is important for the reader to
remember that some limits imposed on the plane may be derived from these.

References
[1] Federal Aviation Administration, “Operations in Reduced Vertical Separation
Minimum (RVSM) Airspace,” FAA 14 CFR Part 91, Appendix G, 23 October 2017.
[2] Federal Aviation Administration, “FAA Reduced Vertical Separation Minimum
(RVSM) Advisory Circular 91-85A,” 21 July 2016.
[3] Silva, S., A Brief History of RVSM, Air Navigation Bureau, ICAO HQ, April 2010.
[4] Federal Aviation Administration, “Air Traffic by the Numbers,” https://www.faa.gov/
air_traffic/by_the_numbers/.
[5] Schroijen, M. J. T., and Slingerland, R., “Propeller Slipstream Effects on Directional
Aircraft Control with One Engine Inoperative,” AIAA 2007-1046, 2007.
Chapter 10 Cruise
Performance
and Flight
Endurance

Chapter Objective
The cruise segment of a flight is when one optimizes the distance covered
per unit of energy capability expenditure. To achieve this optimization, one
needs to understand the capability of the airframe aerodynamics, the engine
operating efficiencies, and possibly add a third system, the propeller. We
will review what impacts the specific air range and the overall measure of
cruise efficiency, discuss range capability, and introduce the potential of elec-
tric propulsion.

Defining Range

O
ne of the most important aspects of an airplane is its ability to cover
a certain distance to perform its mission. One will often see quotes
stating “the airplane’s range is. . .,” “the radius of action of this
plane is. . .,” or “the combat radius is. . .”. These quotes are a means to
define an airplane’s ability to cover a given distance under specified
flight conditions.
We define range as the ability to cover a certain distance on a given load
of fuel and for a given set of flight conditions. We will also cover electric air-
planes, for which range becomes proportional to the available energy carried
and a set of flight conditions.
At a minimum, one must expect the following flight conditions when
defining range capability:
• Engine start and taxi
• Takeoff and initial climb
• Climb to the desired cruise altitude
• Cruise phase
• Descent back to landing
• Possible loiter awaiting landing clearance
• Approach and landing
• Taxi and engine shutdown

411
412 Operational Aircraft Performance and Flight Test Practices

NBAA range

3rd step 200 n miles

H cruise I Fuel reserve
Optimum M Economy N
2nd step
climbs cruise
F cruise G Descent
1st step Descent
D cruise E 5 min hold 3000 fpm max
3000 fpm max
L 5000 ft
C Fuel = 6.7 lb/US Gal Sea level Sea level
Sea level Takeoff 1 min full power J K O
A B Optimum
climbs
Taxi, 10 min idle Approach & land
Fuel = 5 min hold @ 5000 ft

Fig. 10.1 NBAA instrument flight rules (IFR) jet airplane range.

Then, one should expect that the airplane will also carry some reserve
fuel/energy to ensure it can complete the specified mission in case of devi-
ations such as air traffic control (ATC) delays, weather, winds different
from forecast, and so forth. We will address fuel reserves later in this chapter.
One popular range that is often quoted is the National Business Airplane
Association (NBAA) range, used by business airplanes to normalize the
values being quoted by the marketing of various companies. The range
takes on the form shown in Fig. 10.1.
In this chapter, we will concentrate our discussion on the cruise segment
(the level flight part) of the mission and see how one can maximize the dis-
tance covered while minimizing the energy expenditure.
We will focus on two common cruise conditions. The first is the constant
altitude and constant airspeed cruise; this is the easiest condition to
monitor from an ATC and crew perspective. The second condition will be
the constant airspeed and constant lift coefficient cruise. We will see that
this condition results in the largest distance covered for a given initial fuel
condition.
We will cover jet-driven airplanes first, followed by a discussion on
propeller-driven airplanes. After the main material covering the use of fuel
(or gasoline) to cover a given distance, we will address the use of
electric energy.

Useful Parameters
To help with the cruise analysis of airplanes using fuel or gasoline as a
source of energy, we will define some parameters related to fuel weight.
The first parameter is called the fuel-weight fraction z. It is defined as the
fuel weight DWf used during a cruise segment compared to the weight of
the plane at the beginning of that same cruise segment W1 .
DWf W1  W2
z¼ ¼ (10:1)
W1 W1
 CHAPTER 10 Cruise Performance and Flight Endurance 413

For this equation, W2 represents the weight at the end of the segment.
The change in weight is therefore proportional to the fuel burn for the
equation. If other weight changes occur (e.g., military cargo plane dropping
cargo in flight), one can break the cruise segment into a prior to cargo
drop and after cargo drop scenario to compute two different
fuel-weight fractions.
Another useful parameter for cruise analysis is the mass fraction MR,
which is the mass of the airplane at the beginning of the cruise segment com-
pared to the mass at the end of the segment.

W1 W1 1
MR ¼ ¼ ¼ (10:2)
W2 W1  DWf 1z

The airplane’s fuel flow FF represents the quantity of fuel used over time;
it is a change in the mass of fuel over time.

dWf
FF ¼  ¼ Ẇ f (10:3)
dt

Note the negative sign in Eq. (10.3)—the FF is typically represented by a

positive value whereas the change of fuel weight over time is negative. We
saw in Chapter 8 that we can define a specific fuel consumption (SFC) for
the engine in use that is either proportional to the thrust T being developed
by the engine or the power P generated.

Ẇ f ¼ SFCT T (10:4)

Ẇ f ¼ SFCP P (10:5)

The usual units are pounds of fuel burned per hour per pound of
thrust generated (lbfuel /h/lbthrust ) or kilogram of fuel per hour per
Newton (kgfuel /h/N) for Eq. (10.4). For Eq. (10.5), the units are typically
pound of fuel per hour per shaft horsepower (lbfuel /h/SHP) or grams of
fuel per hour per kilowatt (gfuel /h/kW).
Note that this form of energy use does not have units of energy, but rather
mass of fuel used. It is understood that standard fuels have a given amount of
energy per unit mass (e.g., Jet A has 18,500 BTU/lbm or 43.36 MJ/kg), and
the units of mass are easily converted to a volume of fuel (U.S. gallons, Imper-
ial gallons, liters) that becomes the convention to relate the quantity of fuel
used to the energy cost of the fuel. It is the current standard to compare
the efficiency of an airplane in cruise against other planes. As we compare
this approach to that used for electric airplanes later in this chapter, we
will convert the weight of fuel into an equivalent energy.
414 Operational Aircraft Performance and Flight Test Practices

Specific Air Range (SAR) Weight

We now define how efficient an airplane

is at covering a given distance. The measure
needs to integrate the engine efficiency Cruise
Performance
(which may include a propeller) and aerody-
namic efficiency of the airplane. Consider
the cruise performance triangle shown in Drag SFC

Fig. 10.2. We saw that the drag is tied

closely to the airplane weight via the L/D Fig. 10.2 The cruise
ratio. We also saw that the fuel burn is performance triangle.
related to the drag and the engine specific
fuel consumption (SFC).
One way to define the airplane’s cruise efficiency is to use the specific air
range (SAR), which is defined as the air distance covered (dXa , the distance
covered with respect to the air mass) for a given amount of energy (we
start with fuel, dWf ) used.
dXa
SAR ¼ (10:6)
dW f
The SAR has units of distance covered divided by weight of fuel used
(n miles/lb) or (km/kg). Equation (10.6) illustrates the capability of the
plane to cover a distance in an air mass instead of a ground distance; it
does not address winds. It is a measure of the airplane/engine combination
efficiency. We will address winds later in this chapter.
We can rewrite the SAR equation in terms of measured airplane par-
ameters because air distance is equal to true airspeed (TAS) multiplied by
time t, and fuel burn is fuel flow multiplied by time. We get
TAS  t TAS
SAR ¼ ¼ (10:7)
Ẇ f  t Ẇ f
What makes this form of the SAR equation useful is that it is now some-
thing that can easily be measured in flight. In fact, if you were to browse
through most pilot reports of any given airplane, you would see that they
often report one or two cruise conditions, specifying the airspeed and fuel
flow of the plane.

Example 10.1
From a pilot report, an Airbus A400M is cruising at FL310 and at a
Mach number of 0.68 in ISA 258C conditions. The fuel flow at that speed
is 7700 lb/h total. What is the specific air range of the plane?

(Continued)
 CHAPTER 10 Cruise Performance and Flight Endurance 415

Example 10.1 (Continued)

Solution: Under the altitude and temperature conditions stated, one can
convert the Mach number into a true airspeed of
TAS ¼ 394 KTAS
Then, the SAR is simply
394 n miles=hr
SAR ¼ ¼ 0:051 n miles=lb
7700 lb=hr
Is this a sign of efficiency? One aspect that is missing from the previous
example is that the weight of the plane was not specified, nor was it required
to compute a SAR. To compare the relative efficiency of two airplanes in their
ability to carry a payload over a certain distance, one must understand the
relative weight of each one.

We can rewrite Eq. (10.7) in terms of thrust T or power P required to

maintain level flight.
TAS
SAR ¼ (10:8)
SFCT T

TAS
SAR ¼ (10:9)
SFCP P
We saw in Chapter 8 that the thrust required for level flight is equal to the
drag D, and that the drag is proportional to the airplane’s weight W divided
by the airplane’s lift-to-drag ratio E at the current flight condition.
W
T ¼D¼ (10:10)
E
Combining Eqs. (10.8) and (10.10), we get this form of the SAR equation:
TAS E
SAR ¼ (10:11)
SFCT W
We can also derive the equation for propeller-driven airplanes where the
engines are rated in power generated (including propeller efficiency hp , to
convert the shaft power to flight power) as
hp E
SAR ¼ (10:12)
SFCP W
As one can notice, both forms of the SAR equation indicate that the
specific air range is inversely proportional to the airplane weight. Generally,
416 Operational Aircraft Performance and Flight Test Practices

the higher the weight, the smaller the distance covered for a given amount of
fuel used.
Note one difference between Eqs. (10.11) and (10.12): jet-driven airplanes
have a strong cruise efficiency dependency on airspeed flown (both a direct
TAS term and an influence of TAS the lift-to-drag value E), whereas
propeller-driven airplanes tend to have a more indirect effect (influence of
TAS on E and on propeller efficiency hp , and power requirements setting
the SFCP ; see Fig. 10.12 later in this chapter).

Influence of the Airspeed on SAR

As seen in Eqs. (10.11) and (10.12), airspeed will influence the SAR
value. In order to visualize the variation of the SAR vs airspeed, it is useful
to create a basic airplane model. The model airplane, jet-driven, would
have a wing area of 3000 ft2 , a drag model (no compressibility effects)
CD ¼ 0.023 þ 0.045 CL 2 , a jet engine model with a specific fuel consump-
tion (SFCT ) of 0.70 lb/h/lb that is not airspeed or throttle position
dependent (an assumption), and for this example, the airplane weighs
250,000 lb. With this model, we plot the standard sea level performance
(see Fig. 10.3).
As expected, the airspeed for the maximum lift-to-drag condition and
the minimum drag condition align. We can also observe that, although the
SAR curve has the same general form as the L/D curve, its peak occurs at
a higher speed than that for maximum lift-to-drag conditions. This, as
can be seen in Eq. (10.11), indicates the strong influence of the (TAS E)
grouping for jet-driven airplanes, where even if the L/D starts decreasing,
the increasing speed will partially compensate and continue increasing
the SAR.
With the model that we have defined, including the assumptions of no
compressibility effects and constant SFCT , we can mathematically determine
the flight condition for maximum SAR as follows:
   
@SAR @ V
¼ ¼0 (10:13)
@V BR @V SFCT D BR
where V is the true airspeed (TAS), and the subscript BR stands for the best
range conditions. We find that Eq. (10.13) is satisfied when
@D D
¼ (10:14)
@V V
which means that the SAR is maximized when the slope of the drag curve
with respect to airspeed is equal to the ratio of drag over airspeed. Graphi-
cally, Fig. 10.4 shows this represents the point of tangency of a straight line
from the origin to the drag curve, no matter what the actual drag coefficient
equation looks like.
 CHAPTER 10 Cruise Performance and Flight Endurance 417

90,000 18

80,000 16

70,000 L/D 14

60,000 12
Drag (lb)

50,000 10

L/D
40,000 8

30,000 6
Drag
20,000 4

10,000 2

0 0
0 50 100 150 200 250 300 350 400
Airspeed (KTAS)

0.020 18

0.018 16
0.016 14
0.014
L/D 12
0.012
SAR (nm/lb)

SAR 10
0.010 L/D
8
0.008
6
0.006
4
0.004

0.002 2

0.000 0
0 50 100 150 200 250 300 350 400
Airspeed (KTAS)

Fig. 10.3 Sea-level performance of model jet airplane.

Still keeping the same assumption, we can derive the airspeed for BR for a
jet-driven airplane as being equal to
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffi
u W 4 3K
VBR ¼ u
t1 (10:15)
CDo
rSL sS
2
418 Operational Aircraft Performance and Flight Test Practices

90,000 0.02
SAR 0.018
80,000
0.016
70,000
0.014
60,000

SAR (n miles/lb)
0.012
Drag (lb)

50,000
0.01
40,000
0.008
30,000
0.006
Drag
20,000
0.004

10,000 0.002
(D/V)min
0 0
0 50 100 150 200 250 300 350 400
Airspeed (KTAS)

Fig. 10.4 Condition for best range BR.

This is essentially the same equation format as the minimum drag air-
speed equation with an additional fourth root of 3 factor; thus, the speed is
faster than the minimum drag conditions as shown.
pffiffiffi
VBR ¼ 4 3VEm (10:16)
And we find the following values as well:
rffiffiffiffi rffiffiffiffiffiffiffiffi
3 CDo 4
EBR ¼ EM CLBR ¼ CDBR ¼ CDo (10:17)
4 3K 3
Figure 10.4 shows an example of SAR curves vs weight and Mach number
for a turbofan-equipped airplane at a given altitude under standard day con-
ditions. We can see the trends of Eqs. (10.11) and (10.15) in the curves. We
also can see that as the weight of the airplane increases, the SAR reduces at all
Mach numbers, with the greatest impact at low Mach. As the weight
increases, the speed for best range (maximum SAR) increases. The curves
of Fig. 10.5 include the variation of SFC for the engines with varying throttle
setting and Mach.
Repeating the exercise for a propeller-driven airplane, where we
assume the maximum power available remains unchanged with airspeed,
can also be best done by defining a model airplane. The model airplane
will have a wing area of 400 ft2 , a drag model (no compressibility effects)
CD ¼ 0.028 þ 0.04188 CL 2 , a turboprop engine model with a specific fuel
 CHAPTER 10 Cruise Performance and Flight Endurance 419

consumption (SFCP ) of 0.60 lb/h/SHP that is not airspeed or throttle

position dependent (an assumption), and a constant-speed propeller with a
propulsive efficiency hp of 85%, also not airspeed dependent for this
example. The airplane weight here will be 19,000 lb. With this model, we
plot the standard sea-level performance (see Fig. 10.6).
It can be observed that the maximum SAR occurs at the conditions for
maximum L/D. This was expected by observation of Eq. (10.12) and the
assumptions made. Although the models for the propeller efficiency or the
specific fuel consumption may differ somewhat from our constant value
(independent of airspeed) assumptions, the L/D variation with airspeed
has a dominant effect on a propeller-driven airplane.
We can then also state that the BR conditions for a propeller-driven
airplane will be those of maximum L/D, and we can write the following:
½VBR prop ¼ VEm ½EBR prop ¼ Em (10:18)

This is true if the pilot can match the best engine conditions (optimize
SFCP ) and propeller maximum efficiency (combined propeller pitch and
advance ratio) to the airplane’s maximum L/D. The trend from another tur-
boprop airplane is shown in Fig. 10.7. The figure shows a given airplane con-
figuration (sets CDo and K) at a given weight. Then the altitude and
temperature are selected (which sets the relationship between KCAS and
KTAS). The propeller for this example is a constant-speed propeller. (The
propeller governor maintains the propeller RPM selected by the crew.)

0.320
Lo
0.300 ngr
an
ge
0.280

0.260
Specific range – nm/kg

kg

kg
00

0 g
0.240
,0

,00 00 k kg
25

27 , 0 00 g
29 31,0 00 k kg
0.220 3 ,0 00 kg
3 5,0 0
3 ,00 0 kg g
37 9,00 0 k g
0.200 3 ,00 0 k
41 ,00 0 kg
43 5,00
4
0.180

0.160

0.140

0.120

0.100
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80
Mach number – [–]

Fig. 10.5 Example of SAR curves vs weight and Mach.

420 Operational Aircraft Performance and Flight Test Practices

Then, for a given propeller RPM, as the pilot increases the torque, the true
airspeed of the airplane increases.
If we look at specific numbers from Fig. 10.7, say selecting the 160 KTAS
flight condition, we note that increasing the propeller RPM (reducing its

3600 18

3200 16

2800 L/D 14

2400 12
Drag (lb)

2000 10

L/D
Drag
1600 8

1200 6

800 4

400 2

0 0
0 50 100 150 200 250 300
Airspeed (KTAS)

0.900 18

0.800 16

0.700 14
L/D
0.600 12
SAR (n miles/lb)

0.500 10
L/D

SAR
0.400 8

0.300 6

0.200 4

0.100 2

0.000 0
0 50 100 150 200 250 300
Airspeed (KTAS)

Fig. 10.6 Sea-level performance of model propeller-driven airplane.

 CHAPTER 10 Cruise Performance and Flight Endurance 421

Given airplane weight and altitude

1900 RPM 1750 RPM 1600 RPM Propeller speed 0.170
Temp
Torque Fuel Torque Fuel Torque Fuel 1750 RPM 1900 RPM
˚C Ft-Lbs Flow KTAS Ft-Lbs Flow KTAS Flow KTAS 0.168
PPH PPH
Ft-Lbs PPH Prop RPM
1738 394 175 1847 394 173 1949 394 170
1600 366 169 1700 366 167 1800 365 164
0.166
–10 1400 327 159 1500 328 158 1600 328 155
1240 301 150 1300 296 147 1400 298 146 0.164
1290 294 147 1365 292 144

SAR (km/kWh)
Selected torque sets propeller pitch based on selected prop RPM 0.162
1.0
0.160
0.8 1600 RPM
0.158

 0.6 0.156
0.4 Blade 0.154 Engine torque
angle
0.2
15˚ 20˚ 25˚ 30˚ 35˚ 40˚ 45˚ at 0.75 R 0.152
140 145 150 155 160 165 170 175
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 KTAS
V
nD
Lower TAS Higher TAS
Higher RPM Lower RPM

Fig. 10.7 Example of adjustment of propeller RPM and engine torque affecting SAR.

advanced ratio) will result in a torque required to maintained the airspeed

reducing from about 1665 ft-lb at 1600 RPM to 1385 ft-lb at 1900 RPM.
The flight power [airplane’s drag times true airspeed, Eq. (8.33)] is constant
because we fixed the flight speed. Therefore, for this example, the airplane’s
(hp /SFCP ) is optimized by increasing the propeller RPM.

Influence of Altitude on SAR and BR

For a jet-driven airplane, observing Eqs. (10.11) and (10.15), we note that
the maximum specific air range (BR conditions) for a given weight will tend
to increase as the altitude increases (s reduces). The maximum value of the
aerodynamic efficiency E does not increase with altitude, so we will see a
trend for the SAR curves to shift to the higher speeds with increasing altitude.
Figure 10.8 was generated using the airplane model used to define the SAR
curve of Fig. 10.3. We can clearly see a benefit to flying higher and faster
on the airplane SAR. Also note that the benefit rapidly goes away at lower air-
speed, and a higher altitude actually represents a reduction in SAR capability.
Of course, as noted in Chapter 8, the SFC is not necessarily constant with
Mach or altitude; the change is typically gradual, and the general trend of the
SAR curve remains the same. So at first glance, there is an obvious benefit for
jet-driven airplanes to fly at a higher altitude and faster airspeed from a SAR
point of view.
But, as seen in Fig. 10.9, a gain in altitude will result in an increase in
Mach number. At the higher Mach number, drag rise will occur, which
will result in a rapid increase in the zero lift drag due to the presence of
shock waves. The SAR curves (Fig. 10.8) will be altered by this rapid
change in drag, and the maximum SAR may even occur at the beginning
of the drag rise (see Fig. 10.9).
422 Operational Aircraft Performance and Flight Test Practices

Altitude effects
0.03

0.025
20,000 ft

0.02
SAR (n miles/lb)

10,000 ft

Sea level
0.015

0.01

0.005

0
0 50 100 150 200 250 300 350 400 450
KTAS

Altitude effects
0.03

0.025
20,000 ft

0.02
SAR (n miles/lb)

10,000 ft

0.015 Sea level

0.01

0.005

0
0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700
Mach

Fig. 10.8 Altitude effects on jet-driven airplane SAR.

For high Mach cruise, it is often more appropriate to rewrite the SAR
equation in terms of Mach number instead of airspeed. This form of the
SAR equation indicates that, for a given altitude, SAR is maximized by max-
imizing M (L/D) (see Fig. 10.10). Therefore, to maximize range, an airplane
 CHAPTER 10 Cruise Performance and Flight Endurance 423

will often cruise in the transonic Mach range.

pffiffiffi
TAS E aSL u ½M ðL=DÞ
SAR ¼ ¼ (10:19)
SFCT W SFCT W

In the interest of increasing the productivity of an airplane (e.g., number

of flights per day or distance covered per day) or to reduce hourly costs (e.g.,
crew cost), an airline may wish to operate the airplane at a faster airspeed
than the airspeed for maximum range. The airplane’s long-range cruise
(LRC) speed is the airspeed (higher than the maximum range airspeed)
that corresponds to a loss of only 1% in SAR compared to the maximum
value (see Fig. 10.11). The airplane will therefore operate further into the
transonic range.

40,000 0.0350
Dragrise
35,000 0.0300

30,000
0.0250

SAR (n miles/lb)
25,000
Drag (lb)

0.0200
20,000
0.0150
15,000
0.0100
10,000

5000 0.0050

0 0.0000
0.6 0.65 0.7 0.75 0.8 0.85 0.9
Mach
0.400 Lo
ng
ran
ge
0.350 MMO
Specific range – (n miles/kg)

g
00 k
25,0 00 k
g
g
0.300 27,0 9,000 k 0 kg
2 0
31,0 ,000 kg g
33 5,000 k
3 0 kg kg
0.250 37,00 0
39,0041,000 kg kg
0
43,00 ,000 kg
45

0.200

0.150

0.100
0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85
Mach number −[−]

Fig. 10.9 Two examples of drag rise impact on SAR curves.

424 Operational Aircraft Performance and Flight Test Practices

15 Drag Rise 0.04

L/D
14
0.035
13

SAR (n miles/lb)
(L/D), M (L/D)

0.03
12
SAR

11 0.025
M (L/D)

Max SAR
10
0.02
9

8 0.015
0.6 0.65 0.7 0.75 0.8 0.85 0.9
Mach

Fig. 10.10 Example of maximum SAR in the drag rise.

For prop-driven airplanes, if one observes Eq. (10.12) with the baseline
assumptions of constant propeller efficiency hp and engine SFCP , one will
conclude that the SAR is not a function of altitude. Equation (10.18) does
show that the true airspeed for best range for a propeller-driven airplane
will increase with altitude. The reality is a little more complicated than

15 Drag Rise 0.034

L/D
14 0.032

13 0.03
SAR –1%
SAR (n miles/lb)
(L/D), M (L/D)

12 0.028

11 0.026
M (L/D)

10 0.024
Max SAR

LRC

9 0.022

8 0.02
0.7 0.72 0.74 0.76 0.78 0.8 0.82 0.84
Mach

Fig. 10.11 Long-range cruise (LRC) Mach.

 CHAPTER 10 Cruise Performance and Flight Endurance 425

1.0

0.9
Specific air range ~ n miles/lb
0.8

ran T
30,000 129 KCAS

~F
ce

Long range cruise

5,000

um de
2
0.7

du
xim itu
en
174 KCAS
Ma e alt
0
20,00
0.6 ur
ss
0
15,00
Pre

0.5 ise 211 KCAS

0 cru
10,00 um
0.4 xim
5000 Ma
18 20 22 24 26 28303234 36.9
0.3 0 14 16
12 Torque ~ PSI
0.2
80 100 120 140 160 180 200 220 240 260 280 300
True airspeed ~ KT

Fig. 10.12 Example of SAR curves for a turboprop airplane at constant weight.

this, because both the engine and the propeller must turn at their optimal
RPM to generate the maximum combined efficiency [ratio (hp /SFCP )],
and then this must be aligned with the airplane aerodynamic maximum effi-
ciency Em for peak SAR. Figure 10.12 provides an example of a turboprop
airplane in which increasing altitude does help improve range. Looking at
the LRC condition, it shows how the pilot trades L/D (changing KCAS
with altitude) with propeller efficiency and engine SFC (reducing torque
with increasing altitude). Note how the BR true airspeed does not follow
the trend of the simple assumptions (constant hp and SFC) that generated
Eq. (10.18) for this plane. However, the SAR is as defined by Eq. (10.12),
and maximum SAR is still defined by the optimum combination of its
parameters.
An Internet search for the pilot’s report of various airplanes will provide
examples of current design capabilities. Here are a few examples for jet- and
propeller-driven airplanes. (For conversion to energy, we used an energy
density of 11.98 kWh/kg for Jet A fuel.)

• C27J airplane: Once level at 8000 ft, the autopilot precisely maintained
200 kt indicated air speed. Total fuel flow was 966 kg/h, and the 22,160-kg
aircraft maintained 228 kt true airspeed with a static air temperature
of 148C (578F/ISA þ158C). SAR ¼ 0.236 n miles/kg ¼ 0.107 n miles/lb
¼ 0.0363 km/kWh (https://www.flightglobal.com/
flight-test-c-27j-no-small-measure/56418.article).
• SJ30-2 airplane: With the air cooler than standard, and some towering
cumulus ahead, the pilot zipped up to FL450 in 5 min, where the airplane
settled in at Mach 0.80, showing 446 kt through the air on a fuel flow of just
426 Operational Aircraft Performance and Flight Test Practices

770 pph. SAR ¼ 0.579 n miles/lb ¼ 0.197 km/kWh (https://www.

flyingmag.com/pilot-reports/jets/sj30-2-trading-size-speed/).
• CSeries airplane: The CSeries smoothly leveled off at FL280 just 11 min
after brake release. Once level, Mach 0.74 was set in the FMS to establish a
long-range cruise condition. Although well below optimum altitude, total
fuel flow was 4600 lb/h, with a true airspeed of 447 kt (830 km/h).
SAR ¼ 0.097 n miles/lb ¼ 0.033 km/kWh (https://www.flightglobal.com/
engines/flight-test-we-put-bombardiers-cseries-through-its-paces/
122236.article).
• King Air 350i: With full fuel tanks, the 350i has an available payload of
1489 lb; the IFR range decreases to only 1489 n miles. To approximate a
maximum range condition, the pilot retarded the throttles and set a
total fuel flow of 390 lb/h. The aircraft slowed to an indicated airspeed
of 136 kt with a test day true airspeed of 237 kt. SAR ¼ 0.608 n miles/lb
¼ 0.207 km/kWh (https://www.flightglobal.com/
nbaa-turboprop-royalty-beechcraft-350i-king-air/96258.article).
• Bombardier Challenger 350: The pilot used the speed brakes to rapidly
slow to a long-range cruise target of Mach 0.78. Speed brake extension
caused a minor pitch up, which was easily countered with forward yoke
pressure. At the more economical condition, total fuel flow was 1540 lb/h
at 219 KIAS. On the colder than standard day, resultant true airspeed was
443 kt. SAR ¼ 0.288 n miles/lb ¼ 0.098 km/kWh (https://www.
flightglobal.com/in-depth/flight-test-bombardier-challenger-350-/
128062.article).
• Gulfstream 650ER: Once level at 49,000 ft, the aircraft accelerated at climb
power (1.74 EPR) until it reached Mach 0.85, long-range cruise speed. A
total fuel flow of 2300 lb/h held an indicated airspeed of 211 kt, and at a
static air temperature of –66̊C (–87̊F) gave a true airspeed of 478 kt.
SAR ¼ 0.208 n miles/lb ¼ 0.071 km/kWh (https://www.flightglobal.
com/flight-test-gulfstream-goes-farther-and-faster-with-g650er/116063.
article).
• Airbus A380: Indicated airspeed was 305 kt, with a corresponding true
airspeed of 472 kt. Total fuel flow was approximately 13,000 kg/h.
SAR ¼ 0.036 n miles/kg ¼ 0.016 n miles/lb ¼ 0.0056 km/kWh (https://
www.flightglobal.com/flight-test-airbus-a380/69725.article).

Range Equation: Turbojet/Turbofans

The range (or distance or integrated range, X ) covered in the cruise
phase, by using a given load of fuel, can be determined by integrating the
SAR equation.

ð ð 
@X
X ¼ SAR dW ¼  dW (10:20)
@W
 CHAPTER 10 Cruise Performance and Flight Endurance 427

For jet-powered airplanes, the range equation takes on the form

ð
TAS E
X¼ dW (10:21)
SFCT W
The most common cruise condition, and the easiest to monitor from a
pilot’s as well as from an air traffic controller’s perspective, is the constant
altitude and airspeed cruise (V, H ); this is a mandatory cruise mode, and
ATC clearance is required to change altitude (e.g., 14 CFR §91.179).
Another condition of interest is the constant airspeed and lift coefficient
cruise (V, CL ), which as we will see later in this chapter, result in a greater
distance traveled for a given initial condition and fuel burn. A fixed CL and
V result in a fixed E (aerodynamic efficiency, L/D).
Some assumptions will now be made on the variation of the various par-
ameters in the previous equations. Because airplanes (jetliners in particular)
are designed to cruise in a very limited part of their total flight envelope
(altitude/Mach combination), one can easily assume that the SFCT should
be essentially constant during the cruise. On very long-range segments
(large weight change resulting in larger thrust changes, possible significant
temperature changes), one could break down the cruise into smaller
segment and readjust SFCT to be more appropriate.
Therefore, for the two flight conditions of interest, the range equation
reduces to
ð
TAS E
XV , H ¼  dW (10:22)
SFCT W
ð
TAS E dW
XV ,CL ¼  (10:23)
SFCT W
We first look at the (V, H) condition. The details of the integration can be
found in Annex D of [1]. The resulting cruise range equation has the follow-
ing form:
" #
2 Em TAS z E1
XV ,H ¼ tan1   (10:24)
SFCT 2 Em 1  K CL1 E1 z

where E1 and CL1 represent the L/D and lift coefficient at the initial cruise
conditions. To maximize the range for a constant TAS and H cruise, one
must have the following:
• A high maximum lift-to-drag ratio (Em : airplane design)
• A high true airspeed (TAS: drives the design of the airplane)
• Engines with low specific fuel consumption (SFCT : engine design)
• A high fuel-weight fraction (z)
To get a feel for the impact of some major parameters on the range, we
define a model airplane that will start the cruise at 33,000 ft, Mach 0.72 in
428 Operational Aircraft Performance and Flight Test Practices

Table 10.1 Range Example

Condition 1 Condition 2 Condition 3 Condition 4

W(lb) 3,00,000 2,50,000 3,00,000 3,00,000
CDo 0.02 0.02 0.02 0.02
K 0.0531 0.0531 0.0419 0.0531
S (ft2 ) 3200 3200 3200 3200
Em 15.35 15.35 17.28 15.25
SFCt (hr21 ) 0.65 0.65 0.65 0.6
E1 14.83 13.95 16.09 14.83
CL1 0.47 0.39 0.47 0.47

a standard atmosphere. Table 10.1, condition 1 represents the reference

condition. For condition 2, the starting weight was reduced. For condition
3, the wing-induced drag parameter K was reduced (aspect ratio increase)
compared to condition 1. Finally, the SFCT was reduced for condition 4
compared to condition 1. Figure 10.13 presents the range capability of
each case with respect to the fuel-weight fraction z and the average SAR
(here simply the range covered divided by the fuel weight used).
Starting with reference condition 1, we note that as the quantity of fuel
burned increases (z), the range increases as expected. But we also note that
the rate of range increase is not linear; it becomes steeper (see Fig. 10.13a).
This is expected because the weight of the airplane reduces proportionally
to the amount of fuel burn, and the SAR progressively improves for the
same V, H condition (see Fig. 10.14, where z represents the relative
amount of fuel burned from the start of the cruise where z ¼ 0.0). This
benefit of weight reduction vs distance covered is also shown in Fig. 10.13b
as an equivalent average SAR for the flight (given z). A lighter airplane is
more efficient at covering a given range.
Now we look at a different starting weight (condition 2) for the same
airplane as condition 1 and at the distance covered (same flight condition)
versus the fuel mass ratio z, and we note that the distance would be smaller.
This is simply because the initial weight is smaller, and therefore the
amount of fuel burned for the same z would be smaller. At equal weight of
fuel burned (say z ¼ 0.15 for condition 1 ¼ 45,000 lb, equivalent condition
2 z ¼ 0.18), the lighter weight initial condition would cover more range
(1701 n miles for condition 2 vs 1514 n miles for condition 1). This benefit
is also seen on the average SAR for all fuel weight fractions (Fig. 10.13b).
Returning to the airplane of condition 1, but with a better wing (condition
3, aspect ratio increased from 7.5 to 9.5, lower K by 21%), we note an increase
in range for a given amount of fuel used and an improvement in the average
SAR. We also note the benefit is also not proportional to the percent
reduction in induced drag parameter—here, the average improvement in
 CHAPTER 10 Cruise Performance and Flight Endurance 429

SAR was only 7.5%. This difference comes from the fact that the drag due to
lift is relatively low (low percentage of total drag, starting CL relatively low)
for jet-driven cruise conditions (faster than minimum drag).
Finally, we look at an SFCT improvement (e.g., new engine). Condition 4
represents a 7.7% reduction in the SFCT compared to condition 1. For the

(a) 4000

3500 Condition 4
Condition 3
Condition 1
3000
Condition 2
Range (n miles)

2500

2000

1500

1000

500

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4


(b) 0.04
Condition 2
0.039
Condition 4
0.038 Condition 3
Average SAR (n miles/lb)

0.037

0.036
Condition 1
0.035

0.034

0.033

0.032

0.031

0.03
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4


Fig. 10.13 Example of constant speed and altitude cruise.

430 Operational Aircraft Performance and Flight Test Practices

0.0380
M0.72
33,000 ft
0.0370   0.2

0.0360
SAR (n miles/lb)

0.0350
  0.1
0.0340

0.0330

0.0320   0.0

0.0310

0.0300
395 400 405 410 415 420 425 430 435 440 445
KTAS

Fig. 10.14 Increase in SAR as fuel is burned (airplane weight reduces).

same initial flight condition as condition 1, the average improvement in

SAR will be about 8.3%. We therefore note how powerful the improvement
of the engine efficiency is on the overall airplane range improvement
(approximately one for 1% improvement).
For the next flight condition, the specific air range equation can be
integrated by assuming a constant airspeed and constant lift coefficient con-
dition. In this case, the lift-to-drag ratio E is constant and moves out of the
integral. (This assumes no Mach effect during the progression of the
cruise.) The range becomes
ð  
TAS E dW TAS E W1
XV ,CL ¼  ¼ ln (10:25)
SFCT W SFCT W2
This equation is also referred to as the Breguet range equation. The range
becomes proportional to the weight fraction and the ratio of true airspeed
times the lift-to-drag ratio divided by the specific fuel consumption. Equation
(10.25) can be rewritten as
 
TAS E TAS E 1
XV ,CL ¼ lnðMRÞ ¼ ln (10:26)
SFCT SFCT 1z
Of course, flying at a constant lift coefficient is not a natural flight mode,
because the pilot has no instrument to indicate what the lift coefficient is. On
experimental airplane, where special computers are installed, one can easily
 CHAPTER 10 Cruise Performance and Flight Endurance 431

compute the lift coefficient and display its value to the pilot. As an alternative
to the lift coefficient, one could fly at a constant W/d. From the following
equation, one can notice that a constant W/d ratio is essentially equal to a
constant lift coefficient:
 
1
nðW =dÞ ¼ g p S M2 CL (10:27)
2 SL
The only parameters that need to be measured are the Mach number, the
pressure ratio (d, simply the ratio of the pressure at the current altitude over
the reference sea level pressure), the aircraft weight (simply need the initial
weight and the fuel burn to compute), and to be more exact, the load
factor (n, 1 g in normal level flight).
In order to maintain a constant lift coefficient and a constant true air-
speed while the aircraft weight is decreasing (fuel consumed), the aircraft
will have to climb! This flight profile is also called cruise-climb. From the
lift equation,
1
W ¼ L ¼ rSL sV 2 S CL (10:28)
2
As the airplane weight goes down, so must the air density ratio. A
decrease in air density ratio corresponds to an increase in altitude. This
can be seen by the ratio of the weights and using Eq. (10.28) to get
s2 W2
¼ (10:29)
s1 W1
An obvious question from the previous statement: Does this change in
altitude contradict our assumption of level flight used to derive the
Breguet equation? Analysis of the sigma ratio over a cruise with typical
fuel burn will show that the resulting climb angle is very small (often of
the order of 0.025 deg), and it can therefore be neglected.
The next question is: What is the advantage of following the more com-
plicated cruise-climb profile over the constant airspeed and altitude profile?
To answer this, we use the airplane defined by condition 1 of Table 10.1—
ability as a function of fuel consumed for a given set of initial conditions—
and we produce charts comparing the two cruise profiles (see Fig. 10.15).
Notice that the cruise-climb profile results in the longer cruise distance of
the two cruise profiles. The benefit is also shown on the average SAR.
Equation (10.24) does not necessarily give the full picture for the
analysis. It is true that the airplane will fly at a constant true airspeed and lift
coefficient, but this could result in an increase in Mach number and possibly
compressibility effects. Using the data from condition 1, we create Fig. 10.16.
For this example, the airplane will climb from 33,000 ft to a little above
41,000 ft by the time the fuel fraction reaches 30%. This altitude is often
a limiting condition for jetliners, so is the airplane certified to fly to this
432 Operational Aircraft Performance and Flight Test Practices

4000
Condition 1
3500
V, CL
3000
V, H
Range (n miles)

2500

2000

1500

1000

500

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35


0.039
Condition 1
0.038
V, CL
0.037
Average SAR (n miles/lb)

0.036

0.035
V, H
0.034

0.033

0.032

0.031
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35


Fig. 10.15 Comparing two cruise profiles.

altitude? As well, the Mach number will increase as the altitude increases up to
the tropopause. Are there any compressibility effects that could impact the L/D?
There is a benefit to using a cruise-climb condition, but one needs to
clearly review the entire flight profile to see if it can be executed. We also
note that as the fuel fraction increases (as shown in Fig. 10.15), the spread
 CHAPTER 10 Cruise Performance and Flight Endurance 433

in distance covered for a given fraction between the two flight profiles
also increases.

Cruise Altitude Restrictions

Constant airspeed and lift coefficient cruise (the cruise-climb condition)
yields the best range for a given starting altitude and airspeed. Unfortunately,
air regulations restrict flights to constant altitude in order to control separ-
ation between aircraft, and altitude change must be authorized by air
traffic control (ATC). For example, the FAA prescribes the following restric-
tions on flight altitudes:
FAA §91.159 VFR cruising altitude or flight level. Except while holding in
a holding pattern of 2 minutes or less, or while turning, each person operat-
ing an aircraft under VFR in level cruising flight more than 3,000 feet above
the surface shall maintain the appropriate altitude or flight level prescribed
below, unless otherwise authorized by ATC:
(a) When operating below 18,000 feet MSL and—
(1) On a magnetic course of zero degrees through 179 degrees, any odd
thousand foot MSL altitude þ 500 feet (such as 3,500, 5,500, or 7,500); or
(2) On a magnetic course of 180 degrees through 359 degrees, any even
thousand foot MSL altitude þ 500 feet (such as 4,500, 6,500, or 8,500).
(b) When operating above 18,000 feet MSL, maintain the altitude or flight
level assigned by ATC.

42,000 0.730

41,000 0.728

40,000 0.726
Mach
39,000 0.724

38,000 0.722
Altitude (ft)

Mach

37,000 0.720
Altitude
36,000 0.718

35,000 0.716
Condition 1 cruise-climb
34,000 0.714

33,000 0.712

32,000 0.710
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
␨

Fig. 10.16 Altitude change and Mach number evolution in cruise-climb.

434 Operational Aircraft Performance and Flight Test Practices

FAA §91.179 IFR cruising altitude or flight level. Unless otherwise

authorized by ATC, the following rules apply—

(a) In controlled airspace. Each person operating an aircraft under IFR in level
cruising flight in controlled airspace shall maintain the altitude or flight level
assigned that aircraft by ATC. However, if the ATC clearance assigns “VFR
conditions on-top,” that person shall maintain an altitude or flight level as
prescribed by §91.159.
(b) In uncontrolled airspace. Except while in a holding pattern of 2 minutes or less or
while turning, each person operating an aircraft under IFR in level cruising
flight in uncontrolled airspace shall maintain an appropriate altitude as follows:
(1) When operating below 18,000 feet MSL and—
(i) On a magnetic course of zero degrees through 179 degrees, any odd
thousand foot MSL altitude (such as 3,000, 5,000, or 7,000); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any even
thousand foot MSL altitude (such as 2,000, 4,000, or 6,000).
(2) When operating at or above 18,000 feet MSL but below flight level 290, and—
(i) On a magnetic course of zero degrees through 179 degrees, any odd
flight level (such as 190, 210, or 230); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any even flight
level (such as 180, 200, or 220).
(3) When operating at flight level 290 and above in non-RVSM airspace, and—
(i) On a magnetic course of zero degrees through 179 degrees, any flight
level, at 4,000-foot intervals, beginning at and including flight level
290 (such as flight level 290, 330, or 370); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any flight
level, at 4,000-foot intervals, beginning at and including flight level 310
(such as flight level 310, 350, or 390).
(4) When operating at flight level 290 and above in airspace designated as Reduced
Vertical Separation Minimum (RVSM) airspace and—
(i) On a magnetic course of zero degrees through 179 degrees, any odd flight
level, at 2,000-foot intervals beginning at and including flight level 290
(such as flight level 290, 310, 330, 350, 370, 390, 410); or
(ii) On a magnetic course of 180 degrees through 359 degrees, any even
flight level, at 2000-foot intervals beginning at and including flight level
300 (such as 300, 320, 340, 360, 380, 400).

Stepped Climb
With the previously described cruise altitude restrictions, what can be done
to respect air traffic control regulations while trying to achieve conditions
similar to cruise-climb and maximize range? Upon inspection of Figs. 10.15
and 10.17, and as stated previously, one can notice that as the value of z
decreases, the difference in range between the two types of cruise decreases.
Note that the ratio presented in the graph is an example; the exact value
depends on the initial flight conditions and airplane characteristics. The
initial cruise altitude (point 1, Fig. 10.18) is usually limited by the airplane’s
climb capability or inability to reach the next authorized flight level with a
reasonable rate of climb (typically at least 300 fpm). Once the cruise starts,
the airplane is not allowed to follow the cruise-climb profile.
After having cruised for a while at constant V and H, the airplane weight
will be lower (fuel burned) and its climb capability can allow it to climb to the
 CHAPTER 10 Cruise Performance and Flight Endurance 435

0.99

0.98
Relative range

0.97

0.96

0.95

0.94

0.93
0 0.05 0.1 0.15 0.2 0.25 0.3


Fig. 10.17 Relative range vs fuel-weight fraction [(V, H )/(V, CL )].

next authorized level (with ATC permission). “Stepping” to this next altitude
will improve the SAR of the airplane; this is the stepped climb cruise profile.
If these steps are reasonably small in terms of fuel burn (i.e., z), then the
distance covered for a given fuel burn using the stepped climb profile will be
significantly closer to the cruise-climb profile than following a constant
altitude and airspeed profile from the initial condition. The limitation on
how small the steps must be will depend on the authorized flight levels
(cruise altitude restrictions), on the ability for the airplane to climb to the
next level (climb performance, discussed in Chapter 12), and of course on
getting ATC authorization.

Optimum Cruise Altitude

Most jetliners will be limited by climb thrust available when selecting a
desired cruise altitude. These airplanes will thus simply climb as high as

Cruise-climb
2

1 Stepped climb

Fig. 10.18 Stepped climb cruise.

436 Operational Aircraft Performance and Flight Test Practices

800 17
TAS
750 16.5

700 16
Mach limited
650 15.5
TAS (ft/s)

L/D
600 15
L/D

Longest range
550 14.5

500 14

450 13.5

400 13
0 10,000 20,000 30,000 40,000
H (ft)

0.13

0.12

0.11
SAR (n miles/lb)

0.1
Longest range

0.09

0.08

0.07

0.06
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000
H (ft)

Fig. 10.19 Selection of the optimum cruise altitude.

possible to achieve the highest SAR for the mission. But airplanes with signifi-
cantly more available thrust (e.g., fighters) could, in theory, keep climbing to
much higher altitudes and still improve SAR.
We have previously discussed that compressibility effects will be encoun-
tered at higher altitude when VBR reaches the drag rise Mach. Any further
increase in altitude will see the airplane being limited to a given Mach
 CHAPTER 10 Cruise Performance and Flight Endurance 437

number to maximize SAR. Because the airplane was flying faster than the
conditions for maximum L/D (Em ), any further increase in altitude while
maintaining a fixed Mach number will initially result in an increase in L/D
(airspeed slower than the best range condition but faster than minimum
drag condition), so one is effectively maximizing the [M (L/D)] ratio (see
Fig. 10.19). The altitude at which, at the Mach-limited airspeed, the
product [M (L/D)] equals [M Em ] represents the longest range condition.
Any further increase in altitude will not result in an increase in
maximum SAR.
As a reminder, this approach does not consider the climb capability of the
aircraft; it is purely a review of the airplane’s cruise condition vs altitude for
longest range.

Range Equation: Reciprocating/Turboprop

For propeller-driven airplanes, we integrate the SAR [Eq. (10.12)] as a
function of fuel weight burned to get the range [Eq. (10.30)]. The simplest
approach to solve this equation is to assume the airplane will fly at constant
lift coefficient (constant E), that the engine SFCP is constant, and that within
these conditions, the propeller efficiency is unchanged. If those hold true, we
get Eq. (10.31).
ð
hp E
X¼ dW (10:30)
SFCP W
 
hp E W1
X¼ ln (10:31)
SFCP W2

As per the SAR equation, the range equation [Eq. (10.31)] does not show
distinct speed or altitude dependency. The pilot must optimize the combi-
nation of efficiency of three different systems (airframe, engine, and propel-
ler) to achieve maximum range. In theory, purely looking at Eq. (10.31), there
is no range benefit of going to a higher altitude. But as the aerodynamic effi-
ciency is optimized at minimum drag (low calibrated airspeed) and engine
efficiency is typically optimized at high power (high true airspeed), there
can be an advantage of altitude increase on overall range. We offer the
examples shown in Fig. 10.20, which were found in a pilot operating
handbook online.
The best application of Eq. (10.31) this author has seen comes from a
keynote address on the design approach for the Voyager (see Fig. 10.21),
the first airplane to perform an unrefueled flight around the world in Dec.
1986. The 26,366-mile trip took 9 days, 3 min, and 44 s. We extracted a
few key points from [2] to highlight the design process and use of the
range equation to achieve this great feat.
438 Operational Aircraft Performance and Flight Test Practices

W = 8750 lb
Prop 1900 RPM
Fuel = 2224 lb
45 minutes reserve
Zero wind, standard day
25,000

155
KTAS
20,000

167
KTAS
Altitude - Feet 15,000
Maximum 152
cruise power KTAS
172 Maximum
KTAS range power
10,000
150
KTAS
173
KTAS
5000
148
166 KTAS
KTAS
149
KTAS
SL
600 700 800 900 1000 1100 1200 1300 1400
Range - nautical miles

W = 5000 lb
Fuel = 106 gallons
Note: Range includes start, taxi, climb and descent with 45 minutes reserve fuel at economy cruise.
20,000
FT = full throttle
Prop 2100, 2300, 2500 RPM
0
0
Pressure altitude ~ feet

50
30

15,000
/2
/2

10
FT
FT

/2
FT

192 184 171 Cruise TAS~knots

10,000
Power setting

0
IN. HG./RPM

197 19 186 171

300
24.5/2500

24.0/2300

5000
/210

199 188 161

21.0/2

175
20.5

188 178 164 151

SL
400 450 500 550 600 650 700 750 800 850
Range ~ nautical miles (zero wind)

Fig. 10.20 Examples of range capability vs airspeed, engine power setting, and
propeller RPM.

With a singular goal in mind of global non-refueled flight, I initially planned

to design the Voyager for a range well in excess of that required. . .. This
would be a very special aircraft, with a specific design point-optimizing
range with no other generic compromises.

• Ability to just barely takeoff from the world’s longest airport, the 3-mile
runway at nearby Edwards Air Force Bases, California.
• A total weight of 350 pounds for two crews, food and water for the
entire trip.
 CHAPTER 10 Cruise Performance and Flight Endurance 439

• Two engines to allow staging power during the flight. The cruise power
during the last stages of the flight with nearly all the fuel gone would be
less than 15% of that needed for the heavy weight takeoff. Engines are not
efficient at these low power settings, so by shutting off the larger of the two
engines several days into the flight, the important last day would have the
remaining engine running with at least 35% rated power for an
acceptable efficiency.
• The engines and propellers would be “off-the-shelf” with a long history
of reliability.
• The structure would have to be extremely light.

For the airplane efficiency, I would use a slender wing with extremely long
span to distribute the weight lightly and evenly to achieve minimum aerody-
namic drag due to lift. The result would be an airplane efficiency (L/D) of
between 32 and 40.
The last term of the Breguet range equation is where the Voyager was to be
significantly better than any airplane previously built. My goal was to achieve
a fuel weight of 80% of the takeoff weight. . .. The structure would weigh only
9% of the takeoff weight of the airplane.
The graph below [Fig. 10.22] summarizes the measured overall efficiency
of the Voyager aircraft. These flight test data overcome the deficiencies of the
Breguet equation, in that the “constants” which in reality are variables, are all
measured together for a true range prediction.
At the world flight takeoff, at a gross weight of 9,700 pounds, the Voyager
carried 73% of its weight in fuel. . .. A nine day flight, fraught with every
conceivable contingency of weather, turbulence, tailwinds and head winds
tends to be a great average. The performance of the Voyager actually
turned out to be within one percent of its prediction. The additional aero-
dynamic drag, which was experienced because the wing tips were damaged
scraping on the runway during takeoff and the loss of approximately 100
pounds of fuel through a leaky fuel cap, reduced the range to a point
where only 1.5% of the takeoff fuel remained (18 gallons) when the
aircraft landed.

Fig. 10.21 Rutan Voyager. Source: NASA.

440 Operational Aircraft Performance and Flight Test Practices

8
Aft engine operating
Miles per pound of fuel

Both engines operating

4

Takeoff weight
2
Landing weight

0
0 2000 4000 6000 8000 10,000
Gross weight − pounds

Fig. 10.22 Rutan’s Voyager flight test measured efficiency [2].

Fuel Reserves
The Voyager landed with little fuel left in its tanks after completing the
around-the-world flight. In practice, the operator should not plan on using
all the fuel on-board the airplane to complete a mission. The flight should
be planned with the expectation of landing with a minimum fuel quantity,
the fuel reserves. These fuel reserves are defined as the ability of the airplane
to stay aloft for a certain amount of time (30 to 45 min) and a distance (fly to
an alternate) longer than what is planned to be used to complete the intended
mission. The International Civil Aviation Organization (ICAO) defines rec-
ommended values in Annex 6, 4.3.6, Fuel and Oil Supply. These recommen-
dations are used by national authorities to define the minimum operational
standards within a given country. The FAA provides recommendations for
commercial operators under 14 CFR Part 121, for general aviation under
14 CFR Part 91, and for on-demand operators under 14 CFR Part 135.
Here are a few of the requirements for fuel reserves planning in the United
States:
FAA § 121.639 Fuel supply: All domestic operations
No person may dispatch or take off an airplane unless it has enough fuel—
(a) To fly to the airport to which it is dispatched;
(b) Thereafter, to fly to and land at the most distant alternate airport (where required)
for the airport to which dispatched; and
(c) Thereafter, to fly for 45 minutes at normal cruising fuel consumption or, for
certificate holders who are authorized to conduct day VFR operations in their
operations specifications and who are operating nontransport category airplanes
type certificated after December 31, 1964, to fly for 30 minutes at normal cruising
fuel consumption for day VFR operations.
 CHAPTER 10 Cruise Performance and Flight Endurance 441

§ 121.645 Fuel supply: Turbine-engine powered airplanes, other than

turbo propeller: Flag and supplemental operations
(a) Any flag operation within the 48 contiguous United States and the District of
Columbia may use the fuel requirements of §121.639.
(b) For any certificate holder conducting flag or supplemental operations outside the
48 contiguous United States and the District of Columbia, unless authorized by the
Administrator in the operations specifications, no person may release for flight or
takeoff a turbine-engine powered airplane (other than a turbo-propeller powered
airplane) unless, considering wind and other weather conditions expected, it has
enough fuel—
(1) To fly to and land at the airport to which it is released;
(2) After that, to fly for a period of 10 percent of the total time required to fly from
the airport of departure to, and land at, the airport to which it was released;
(3) After that, to fly to and land at the most distant alternate airport specified in
the flight release, if an alternate is required; and
(4) After that, to fly for 30 minutes at holding speed at 1,500 feet above the
alternate airport (or the destination airport if no alternate is required) under
standard temperature conditions.

§ 121.643 Fuel supply: Nonturbine and turbo-propeller-powered air-

planes: Supplemental operations
(a) Except as provided in paragraph (b) of this section, no person may release for flight
or takeoff a nonturbine or turbo-propeller-powered airplane unless, considering
the wind and other weather conditions expected, it has enough fuel—
(1) To fly to and land at the airport to which it is released;
(2) Thereafter, to fly to and land at the most distant alternate airport specified in
the flight release; and
(3) Thereafter, to fly for 45 minutes at normal cruising fuel consumption or, for
certificate holders who are authorized to conduct day VFR operations in their
operations specifications and who are operating nontransport category
airplanes type certificated after December 31, 1964, to fly for 30 minutes at
normal cruising fuel consumption for day VFR operations.
(b) If the airplane is released for any flight other than from one point in the contiguous
United States to another point in the contiguous United States, it must carry
enough fuel to meet the requirements of paragraphs (a) (1) and (2) of this section
and thereafter fly for 30 minutes plus 15 percent of the total time required to fly at
normal cruising fuel consumption to the airports specified in paragraphs (a) (1)
and (2) of this section, or to fly for 90 minutes at normal cruising fuel consumption,
whichever is less.

§91.151 Fuel requirements for flight in VFR conditions

(a) No person may begin a flight in an airplane under VFR conditions unless
(considering wind and forecast weather conditions) there is enough fuel to fly to
the first point of intended landing and, assuming normal cruising speed—
(1) During the day, to fly after that for at least 30 minutes; or
(2) At night, to fly after that for at least 45 minutes.

§91.167 Fuel requirements for flight in IFR conditions

(a) No person may operate a civil aircraft in IFR conditions unless it carries enough
fuel (considering weather reports and forecasts and weather conditions) to—
442 Operational Aircraft Performance and Flight Test Practices

(1) Complete the flight to the first airport of intended landing;

(2) Except as provided in paragraph (b) of this section, fly from that airport to the
alternate airport; and
(3) Fly after that for 45 minutes at normal cruising speed or, for helicopters, fly
after that for 30 minutes at normal cruising speed.

Similar requirements are defined by EASA for European operators.

An example is:

NCC.OP.130 Fuel and oil supply—aeroplanes

(a) The pilot-in-command shall only commence a flight if the aeroplane carries
sufficient fuel and oil for the following:
(1) for visual flight rules (VFR) flights:
(i) by day, to fly to the aerodrome of intended landing and thereafter to fly for
at least 30 minutes at normal cruising altitude; or
(ii) by night, to fly to the aerodrome of intended landing and thereafter to fly
for at least 45 minutes at normal cruising altitude;
(2) for IFR flights:
(i) when no destination alternate is required, to fly to the aerodrome of
intended landing, and thereafter to fly for at least 45 minutes at normal
cruising altitude; or
(ii) when a destination alternate is required, to fly to the aerodrome of
intended landing, to an alternate aerodrome and thereafter to fly for at
least 45 minutes at normal cruising altitude.
(b) In computing the fuel required including to provide for contingency, the following
shall be taken into consideration:
(1) forecast meteorological conditions;
(2) anticipated ATC routings and traffic delays;
(3) procedures for loss of pressurisation or failure of one engine while en-route,
where applicable; and
(4) any other condition that may delay the landing of the aeroplane or increase fuel
and/or oil consumption.
(c) Nothing shall preclude amendment of a flight plan in-flight, in order to re-plan the
flight to another destination, provided that all requirements can be complied with
from the point where the flight is re-planned.

The reserve fuel is planned to ensure the airplane can safely execute the
mission. It does not mean that it has to land with an exact amount of
reserve fuel. (There can be more or less fuel than planned upon landing.)
The operators must be fully aware of the meaning of the reserve fuel on
board the airplane and what can be accomplished with it. We offer three
examples of fuel mismanagement in the incident boxes that follow.

Flight 980, a DC-9-33F operated by Overseas National Airways, departed JFK

International on 2 May 1970 en route to St. Maarten in the Antilles. The alter-
nate airport was St. Thomas, 100 n miles west of St. Maarten. The flight was
scheduled to occur at FL290, and the calculated reserve for this flight was 6400
lb, enough to fly to the alternate plus 30 min holding. The plane carried 28,900
lb of fuel, 900 lb more than the scheduled trip plus the calculated reserve. Bad
 CHAPTER 10 Cruise Performance and Flight Endurance 443

weather en route forced the aircraft to fly at a lower altitude than planned and
to divert around thunderstorms, increasing the flight distance. The aircraft
arrived near its destination with less than 4500 lb of fuel.
The weather at St. Maarten was near minimums, and after three missed
approaches, the crew decided to fly to their alternate. The aircraft suffered
fuel starvation and had to be ditched in the Atlantic Ocean approximately
halfway between its planned destination and its alternate.
A probable cause of the accident was attributed by the NTSB as pilot
improper in-flight decision/fuel management (Report DCA70AZ007).

23 July 1983—Fuel starvation incident, Air Canada Boeing 767: The fuel
computer was not functional. The alternative means to verify fuel quantity
was with a “dripping method” (a little like using a dipstick to measure a
fluid height), which gives fuel volume in liters. The refueler used a conversion
factor of 1.77 lb/l to compute fuel on board. He should have used 0.8 kg/l. (It
was a brand new airplane for Air Canada, which now used kilograms for fuel
rather than pounds.) So, if a Boeing 767 runs out of fuel at 41,000 ft, what do
you have? Answer: A 132-ton glider with a sink rate of over 2000 ft/min and
marginally enough hydraulic pressure to control the ailerons, elevator, and
rudder. The pilots managed a gliding landing to the decommissioned Gimli
airfield (the Gimli glider), where car racing was occurring.

Gimli
Winnipeg
Montreal
Ottawa

On Saturday, 19 Feb. 2005, a British Airways flight took off from Los Angeles’
LAX airport, destined for Heathrow, with 351 passengers and crew aboard.
Shortly after takeoff, with the aircraft not more than 100 ft over the ground,
controllers notified the pilot that a shower of sparks could be seen coming
out of one of the engines. The pilot responded by throttling back, but the
engine continued to overheat, and the crew decided it had to be shut down.
After circling the Pacific for a few minutes while the captain contacted
BA’s control center, the crew decided to continue the 11-hr, 5000-mile
444 Operational Aircraft Performance and Flight Test Practices

flight to Heathrow on three engines, rather than turn back and face a
minimum 5-hr delay, at an estimated cost of nearly $200,000. Just three
days before, a new EU regulation had come into force that would have
required British Airways to compensate the passengers for long delays or
cancellations.
The crew knew that the aircraft would burn more fuel because it would
be unable to climb to FL360, its assigned altitude. Instead, it was forced stay
down at FL290, and with extra rudder drag due to the differential thrust
created by the engine shutdown. As the aircraft made its way to Heathrow
over the Atlantic, he realized he wouldn’t have enough fuel and requested
an emergency landing at Manchester airport.

Source: http://www.aaib.gov.uk/cms_resources.cfm?file=/Boeing%20747-
436,%20G-BNLG%2006-06.pdf

Impact of Wind on Range

The SAR model that was established allows us to compute an air range
that the airplane can achieve for a given set of flight conditions. The
ground range is the variable that really counts because an airplane is typically
dispatched to go from point A to point B. The ground range will depend on
the amount of wind (head or tail) the airplane will encounter during the trip.
The impact of the wind can be estimated as follows. Let us define a wind
fraction as the ratio of the wind component to the airplane’s true airspeed
(V, or TAS).
VW
vf ¼ (10:32)
V
Then, the true ground speed becomes

TGS ¼ TAS 1 + vf (10:33)

A headwind would have a negative value, resulting in a slower ground

speed than the true airspeed. For a given fuel burn and initial flight condition
 CHAPTER 10 Cruise Performance and Flight Endurance 445

(weight, altitude, airspeed), one gets a given airtime. Range is airspeed

multiplied by time, so we note that the ground range XGR will then
become proportional to the air range Xair and the wind fraction

XGR ¼ Xair 1 + vf (10:34)

We can then write an equation for the specific ground range (SGR) as a
function of SAR.

SGR ¼ SAR 1 + vf (10:35)

One can see that tail winds will greatly help with range, but head winds
are detrimental to the mission. One way to reduce the impact of the head
winds is to increase the cruise airspeed [reduce the wind fraction,
Eq. (10.32)], but this may also mean the airplane will be deviating from its
best range airspeed. Let’s set a speed ratio y as the ratio of the airspeed
flown to that of the best range airspeed. Then also adjust wind fraction as
a function of the best range airspeed.
V VW
y¼ vf ¼ (10:36)
VBR VBR

Combining Eq. (10.36) with Eqs. (10.26) and (10.34), one would get the
following (see [1] for full derivation):
pffiffiffi  
2 3 Em VBR y2 1
XGR ¼ y + vf ln (10:37)
SFCT 3 y4 þ 1 1z

This equation allows one to trade increased (or reduced) airspeed to

maximize range in the presence of winds (negative sign for head winds).
This can be used to determine small speed changes where the assumption
of constant SFCT and Em holds true. To better visualize the benefit of deviat-
ing from the best range airspeed to address winds, we define a relative range
XR that is the best ground range corrected for winds to the best ground range
uncorrected for winds (derivation also in [1]); see Fig. 10.23.
  !
4 y2 y + vf
XR ¼ (10:38)
3 y4 þ 1 1 + vf

Let’s use sample numbers to see the benefits or penalty of deviating from
best range airspeed in the presence of winds. We take a single-aisle jetliner
with a typical cruise airspeed of Mach 0.78 at 36,000 ft; this is a cruise air-
speed of 447 KTAS. A headwind of 45 kt (vf ¼ 20.1) would require the air-
plane to accelerate by 4% (an increase of almost 18 kt) to minimize the impact
of the head wind on the ground range. The improvement in ground range
would be 0.2% compared to flying at BR conditions. There is still an
446 Operational Aircraft Performance and Flight Test Practices

1.005 f = −0.10
f = −0.05
1 f = 0.0
f = 0.05
0.995
Relative range (XR)

f = 0.10
0.99

0.985

0.98

0.975

0.97
0.9 0.92 0.94 0.96 0.98 1 1.02 1.04 1.06 1.08 1.1
Relative airspeed ()

Fig. 10.23 Benefit of deviating from BR under windy conditions.

impact on ground range because the winds would reduce the ground range to
90% of the no-wind condition, but going 4% faster would result in a 90.18%
ground range, which is some benefit.
We also note that this approach may not even be a possibility if the
airplane is flying into the transonic region and there is a large drag rise.
The operator is then simply stuck with the winds at the cruise altitude.

Altitude Selection with Winds

Trading airspeed for winds may not offer much advantage, as discussed
previously. But we also know that winds do change with altitude, and this
offers the operator a chance to optimize the SGR. In the United States,
NOAA offers an Aviation Weather Center web page (https://www.aviation
weather.gov), which allows operators to estimate winds along the planned
route and trade altitude for wind speeds (see, e.g., Fig. 10.24). If one does
the ratio of the BR airspeed between two authorized flight levels (say
FL310 and FL350), the ratio of the two speeds would be
rffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VBRFL310 sFL350
¼ ¼ 0:927
VBRFL350 sFL310
If one assumes the BR airspeed at FL350 was 445 KTAS, then the BR
airspeed at FL310 would be 412.5 KTAS, a difference of 32.5 KTAS (7.3%).
Both conditions are flown under BR conditions (EBR ), so under no wind con-
ditions, the airplane at FL350 would fly the furthest. Now imagine there is no
 CHAPTER 10 Cruise Performance and Flight Endurance 447

FL Wind
240 65 kt
300 75 kt
340 75 kt
390 80 kt
450 60 kt

Fig. 10.24 Wind aloft (FL300 shown) over the United States on 11 May 2020. Source: NOAA.

wind at FL310, but 45 kt (10%) of head wind at FL350; the airplane flying at
FL310 would then actually fly the longest ground distance of the two air-
planes. The operator can optimize the cruise altitude. An example from
Boeing, shown in Fig. 10.25, was found on the Internet, showing further
tradeoff with speed and altitude for representative jetliners.

Payload-Range Diagram
Most airplanes will not be able to carry a full payload while carrying
maximum fuel; they will most likely be maximum takeoff weight limited. A
compromise must therefore be sought between the amount of payload to
be carried and the amount of fuel to be loaded in order to cover a desired

Example of increasing Tailwind at 31,000 ft Example of increasing headwind at 35,000 ft

78 78
Wind =
40
76 Wind =
76
Ground fuel mileage

Ground fuel mileage

30 35K, Wind LRC, 35K 35K, Wind = 0

= 0 LRC, 35K
74 Wind =
20 74 Wind = –10
Wind =
72 10 72 Wind = –20
31K, Win 31K, Wind = 0
d =0 Wind = –30
70 70
LRC, 31K Wind = –40
68 68
LRC, 31K
66 66
64 64
.80 .81 .82 .83 .84 .85 .86 .80 .81 .82 .83 .84 .85 .86
Mach number Mach number

Fig. 10.25 Trading cruise speed and altitude with winds to maximize SGR. Source: Boeing.
448 Operational Aircraft Performance and Flight Test Practices

80,000
MZFW
70,000

60,000
Payload weight (lb)

50,000

M
TO
W
40,000

30,000

20,000

Ma ume
vol
x fu
10,000

el
0
0 1000 2000 3000 4000 5000 6000 7000
Range (n miles)

Fig. 10.26 Payload-range diagram.

range. A payload-range diagram (see Fig. 10.26) is a

chart of range that can be covered by the airplane Note: This diagram is not
with a given payload. appropriate for carrying
The payload-range diagram can rapidly identify external payload where the
drag of the aircraft varies
all the possible combinations of payload and range
depending on the size, shape,
that a given aircraft can do based on a design and location of the
mission. For example, ISA condition, 10 min external payload.
taxi-out, takeoff to 1500 ft AGL using Flap 10, 250
KCAS/Mach 0.70 climb, Mach 0.78 step cruise, Mach 0.70/250 KCAS
descent, Part 121 reserves.
The boundaries of a payload range diagram are:

• Maximum payload (maximum zero fuel weight, MZFW) that can be

carried by the airplane. In the payload weight–limited range, fuel is added
to meet the desired range to be covered.
• At a given point, the airplane will become maximum takeoff weight
(MTOW). To increase the range, payload will need to be traded for fuel
(reducing payload while increasing fuel, pound for pound).
• At one point, the fuel tanks will be full (maximum fuel volume). The only
way to further increase the range of the airplane will be to reduce the
payload so as to reduce the airplane drag due to weight.

The point where zero payload is reached is called the ferry range of the
airplane. The zero payload line, all ranges, represents the airplane operating
weight empty (OWE). If the OWE is altered for any reason, the payload-range
 CHAPTER 10 Cruise Performance and Flight Endurance 449

chart will change. (An increasing OWE will result in a lowering of the payload
for a given range.) Figure 10.27 represents a payload-range diagram created
by Airbus for the A350-900 provided in the airport planning manual (modi-
fied by author). Of interest, it shows two fuel capacity conditions with differ-
ent ferry ranges, but both are limited by the same MTOW.
Some manufacturers will offer multiple MTOW variants of a given air-
plane. That varying MTOW will adjust the line limited by MTOW on the
payload range. Figure 10.28 is an example of payload range capability vs
takeoff weight (TOW). One can see at a glance the general advantage of
the payload-range diagram. Although not a guarantee of performance, it pro-
vides a good summary of the capability of the airplane.

Flight Endurance
At the end of the flight, an airplane may have to hold in the pattern while
awaiting clearance to land. In this case, distance is not an issue; time becomes
the driving factor for defining the airspeed to use. Minimum fuel burn while
holding is achieved when the rate of fuel consumption per unit of time is
minimized. This can be expressed as the following (for jet airplanes):
@t 1 E
 ¼ ¼ (10:39)
@W SFCT Tr SFCT W
Therefore, fuel burn per unit of time is minimized while flying at con-
ditions of maximum (E/SFCT ), which is essentially maximum lift-to-drag

MTOW − 280 000 kg (617 295 lb); MFC 140 795 1 (37 195 US gal)
Airbus A350-900 MTOW − 280 000 kg (617 295 lb); MFC 166 488 1(43 982 US gal)

Range (km)
2000 4000 6000 8000 10,000 12,000
00 14,000 16,000 18,000 20,000 22,000
70

60
Maximum structural payload
120
120
50
Payload (x 1000 kg)

Payload (x 1000 lb)

100
40
325 Passengers 80
30
60

20 Standard day conditions

40
Typical international flight profile
95 kg per passenger including baggage
10 20
Basic configuration with crew rest compartments
and other optional features
0 0
500 1500 2500 3500 4500 5500 6500 7500 8500 9500 10,500 11,500 12,500

Range (nm)

Fig. 10.27 A350-900 payload-range. Source: https://www.airbus.com/aircraft/

support-services/airport-operations-and-technical-data/aircraft-characteristics.html.
450 Operational Aircraft Performance and Flight Test Practices

Standard day, zero wind Payload/Range

Mach 0.84 cruise
Step climb at 2000 feet increments 777-300ER/GE90-115BL
Normal power extraction and air conditioning bleed
Typical mission rules Consult using airline for specific operating
procedure and OEW prior to facility design
540

240 530
Max zero fuel weight = 524,000 lb (237,682 kg)

Bra ss w 00)
520

gro 10 )
235

ke eig
rel ht
(x
510

lb

ea
77
230

se
5(
(kg )
75

3
0(
500

52
72

34
225

5(
70

0)
32
0(
490
OEW plus payload

1000 pounds

9)
(1000 kilograms)

67

31
220

5(

8)
65

30
480

0(

6)
62

29
215

5(

5)
60
470

28
0(

3)
27
210
57

2)
5(
460
55

26
0(

1)
205
24
52

450
9)
5(

320,8

Fuel
23
50

200
8)
0(

lb (k ,540)
440

capa
60 (1
22
47

7)
5(

g)
195

city
430

45
21
45

5)
0(
20

190 420
4)

410
0 1 2 3 4 5 6 7 8 9 10
1000 nautical miles range

Fig. 10.28 Takeoff weight impact on Boeing 777-300ER payload range diagram. Source:
https://www.boeing.com/resources/boeingdotcom/commercial/airports/acaps/777_2lr3er.pdf.

condition (or minimum drag, Em ) airspeed [see Eq. (8.25) in Chapter 8]. We
can also see that the best condition for this flight mode will also be one where
the airplane configuration yields the largest Em , which is with landing gear
and flaps up. The time aloft as a function of fuel burn can be estimated by
maintaining the L/D and assuming constant SFCT .
 
E 1
t¼ ln (10:40)
SFCT 1z

For propeller-driven airplanes, where fuel flow is proportional to the

engine power P required to maintain the flight condition (Pr ¼ D V ¼ hp P),
the time vs fuel burn will be

@t 1 hp hp E
 ¼ ¼ ¼ (10:41)
@W SFCP P SFCP Pr SFCP W V

We are again faced with three systems that need to be optimized (aero-
dynamics of the airplane, efficiency of the propeller, and efficiency of the
engine) to increase the air time. The general condition where Eq. (10.41) is
maximized (tmax ) will generally be near the condition for minimum power
required (MP) for flight [see Eq. (8.38) in Chapter 8], an airspeed slower
 CHAPTER 10 Cruise Performance and Flight Endurance 451

than minimum drag.

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffi
2W 4 K 4 1
VMP ¼ ¼ V (10:42)
rSL s S 3 CDo 3 Em
 
hp EMP 1
tmax ¼ ln (10:43)
SFCP VMP 1z

Electrical Propulsion: A Step Change?

At the time of writing this textbook, electrical propulsion was starting to
take the forefront of new airplane discussions. Designers were looking at
small regional airplanes as the next step (aviation electrification), after
some successful prototype work on small single-engine airplanes and the
Airbus E-Fan. Some startup companies were bringing products to market,
or at least trying, because the endeavor of designing, certifying, and produ-
cing a new Part 23 or Part 25 airplane is a cash-intensive one, and some com-
panies were not able to continue.
As a way to introduce the subject of electrical propulsion, we have elected
to discuss the cruise performance of an airplane using traditional engines (gas
turbines and piston-props) vs the initial contenders of all electric and
hybrid-electric engines.
Our first question is: How efficient is an airplane in converting the energy
stored on board into propulsive energy? This includes not just the thermal
energy efficiency of the system, but the entire energy of the fuel and/or
battery into thrust multiplied by airspeed and time.
For this exercise, the power will be expressed in kW because it is the pre-
ferred unit for electrical propulsion. We start with a single-aisle jetliner.
Table 10.2 shows typical cruise performance from a single-aisle jetliner
(data from Internet). The table contains the performance for a range of alti-
tudes (32,000 to 37,000 ft) and a range of weights (34,000 to 62,000 kg).
For each weight/altitude condition, we find the following data of interest:
• Cruise Mach (in this case, it is the long-range cruise condition, 1% less
efficient than max range)
• The corresponding true airspeed (KTAS)
• The fuel flow per engine (in kg/h)
If we pick an airplane that weighs 50,000 kg (110,000 lb), cruising at
36,000 ft at Mach 0.745, the table specifies a fuel flow of 1071 kg/h/engine
or 2142 kg/h (4712 lb/h) total for this twin-engine airplane. Jet fuel contains
about 18,550 BTU/lb; thus, in the cruise condition specified, the airplane is
consuming
BTU lb BTU
18,550  4712 ¼ 87,407,600 ¼ 25,616:6 kW
lb hr hr
 452
Table 10.2 Cruise Table for Single-Aisle Jetliner

Operational Aircraft Performance and Flight Test Practices

Long Range Cruise Table 37000 FT to 32000 FT
PRESS ALT WEIGHT (1000 KG)
(1000 FT) (STD
TAT) 62 60 58 56 54 52 50 48 46 44 42 40 38 36 34
%N1 92.3 90.4 89.0 87.7 86.5 85.4 84.5 83.6 82.8 82.0 81.2 80.2 79.2
MAX TAT 215 27 21
3 KIAS 239 240 240 240 240 240 240 240 240 240 239 237 233
(233) MACH .742 .744 .745 .745 .745 .745 .745 .745 .745 .744 .742 .735 .725
FF/ENG 1293 1226 1169 1118 1073 1032 997 965 936 908 878 844 809
KTAS 426 427 427 427 427 427 427 427 427 427 426 422 416
%N1 93 5 91.5 89.8 88.5 87.3 86.2 85.2 84.3 83.5 82.7 82.0 81.2 80.3 79.4 78.3
MAX TAT 220 211 24
36 KIAS 244 245 245 246 246 246 246 246 246 246 245 245 242 239 235
(233) MACH .741 .743 .744 .745 .745 .745 .745 .745 .745 .745 .744 .742 .736 .727 716
FF/EXG 1395 1321 1258 1203 1154 1110 1071 1036 1004 975 947 918 884 849 812
KTAS 425 426 427 427 428 426 427 427 427 427 427 426 422 417 411
%N1 91.0 89.6 88.4 87.3 86.3 85.3 84.5 83.7 83.0 82.3 81.5 80.6 79.8 78.8 77,7
MAX TAT 27 21
35 KIAS 251 251 251 252 251 251 251 251 251 251 250 248 245 241 237
(230) MACH .744 .745 .745 .745 .745 .745 .745 .745 .745 .744 .743 .737 .728 .718 ,705
FF/ENG 1356 1297 1244 1197 1154 1115 1081 1050 1021 992 963 929 894 857 818
KTAS 429 429 430 430 430 429 429 429 429 429 428 425 420 414 406
(Continued)
 Table 10.2 Cruise Table for Single-Aisle Jetliner (Continued)

Long Range Cruise Table 37000 FT to 32000 FT

PRESS ALT WEIGHT (1000 KG)
(1000 FT) (STD
TAT) 62 60 58 56 54 52 50 48 46 44 42 40 38 36 34
N1 89.4 88.3 87.3 86.3 85.5 84.7 83.9 83.2 82.5 81.8 81.0 80.1 79.2 78.2 77.1
MAX TAT 2

CHAPTER 10
34 KIAS 257 257 257 257 257 257 257 257 257 256 254 251 247 243 238
(228) MACH .745 .745 .745 .745 .745 .745 .745 .745 .744 .742 .737 .729 .719 .707 .693
FF/ENG 1337 1286 1240 1198 1161 1127 1096 1067 1038 1009 974 939 902 864 823
KTAS 431 432 432 431 431 431 431 431 431 430 427 422 416 409 401

Cruise Performance and Flight Endurance

%N1 88.2 87.2 86.3 85.6 84.8 84.1 83.4 82.8 82.1 81.3 80.5 79.6 78.6 77.6 76.5
MAX TAT
33 KIAS 263 263 263 263 263 263 263 263 262 260 257 253 249 244 239
(227) MACH .745 .745 .745 .745 .745 .745 .745 .744 .742 .737 .729 .720 .709 .696 .682
FF/ENG 1329 1284 1244 1207 1174 1143 1114 1086 1055 1021 985 948 910 869 829
KTAS 434 433 433 433 433 433 433 433 432 429 424 419 412 405 397
%N1 87.2 86.4 85.6 84.9 84.3 83.6 83.0 82.3 81.5 80.8 79.9 79.0 78.0 76.9 75.8
MAX TAT
32 KIAS 269 269 269 269 269 269 269 268 266 263 259 255 251 245 240
(225) MACH .745 .745 .745 .745 .745 .745 .744 .742 .737 .729 .720 .710 .698 .684 .671
FF/ENG 1329 1290 1254 1222 1191 1162 1134 1102 1067 1031 995 956 916 875 834
KTAS 435 435 435 435 435 435 435 433 430 426 421 415 408 400 392
Max TAT not shown where %N1 can be set in ISA þ 308C conditions.
Increase/decrease %N1 required by 1 % per 58C above/below standard TAT.
Increase/decrease fuel flow 2% per 108C above/below standard TAT.
Increase/decrease KTAS by 1 knot per 18C above/below standard TAT.

453
Shaded area approximates optimum altitude.
454 Operational Aircraft Performance and Flight Test Practices

To see how efficient the airplane powerplant is at converting the energy

provided by the fuel into propulsive energy, we estimate the power required
to maintain level flight. Single-aisle jetliners typically cruise at lift-to-drag
(L/D) conditions of 14 to 17; let’s assume that for this example, the L/D
was 16. The lift required in level flight is equal to the airplane weight, thus
the drag of the plane is
L W W
¼ ¼ 16 ! D ¼ ¼ 6875 lb
D D 16
From these numbers, we can derive the engine thrust specific fuel con-
sumption (SFC). We get
4712 lb=hr lb=hr
SFC ¼ ¼ 0:685
6875 lb lb
This is typical of the 1980s to 1990s turbofans used on single-aisle jets.
The airplane, for this example, is traveling at Mach 0.745 for a true airspeed
of 427.4 kt. To maintain level steady-state flight, the thrust produced must
equal the drag of the airplane. We then compute the power required to main-
tain level flight, which is simply the thrust produced times the true airspeed.
ft lb
Pr ¼ T  V ¼ 4,959,324 ¼ 6,723:9 kW
sec
Because the airplane consumes 25,616.6 kW of fuel energy to produce
6,723.9 kW of propulsive energy, the efficiency of the powerplant to
convert the energy is simply the ratio of the two, or
6,723:9
Efficiency ¼ ¼ 26:2%
25,616:6
The majority of the rest of the energy consumed is wasted heat energy
out of the tail pipes. (There is a small fraction of the energy that is used
for pressurization, hydraulic power, and electrical power.)
We do the same for a small 19-passenger regional turboprop, Table 10.3.
This airplane, weighing 15,000 lb in cruise at 15,000 ft (typical for shorter
range turboprop airplanes) burns 298 lb/h/engine while traveling at 226
KTAS, for a total of 596 lb/h. These small airplanes have a typical cruise effi-
ciency of 12 to 14, so we pick 13 for this analysis.
The power from the fuel is
BTU lb BTU
18,550  596 ¼ 11,055,800 ¼ 3240:1 kW
lb h h
The drag of the airplane is
L W W
¼ ¼ 13 ! D ¼ ¼ 1154 lb
D D 13
 Table 10.3 Cruise Table for Turboprop

LONG RANGE CRUSIE

FLT LVL WEIGHT (LB)
11,500 12,000 12,500 13,000 13,500 14,000 14,500 15,000 15,500 16,000
15 200 201 201 203 204 206 207 209 210 212
206 206 207 209 210 211 213 215 216 217

CHAPTER 10
352 355 358 363 367 372 376 383 387 392
51.4 52.1 53.0 54.6 55.8 57.2 58.6 60.5 61.7 63.1
50 194 194 195 197 199 201 203 203 204 205
210 210 211 213 215 218 219 220 220 221

Cruise Performance and Flight Endurance

327 330 333 339 345 352 358 361 363 367
49.6 50.3 51.3 53.1 54.8 57.0 58.6 59.4 60.3 61.3
100 184 184 186 187 188 189 190 191 192 193
214 215 216 218 219 220 222 223 223 224
296 299 303 308 312 316 322 327 330 334
46.8 47.5 48.9 50.3 51.5 52.5 54.3 55.6 56.6 57.7
140 176 176 177 178 179 180 181 181 182 183
218 218 219 220 222 223 224 225 225 226
273 276 280 284 289 294 299 303 307 312
44.5 45.3 46.3 47.6 49.1 50.5 51.7 52.8 53.9 55.3
150 173 174 174 175 177 178 179 179 180 181
218 219 220 220 222 224 225 226 226 227
267 270 274 278 284 289 294 298 302 307
43.8 44.7 45.7 46.9 48.5 50.0 51.3 52.5 53.6 55.0

455
456 Operational Aircraft Performance and Flight Test Practices

And the power required for level flight is

ft lb
Pr ¼ T  V ¼ 440, 179 ¼ 596:7 kW
sec

For an efficiency of

596:7
Efficiency ¼ ¼ 18:4%
3240:1

Considering the current (and future) cost of energy (see Fig. 10.29), one
would expect that improved efficiency in energy consumption would go a
long way to improving the direct operating cost of an airplane.
When one considers an all-electric airplane in which the energy is stored
in batteries, one could expect a system similar to the one shown in Fig. 10.30.
When applying this system to the small turboprop example from earlier in
the chapter, one could find that the improved efficiency would reduce the
rate of energy use by a factor of 4 if everything else stayed the same (cruise
speed, altitude, weight, and aerodynamic efficiency).
Where the comparison fails a little is that, contrary to a conventional air-
plane where the weight reduces as the fuel is burned and thus the drag
reduces as well, the battery-powered electric airplane will maintain the
same weight throughout the flight from takeoff to landing. So, some of the

200

180
Average cost (US$) per barrel of fuel

160

140
nd
Tre

120

100

80

60

40

20

0
1940 1950 1960 1970 1980 1990 2000 2010 2020
Year

Fig. 10.29 Evolution of Jet A fuel prices over time.

 CHAPTER 10 Cruise Performance and Flight Endurance 457

All electric ≈ 75–80% efficiency Propeller

85% eff.

Battery pack DC Controller AC Electric motor

0.30 kWh/kg 99.5% eff. 97% eff.

98% eff.
Power cables
>99% eff.

Fig. 10.30 Example of an all-electric propulsion system.

benefit of the much more efficient engine (electric motor) is lost due to the
weight not reducing.
At the time of writing of this chapter, the batteries available for the avia-
tion industry were at about 0.22 kWh/kg of specific energy with an expected
target of 0.30 kWh/kg by the time the first practical purpose-built, all-electric
regional airliner reaches certification in 2025. In comparison, the specific
energy of Jet A fuel is 11.98 kWh/kg (18,550 BTU/lb), about 40 times
higher. So for the same amount of energy on board the airplane (fuel vs
battery), accounting for the higher overall efficiency of the electrical propul-
sion, one would expect an airplane of a given size to weigh significantly more
to perform the same mission.
We provide the following example to discuss numbers. This is a conver-
sion of a small turboprop from fuel powered to electricity powered. We
create a SAR equation for the electric airplane that presents the distance
covered per amount of energy used (km/kWh). The equation takes on the
following form of the airplane true airspeed (TAS) divided by the power
from the system P:

TAS
SAR ¼ (10:44)
P
The power provided by the system must equal the power required to
sustain the flight (Pr = D TAS). This can be represented by

Pr ¼ D TAS ¼ hp hmotor hsyst I V (10:45)

where:
Pr ¼ Power required for level flight
D ¼ Drag
TAS ¼ True airspeed
hp ¼ Propeller efficiency
hmotor ¼ Motor efficiency
hsyst ¼ System efficiency to carry power from battery
I V ¼ Power from the battery (current multiplied by voltage)
458 Operational Aircraft Performance and Flight Test Practices

The airplane selected has a flight weight of 10,000 lb. The airplane long-
range cruise speed at 10,000 ft pressure altitude is 175 KTAS; under these
conditions, the airplane SAR is 0.41 n miles/lb. (Using a specific energy for
Jet A fuel of 11.98 kWh/kg, one gets a SAR of 0.138 km/kWh.) In this con-
figuration, the airplane has two pilots and nine passengers (weight of 2200 lb)
and carries about 2000 lb of fuel. This amount of fuel represents 10,873 kWh
of energy. Assume this airplane burns 1500 lb of its total fuel and maintains
cruise conditions (ratio hp E/SFCP ); the SAR at the end of the cruise segment
is then

W1
SAR2 ¼ SAR1 ¼ 0:162 km=kWh
W2

This gives an average SAR of approximately 0.15 km/kWh (0.446 n

miles/lb of fuel). This would convert to a distance covered of about 665
n miles (1231 km). To complete the analysis, assume the propeller effi-
ciency is 80%, the L/D during the cruise is 12, and the SFCP is 0.765
lb/h/SHP.
To do the first conversion (replace the turborprop engines and remove
fuel, add electric motor and batteries at 0.3 kWh/kg), and respecting the
initial weight of 10,000 lb plus keeping the same propellers, we would end
up with about 275 kWh of batteries (918 kg) (see Fig. 10.31). The electric
motor has an efficiency in cruise of 96%, and the system can transfer the
power to the motor from the batteries at about 97% efficiency. With a
cruise weight of 10,000 lb that remains constant throughout the flight, the
modified plane would have a SAR of 0.73 km/kWh.
This would be a cruise efficiency improvement by a factor of a little over
5, which is impressive. But the amount of energy on board being 40 times less,

Baseline Converted

Fig. 10.31 Conversion of a small turboprop into an all-electric airplane.

 CHAPTER 10 Cruise Performance and Flight Endurance 459

$0.350
Cost of electricity in various areas
Hawaii
$0.300

$0.250 Sweden

Alaska
$0.200 California
$/kWh

New York
$0.150 Maine

Oregon
$0.100 Washington

$0.050 Quebec

$0.000
0 1 2 3 4 5 6 7 8
$/USG Jet A

Fig. 10.32 Cost of electricity vs cost of fuel, same energy.

the actual range capability of the airplane, using 75% of the energy (same as
the preconversion airplane), would result in a distance of 150 km, about 12%
of the turboprop airplane. Further range could be gained by the electric
airplane with the addition of batteries, but this would result in a heavier
weight overall (including structural reinforcement).
Where a conversion to an electric airplane, with today’s battery technol-
ogy, would really shine is on short routes where the turboprop’s inefficiency
would impact the airplane. On a 150-km route, the electric airplane will
use 207 kWh to complete the cruise; the turboprop airplane would require
335 lb of fuel (50 U.S. gal, 1821 kWh). To compare the cost of operation
between the two, we offer the conversion chart in Fig. 10.32 for equivalent
energy ($3 per U.S. gal is approximately equal to $0.082/kWh of electricity).
The turboprop ops would cost $150 of fuel whereas the electric airplane
would cost $17 of electricity. The reality on the operating cost will, of
course, depend on the markets and the price for fuel vs electricity—some
regions will see tremendous benefits (low cost of electricity) whereas
others will still see an improvement in operating cost.
As the price of fuel has gone up over the years, older-generation turbo-
props have fallen out of favor due to their higher operating cost (cost per
seat-mile). Without these small airplanes, many small communities lost the
air service they had, and major airlines have tended to favor the
hub-and-spoke approach to help fill larger planes. There is still a need for
shorter routes (200 km and above) for both passengers and cargo that can
460 Operational Aircraft Performance and Flight Test Practices

Fig. 10.33 Heart Aerospace ES-19 all-electric regional airplane.

be filled only by smaller planes like thin routes (small number of passenger or
cargo), interisland, distant suburb to downtown airports, inter –small cities,
and so forth. An airplane with reduced operating cost and good field per-
formance would allow these markets to be served.
At this time of writing this textbook, Heart Aerospace was developing the
ES-19 all-electric airplane (see Fig. 10.33) to meet the demand and the gov-
ernmental mandates of the Nordic countries. (Norway has mandated all-
electric domestic flights by 2040; Sweden has mandated domestic flights be
fossil fuel–free by 2030.) The airplane is being designed specifically from
the start to be all electric and of relatively conventional design to ease the cer-
tification acceptance and market introduction. The airplane will also have
excellent field performance. All-electric airplanes have no in-flight emissions.
Note that when an airplane becomes very efficient (low energy cost),
other “hidden” costs show up more clearly, such as the following:

• Crew cost becomes dominant.

• The energy cost to power systems (avionics, environmental control,
pressurization) that used to be background noise is now starting to show
up in the energy equation. For example, on gas turbine engine airplanes,
going up in altitude improves overall energy consumption; however, on an
all-electric plane, the cost of pressurization can become important in the
overall energy equation.

On the good side, we see the following:

• Clean flight, zero emission.

• Ideal for short ranges where turboprop inefficiencies show up. (Idle power
per engine on the ground typically consumes 0.5 to 1.0 kWh.)
 CHAPTER 10 Cruise Performance and Flight Endurance 461

• Maintenance costs are significantly reduced. An Internet search indicates

that small turboprop engines can easily cost $300,000 each to acquire.
Then they need overhaul every 3000 to 4000 flight hours at a cost of
approximately $250,000/engine. Electric motors of the same size may cost
only $50,000 to acquire and have an essentially unlimited life (compared to
the typical 20,000 flight hours of economical life of an airplane).
• The current battery technology (Li-ion) has shown a yearly improvement
of about 5% in specific energy and a fairly rapid decrease in acquisition cost.
The next generation (Li-S) promises levels of 0.40 kWh/kg (as of 2020) and
Li-air has the possibility of achieving 0.8 kWh/kg; however, these last two
have not yet matured enough for aviation application in the
immediate future.
This author believes the industry must invest now to sort out the details
of the architecture and certification requirements for electric airplanes. The
airplanes will improve in range and size as batteries improve. There is also the
possibility of future technology (energy storage or generation) that may come
along. Gasoline/fuel-powered airplanes are reaching a plateau with little
improvement (less than 1% per year) achievable, and this at significant cost
to the manufacturers. For now, however, they are still the standard.

Hybrid-Electric: An Evolutionary Change

The low specific energy of batteries has led some designers to consider
hybrid-electric configurations where a reciprocating engine or gas turbine
is combined with electric power from batteries. The general concept is to
use, say, a smaller gas turbine than required for the most demanding part
of the flight (typically takeoff and go-around) but sufficient to support the
longer sections of the flight like cruise where power demand can be signifi-
cantly less. The missing power for takeoff and possibly the climb to cruise
is then provided by batteries. This smaller turbine would then be turning
to near its maximum continuous power for most of the flight where its

Prop eff.
Hybrid-electric ≈ 40–45% efficiency 0.85

Fuel AC Controller AC Electric motor

18,550 BTU/lb 99% eff. 96% eff.
11.98 kWh/kg Generator
95% eff.
Gas turbine
55% eff.
Batteries

Fig. 10.34 A hybrid-electric configuration (serial connection).

462 Operational Aircraft Performance and Flight Test Practices

Fig. 10.35 Zunum ZA-10. Source: Zunum.aero.

efficiency is highest. (We saw in Chapter 8 that a gas turbine at low power can
have very low efficiency.) When combined in this manner, one can generally
expect an increase in the overall efficiency (see Fig. 10.34).
The configuration of Fig. 10.34 lends itself well to distributed power
where a single gas generator, supplemented by batteries, could feed multiple
electric motors. An example of this approach was the Zunum ZA-10 regional
airliner concept (see Fig. 10.35). The airplane had a single gas turbine tur-
boshaft located in the aft fuselage providing electrical power to two ducted
fans on either side of the fuselage. The gas turbine was supplemented by bat-
teries for takeoff to achieve the proper power. This author had a chance to
briefly examine the concept, which showed promise. For one thing, the
concept of engine failure on takeoff was addressed by redundancy in the
system; should the gas turbine fail during the takeoff, the airplane had
enough power from the batteries to not see a reduction in the takeoff
power and to perform up to a 100-n-mile mission with no reserves. This
meant the airplane could have easily come back to land with no loss of per-
formance. Because electric motors degrade more gently then gas turbines
(losing partial power), the airplane offered increased safety.
The hybrid-electric airplane still burns fuel, so the impact on the environ-
ment is similar to that of the engine it uses. But the addition of battery power
and increased efficiency of its gas turbine (or reciprocating engine) turning
closer to its optimum value meant that there was an overall reduction in
emissions for a given flight. Part of the efficiency comes from the reducing
weight as the fuel is burned, of course.

The Case for the Reduced Energy Plane: The Hydrogen Plane
The specific energy of the batteries imposes a weight issue on modern
electric airplanes, and it has so far kept their size small. An alternative
source of energy that also shows some promise is hydrogen. Hydrogen has
a specific energy of 33.314 kWh/kg, almost three times more than Jet A
fuel, so an airplane with a given aerodynamic efficiency and a given power-
plant efficiency has, in theory, a chance to have a much lower operating
weight with the use of hydrogen. Another interesting aspect is that the emis-
sions from burning hydrogen are water vapors with no CO2 .
 CHAPTER 10 Cruise Performance and Flight Endurance 463

Unfortunately, hydrogen also has a low energy density—liquid hydrogen

has 2.36 kWh/l compared to 9.2 kWh/l for jet A fuel. This means the
liquid hydrogen (LH2 at –2538C) would take more than three times the
volume of fuel. The scenario is now reversed. Liquid hydrogen airplanes
will tend to become larger to carry the fuel required for the flight. On
small planes, the extra volume would have a detrimental impact on the
airplane parasite drag.
This author worked on an LH2 plane project in the late 2000s, the
H-Plane (see Fig. 10.36). The concept was simple: take advantage of the
much-reduced weight of the energy (through the use of liquid hydrogen)
to build a cargo airplane with a trans-Pacific range that could fly from the
west coast of the United States to Beijing (5500-n-mile design mission) in
about the same time as the current cargo competitors could using the promi-
nent cargo airplane of the time, the Boeing 747-200. The 747 had to do a
refueling stop in Alaska, so we could take advantage of the nonstop
mission design and slow down the airplane (reduce parasitic drag). The air-
plane ended up as a point design 120 KCAS climb-cruise-descent airplane
with a cruise altitude of 60,000 to 65,000 ft. The low dynamic pressure
point allowed us to trade structure weight even more. The cruise speed at
high altitude was Mach 0.70, or 400 KTAS in a standard atmosphere. The air-
plane had a projected maximum L/D of 32. Because the airplane flight air-
speed was of limited range, the wing twist distribution of the fore and aft
wings could be optimized to improve the overall aerodynamic efficiency.
The airplane would use a new, variable compression engine (constant
chamber pressure with increasing altitude) that maximized the efficiency in
cruise. Finally, the airplane was to be equipped with an unducted fan
around the fuselage. The expected energy required for the flight would
have been equivalent to 25,000 lb of LH2 , a very small fraction of the
takeoff weight. It was estimated at the time that the 747-200 would have
required more than 500,000 lb of fuel to accomplish the same mission.

Fig. 10.36 The H-Plane.

464 Operational Aircraft Performance and Flight Test Practices

Fig. 10.37 Aerovironment Global Observer H2 -powered airplane. Source: NASA (https://www.
nasa.gov/aeroresearch/tech-excellence/2010/global-observer-wing) or Aerovironment.

The team built a small prototype engine and a scaled model before funding
ran out. The idea was presented to the FAA with a concept of an unmanned
cargo, and it was well received due to the over-water flight.
In the same timeframe, other hydrogen airplanes came about, including
the unmanned Aerovironment Global Observer (see Fig. 10.37), designed

Fig. 10.38 Airbus hydrogen concept planes, 2020 announcement. Photo: Airbus.
 CHAPTER 10 Cruise Performance and Flight Endurance 465

to use hydrogen and perform long-endurance (up to 1 wk in the air) missions

between 55,000 and 65,000 ft. The airplane suffered a crash in April 2011.
The program was sponsored by the Pentagon and was terminated in 2012.
The Global Observer used LH2 as fuel to feed an internal combustion
engine driving four electric motors (two per wing, hybrid-electric configur-
ation). It had a wingspan of 175 ft.
It is also fitting to mention that at the time of publication of the book in
late 2020, Airbus announced they were studying 3 concept hydrogen air-
planes, Figure 10.38. The future will be interesting.

References
[1] Asselin, M., An Introduction to Aircraft Performance, AIAA Education Series, AIAA,
Reston, VA, 1997.
[2] Rutan, B., “Voyager Milestone: Non-refueled Flight Around the World, The Design
Approach,” Keynote address 2, AGARD-CP-547, Nov. 1993.
Chapter 11 Cruise and Drag
Performance
Testing

Chapter Objective
We looked at the components that form the drag and thrust models in
Chapter 8. We included the aspect of fuel burn and its impact on weight
throughout the flight in Chapter 10. We also discussed electric power and
the constant weight model. All these models have trends and expected beha-
viors, and each system (airframe, engine, and propeller) may interact with each
other. It is now time to integrate the models and validate by testing, possibly
adjusting the models to best match the performance measured. In this
chapter, we will focus on level flight and verify that the excess thrust model
provides good results; that is, the drag model is well integrated with the
engine/propeller model.

Creating a Drag Model from Flight Test Results

O
ne thing this author has noticed over the years is that two people
looking at the same flight data trace to do drag data reduction
will get two different results. I have tried this in class with students
looking at the same data and I always get the same “scatter.” The results do
not vary by much, but they are different. The secret to creating a good model
is that consistency in the data the reduction procedure will produce a more
representative solution and this is usually best done when a single individual
reduces the data. The entire flight envelope can be separated into multiple
segments with a drag model that fits that envelope best, and then the work-
load can be split among multiple people, each handling one envelope. If more
than one person is involved in creating drag models, the group must agree on
the ground rules for the thrust-drag bookkeeping methodology ahead of time
(i.e., what is considered a thrust, could it simply be the engine deck from
the manufacturer, and what corrections are applied to the model to match
the demonstrated performance, as discussed in Chapter 8).
We typically create two fundamental drag models for Part 25 airplanes (or
Part 23 small airplanes), one low speed and one high speed. The low-speed
drag model focuses on the speeds that create the certification models (most
adverse center of gravity position, focused on the takeoff and landing

467
468 Operational Aircraft Performance and Flight Test Practices

speed part of the envelope). The second one, the high-speed drag model, will
focus on the climb/cruise/descent phase, often dealing with compressibility
effects, higher altitudes, and possibly using a less adverse center of gravity
position. This chapter follows this approach.

Testing for Low-Speed Drag

The focus of the low-speed drag testing campaign is on creating a model
to be used for certification and creating the airplane flight manual (AFM),
although exploratory testing is usually done earlier to get a first look at the
test vs preflight estimates or to support the creation of the simulator aero
model. For this reason, one needs to take appropriate steps to ensure the
testing is done right the first time to minimize the number of repeat flights.
The simplest approach for drag analysis is to start with an “approved/
accepted” engine deck (an engine model) that defines thrust (or power) vs
a set of parameters (airspeed, altitude, temperature, throttle position, fuel
flow, turbine speed, etc.) and then fit your drag curve. This is the starting
point, not the final model; we will expand on this in the section “Data
Reduction: Drag Polar Fitting.” Then, for a turbojet/turbofan-equipped air-
plane, we get this initial drag model:
D ¼ Tengine deck (11:1)

There are some limits to this approach that the performance engineer will
need to address. Note we did not include the effects of angle of attack, for
example. One can include the angle of attack in a slightly more complicated
model, but there may be no benefit depending on how one uses the model. In
the end, what is important is, does the model created produce repeatable
results? Including angle of attack may help compare the drag model produced
to the wind tunnel/computational fluid dynamics (CFD) models.
For a reciprocating, turboprop, or electric engine with a propeller or a fan,
one can start with the power model from the engine manufacturer and the
propeller model from the propeller manufacturer.
h pmodel Pengine deck
D¼ (11:2)
V
As specified in Chapter 8, propellers can have a large impact on the air-
plane lift and drag (see Fig. 11.1), and one needs to decide before testing if the
proposed model from Eq. (11.2) will provide an acceptable model for the task
(create a certification model) or not; if not, what else needs to be included in
the model. We will discuss this in the “Data Reduction: Drag Polar Fitting”
section.
To execute the test and help reduce test scatter, it is best to provide the
test crew with acceptable test limits. Limits that are too tight would lead to
unnecessarily long test execution as the crew tries to stabilize on conditions
 CHAPTER 11 Cruise and Drag Performance Testing 469

Propeller thrust

Propeller feathered

Lift

Prop feathered
Failed engine

Fig. 11.1 Impact of propeller slipstream on wing lift.

for the desired test duration; limits that are too loose would result in more
postflight data reduction/correction. The following test limits have usually
provided a good compromise for drag testing:
• Maximum airspeed variation during the test: + 2 KIAS.
• Maximum altitude variation: + 50 ft.
• Maximum temperature variation: + 18C.
• Minimum test time: 2 min or one phugoid period.
W A phugoid is an airplane motion in which altitude (potential energy) is

traded for speed (kinetic energy) in a cyclic pattern but where the total
energy remains essentially constant. It is a basic flight dynamics mode of
an airplane.
• The pilot should never touch the throttle during the test. We need the
engine thermodynamics to remain stable during the execution of the test.
• Desired propeller settings, if applicable (blade pitch, engine torque, etc.).
Note that testing done near the minimum drag airspeed can be difficult.
The drag curve vs airspeed is typically very flat (see Fig. 11.2), and small
atmospheric disturbances can easily push the airplane away from the target
condition. The pilot would then try to adjust thrust to counter accelera-
tion/deceleration, but these would be seen as “digital/step” throttle (throttle
lever angle, TLA) input, after which it takes some time for the engine to spool
up or down as desired. The drag curve is very flat, so there may still be a large
 470
Operational Aircraft Performance and Flight Test Practices
Calibrated airspeed (kt) Pressure altitude (ft) End of test 10,000
Calibrated airspeed (kt)

201.5
201.0 10.4

Pressure altitude
200.5

(ft) (103)
10.3
200.0
199.5
9000
10.2
199.0
198.5 10.1
198.0 8000
TLA (deg) N1 (%) Small pilot input
11 65

Drag (lb)
10
60 7000
TLA (deg)

N1 (%)
8 55 Test
7
6
50
6000 condition
5 45
Nz (g)
0.04 5000
0.02
0.00
Nz (g)

–0.02
–0.04
–0.06 Atmospheric disturbance 4000
–0.08 100 150 200 250 300 350
0 10 20 30 40
Time (s)
KCAS

Fig. 11.2 Example of testing near minimum drag.

 CHAPTER 11 Cruise and Drag Performance Testing 471

speed change for the small TLA change. For this reason, one typically accepts
a lesser stabilized time on condition and the presence of more data scatter.

Test Scope, Preparation, and Instrumentation

The main purpose of low-speed drag testing is to collect data for the
AFM. One should first select all configurations (flaps/slats position, gear
position, and other special features) that one intends to certify. For takeoff
configurations, the target airspeed condition will be near 1.13 VSR (see
takeoff Chapter 15). This is a stall speed ratio, so the condition will be
done in essentially a constant angle of attack and thus constant lift coeffi-
cient. One should bracket this test condition with one or two speed points
lower and two or three speed points higher. One would then expect four
to six test points near the design point to allow us to build a drag polar
model around this condition. For landing, the reference airspeed will be
near 1.23 VSR , and the same scope is repeated. For flaps up conditions,
drift down conditions are of interest. Beyond these expected certification
test points, one can also develop/verify the drag polar of the plane with
speeds up to VMO , up to a reasonable altitude where compressibility
effects are not significant.
For type certification (FAA/TCCA/EASA), some form of conformity
(official description of the airplane such as records of build per engineering
followed by inspection records) is required to validate that the test being
done will represent future planes to be built; without this, the applicant
may actually have to repeat the testing in order to show compliance. If the
airplane configuration changes during the flight test program after the drag
testing has been done, the applicant must either reconcile the difference to
show no impact or retest.
Beyond the conformity, one would benefit from performing an aerody-
namic inspection of the airplane prior to testing. The test airplane surface
should be cleaned, even if not painted. Items like aerodynamic seals
between control surfaces tend to degrade over time and should be replaced
prior to the testing. Flight control rigging should be rechecked.
The engines used on the test airplane should have engine serial number
specific thrust deck representing the best calibration of those engines so as to
come up with a reasonable drag model. Each engine must have a trim applied
to it so that it falls within specified limits; this trim is applied to the
mathematical models.
As part of the drag testing, one must plan for the preflight weigh. The
airplane must be in the right configuration (aerodynamic build and
onboard equipment) for the test, and this configuration must be tracked care-
fully. Daily work on the airplane during the drag test campaign can cause
surprises [e.g., weight change, aerodynamic changes, engine change including
engine controller/full authority digital engine control (FADEC) software
changes].
472 Operational Aircraft Performance and Flight Test Practices

The minimum expected instrumentation this author sees for the purpose
of performing dedicated drag testing would include:
• Pressure altitude
• Calibrated/indicated airspeed
• True airspeed
• Mach number
• Ground speed
• Static air temperature
• Airplane weight and CG [could be a fuel weight vs time with a known load
sheet (weight of airplane and crew)]
• Vertical acceleration (load factor) at the center of gravity
• Engine parameters (anything needed to derive thrust or power from the
engine)
• Propeller parameters (propeller RPM, blade pitch)
Useful parameters to track test point stability include:
• Angle of attack and angle of sideslip
• Longitudinal acceleration
• Pitch and roll attitude
• Flight control position and trim setting
• Heading
• Vertical speed

Test Monitoring
Drag testing is not the most exciting flight an aircrew can do in a flight
test program. It requires the pilot to stabilize on condition and maintain it
for a long period of time (typically at least 2 min), then proceed to the next
condition, often only a few knots faster/slower, and repeat the point.
Timely feedback to the crew, from either telemetry or a dedicated test engin-
eer on board the airplane, on the appropriateness of the test point flown will
greatly help reduce the number of repeat points and could even shorten a
given test point if all monitored parameters are stable to within the test
tolerances.
Figure 11.3 shows an example of an onboard screen that can be used by
the test crew or in telemetry to verify that the test being executed was within
the criteria provided for test condition stability. The screen is meant to
provide a quick-glance review of the condition. More parameters can be
tracked if the screen can accommodate.
One item to verify for the quality of the test condition is whether there is
any sideslip present. Sideslip means that the airplane is not in its expected
level flight drag condition, and the test will record slightly more thrust
required than expected. The test plane may not be equipped with an exper-
imental or production air data probe measuring the sideslip angle, but
 CHAPTER 11 Cruise and Drag Performance Testing 473

10,100 Tolerances

Altitude (ft) 10,090

10,080 ± 50 ft
10,070
10,060
212
Airspeed
(KCAS)

211
210 ± 2 kt
209
0.39
0.38 ± 0.005
Mach

0.37
0.36
0.35
35
Engine N1 (%) SAT (deg C)

30 ± 1°C

25
75
70 Fixed engine setting
65
60
55
0 20 40 60 80 100 120 140 160

Fig. 11.3 Example of test monitoring data screen.

the crew can still monitor the airplane sideslip by verifying that the lateral
acceleration of the airplane is “zero.” This can be done by looking at the
value of the lateral acceleration ny from an inertial reference unit or simply
by observing the slip-skid indicator on the pilot’s attitude display (see
Fig. 11.4).
Figure 11.5 is a proposed test card to track the individual test points.
This card could be appropriate for a turbofan-equipped airplane if the air-
plane is equipped with a continuously recording data acquisition system

Skid/slip
(sideslip)
indicator

Digital display Mechanical display

Fig. 11.4 Verifying for zero sideslip during drag testing.

474 Operational Aircraft Performance and Flight Test Practices

Fueled aircraft weight at engine start:

Aircraft weight at engine shutdown:
Fuel weight on board at engine start:
Test Start Altitude Airspeed Fuel remaining Temperature N1 Fuel flow Test
point time (ft) (KIAS) Start End (°C) (%) Start End time
(hh:mm:ss) (lb) (lb) (lb/hr) (lb/hr) (min:s)

Fig. 11.5 Proposed test card for cruise testing.

(captures more parameters than what is hand recorded by the flight test
engineer on this card). The card gives a postflight record for the performance
engineer to quickly find the test data. The information recorded on the card
is as provided at the flight test engineer’s station and most often is uncor-
rected data.

Posttest
As the airplane comes back from test, one must have planned for
the postflight weight before any work on the plane starts, if this is desired.
(A postflight weight is not mandatory; it is a means to validate the change
in weight during the flight so as to reduce the data reduction scatter.)
A general review of the airplane’s configuration should also be done to
recheck aerodynamic cleanliness (aero seals, paint, etc.). One should also
review the test instrumentation to make sure the data are valid.

Data Reduction: Corrected Test Weight and Center of Gravity Location

One large variable in the data analysis to create a drag model is the airplane
weight at the time of the test. As we saw in Chapter 8, weight has a direct
impact on drag, especially at low speed. In Chapter 7, we also saw the
impact of weight on the scatter of the test results for the creation of a
maximum lift coefficient. Figure 11.6 shows the typical weight trends that
could be recorded on the airplane over the course of a full flight. An expected
weight variation for an airplane burning fuel/gasoline is that the weight will go
down with time when engines are operating. Note in Fig. 11.6 that the weights
can plateau (flattening of the trend to near-zero change), fluctuate, or even
have a marked increase depending on how the airplane weight is computed
in flight.
For this reason, before one initiates dedicated testing for drag, it is rec-
ommended that the test include a preflight and postflight weighing. This
allows the performance engineer to recalibrate the weight estimate with
good starting and end points. This is still recommended even if a good,
 Pressure altitude (ft)
50
Pressure altitude

40
(ft) (103)

30
20
10
0
Mach

CHAPTER 11
0.9
0.8
0.7
0.6
Mach

0.5
0.4

Cruise and Drag Performance Testing

0.3
0.2
0.1
0.0
Weight (lb)
185
Weight plateau
180
(lb) (103)
Weight

As weighed Weight fluctuation

175
preflight? Weight growth
170
165
0 2 4 6 8 10 12
As weighed
Time (s) (103)
postflight?

Fig. 11.6 Weight recorded during a flight.

475
476 Operational Aircraft Performance and Flight Test Practices

calibrated experimental fuel flow meter is installed in the system for each
engine. The errors that can be spotted include:
• A wrong airplane configuration, including center of gravity (CG) location,
not as specified in the load sheet
• Drifting production fuel flow meters during the test or simply normal
production tolerance (can be as much as +2%)
Sometimes, especially for small airplanes, it may be your only source of
weight for the fuel used during the flight.
All electric airplanes do not have the unknown of a fuel weight change
during the test, so a preflight weighing, sometimes a pretest campaign weigh-
ing, is enough. It remains very important for both conventional and all-
electric airplanes that the airplane configuration be carefully tracked once
weighed, prior to the test completion. Very often in a test program, small
modifications keep being made to the plane as the design matures; these
will impact the weight of the plane and must be accounted for.
Another item that must be tracked during the test and adjusted by postt-
est analysis is the location of the center of gravity (CG) during the test. As
discussed in previous chapters, the CG location impacts the trim require-
ments of the airplane and thus the drag of the airplane. If the testing is
done for certification, one must perform corrections to the most adverse
CG location, which, as we saw in Chapter 7, is the most forward location
for a conventional layout (wing and aft tail). For climb (excess thrust), this
is per FAA 14 CFR Part 25, 25.117 (underline added by author):
25.117 Climb: General. Compliance with the requirements of §§25.119 and
25.121 must be shown at each weight, altitude, and ambient temperature
within the operational limits established for the airplane and with the
most unfavorable center of gravity for each configuration.

For noncertification testing, such as cruise performance, the airplane manu-

facturer may opt for a more favorable (mid) CG that would represent an
average loading of the airplane and provide the operator with a means to
correct the drag/fuel burn of the airplane when loaded more forward.

Data Reduction: Air Data Correction

The foundation of the drag model is a known calibrated airspeed.
Although most air data systems will give reasonable indication during the
early parts of a flight test program, one should definitely complete the air
data calibration campaign prior to starting dedicated drag and cruise
testing. If the system uses static source error corrections (SSECs) within
the air data computer (ADC), it reduces the crew’s workload to have the
latest calibration values in the ADC so they can target the right combination
of speed and altitude to be flown. This calibration is not required, however;
the performance engineer can apply the correction to the recorded air data
prior to the creation of the drag model.
 CHAPTER 11 Cruise and Drag Performance Testing 477

Data Reduction: Altitude Variation Correction

Although the flight crew may have done an excellent job at tracking the
altitude, there may be a small increasing/decreasing trend over the time
frame that will be used for the data collection/reduction (see Fig. 11.7).
This trend represents either an excess thrust or a thrust deficit, depending
on the gain or loss in altitude. This excess thrust is equivalent to a weight
component. (We will discuss this in more detail in Chapter 12.)
Note how we averaged the rate of climb in Fig. 11.7; in the creation of the
drag model, we are not interested in all the small fluctuations in the air data
that may be coming from atmospheric disturbances, but are more into the
long-term behavior of the test. For this same reason, the reference airspeed
for the test condition of Fig. 11.7 would be set to 132 KTAS for the data
reduction. With this averaged rate of climb, we can model this excess

10,600

10,580
Geometric altitude (ft)

10,560

10,540
Average
RCg = 34.3 fpm
10,520

10,500

10,480
135

134
Airspeed (KTAS)

133

132

131

130
0 20 40 60 80 100 120 140 160 180
Time (s)

Fig. 11.7 Small climb during a level flight test condition, good speed tracking.
478 Operational Aircraft Performance and Flight Test Practices

thrust as follows:
RCg
½DT Climb ¼ W ¼ W sinðgÞ (11:3)
V
where RCg represents the geometric rate of climb (a gain in potential energy),
V is the true airspeed of the airplane, W is the airplane weight, and g is the
airplane’s flight path. Equation (11.3) can be used to do small corrections
to an otherwise steady test condition; one must remember that every correc-
tion done to data collected represents a possible source of scatter. (Even if the
correction is meant to reduce the overall scatter of the data, a very steady
point is always best.)
To bring some numbers into the discussion about the correction, assume
the test airplane that flew the condition in Fig. 11.7 weighed 20,000 lb; delta
thrust due to the climb would be 51 lb (in excess!, this is a climb). The air-
plane in such condition had a lift-to-drag (L/D) near 10, which means the
expected drag level would be about 2000 lb. Therefore, the correction
applied to the drag is of the order of (51 lb/2000 lb) ¼ 2.5% for this
“small” recorded rate of climb—still a fairly large correction.
The flight path angle g in Eq. (11.3) will be discussed in more detail in
Chapter 12. Here, it is sufficient to say that if the rate of climb measured is
based on pressure altitude measurement, it must then be converted to the
true geometric rate of climb RCg
 
1 RCg
g ¼ sin (11:4)
V
The calibrated rate of climb RCC measured by the airplane is equal to the
change of pressure altitude with time
 
DHp
RCC ¼ (11:5)
Dt
To obtain the geometric rate of climb from the calibrated rate of climb,
one must correct the measured condition back to standard atmosphere con-
ditions as follows:
 
SAT
RCg ¼ RCC (11:6)
SATstd
where SAT is the static air temperature at the test altitude, and SATstd is
the static air temperature of the standard atmosphere at the same pressure
altitude.
Other sources of altitude can be used to do this exercise. The GPS system
provides a source of geometric altitude and can therefore be used directly in
the analysis. An inertial reference system (IRS) can also be used for change in
altitude, but because of the natural drift tendency of the position signal, the
value is often corrected with another source. One source this author saw on
 CHAPTER 11 Cruise and Drag Performance Testing 479

an airplane program was the baro-corrected inertial altitude. This altitude

source was essentially corrected with the pressure altitude source, and thus
was not a true inertial altitude. It also needed to be corrected for temperature
deviations per Eq. (11.6).

Data Reduction: Acceleration Correction

Even while flying at a constant airspeed, the aircraft may be subjected to a
change in ground speed (inertial speed, change in kinetic energy) due to a
shift in winds or excess thrust for the test condition (see Fig. 11.8). This
change in inertial speed is an acceleration/deceleration; therefore, it is a

17,020
17,010

17,000
Pressure altitude (ft)

16,990

16,980

16,970

16,960

16,950

16,940
16,930
230
228

226
224
Ground speed (kt)

222
220
218
216
214
212
210
1640 1660 1680 1700 1720 1740 1760 1780 1800
16: 1354.429 Elapse time (s)

Fig. 11.8 Stable altitude condition with small acceleration.

480 Operational Aircraft Performance and Flight Test Practices

force that must be accounted for in the analysis. This force is

 
W DVG
½DT accel ¼ (11:7)
g Dt
The expression (DVG /Dt) represents inertial acceleration during the test
segment. It is usually recorded by an INS, or it can be derived from the
change in ground speed measured by a GPS system.
Under the condition of steady winds, one can use the recorded true air-
speed in lieu of the ground speed to obtain the acceleration component;
however, there are cases where, under a slow shifting wind component, the
airplane may remain at a constant true airspeed (within test tolerance) and
yet still accelerate/decelerate. Figure 11.9 is such an example, where the
first part of the test of the test condition is at an almost constant KTAS
(within +2 kt—note that we typically use KCAS as a pilot reference) but
the ground speed is reduced by about 6 kt. The second part of the test had
a more stable ground speed, and less correction would be applied to the
test condition.

Data Reduction: Drag Polar Fitting

The next step is to perform the first set of data reduction and see how well
the combination of airframe drag and engine/propeller thrust models
matches the actual test results. For this section, we use an example. A test
airplane adopts a certain airplane configuration (flaps position, gear position,
any other system activation like anti-ice, and so on) and flight condition (see
Fig. 11.10). We used a jet airplane with the engine producing thrust. The
same exercise can be done with a propeller-driven airplane provided the
right information is available (such as engine torque, propeller RPM, propel-
ler pitch, etc.).
This is a slow flight condition with quasi-steady airspeed and altitude.
Upon review of the test data, we accept the test point as having been flown
per the test tolerance (airspeed stability, altitude stability, temperature,
etc.) provided in the test plan. From this total test time of a little less than
2 min, we identify two segments of interest. Segment 1 is 29 s long and
shows fairly stable airspeed (KIAS) and altitude hold (Hpi) as well as stable
ground speed (KTGS). The second segment includes the first one but is
longer (57 s long). Over the second segment, the altitude variation and accel-
eration are slightly longer. Doing this will give us two data points for one test
point flown (performance engineers like data).
We start by identifying the average condition for each segment, which we
insert in Table 11.1. Some data, like the fuel flows per engine, were not shown
in Fig. 11.10 but are included in Table 11.1. We note that the pilot maintained
a constant thrust setting (no change in the fan speed, N1L in Fig. 11.10);
this helps to compute the thrust of the test condition (thermally stable
engine).
 End of
test
Pressure altitude (ft) Total air temperature (°C)
Pressure altitude (ft) (103)

39.10 –16.6

Total air temperature

39.05 –16.8
–17.0
39.00

(°C)
–17.2
38.95
–17.4
38.90 –17.6
38.85 –17.8

CHAPTER 11
TAS (KTAS) Ground speed (kt)
460 608
459 606

Ground speed
TAS (KTAS)

604
458

(kt)
602

Cruise and Drag Performance Testing

457
600
456 598
455 596
TLA (deg) Fuel flow (pph)
40 2200
2000
30
TLA (deg)

1800

Fuel flow
(pph)
1600
20
1400
Less stable More stable 1200
10
More ground speed change Less ground speed change 1000
0 800
0 100 200 300 400
Time (s)

Fig. 11.9 Level flight drag point example.

481
482 Operational Aircraft Performance and Flight Test Practices

Segment 1
KIAS Hpi (ft)
135 10.22

Hpi (ft) (103)

134 10.20
133 10.18
KIAS

132 ROC = –46.3 fpm 10.16

131 10.14
ROC = –31.0 fpm
130 10.12
Segment 2
TAT (°C)
12.4
12.2
12.0
(°C)
TAT

11.8
11.6
11.4
KTGS KTAS
164
162 Accel = 0.038 kt/s
160 Accel = 0.054 kt/s
KTGS
KTAS

158
156
154
N1L (%)
66.2
N1L (%)

66.1
66.0
65.9
65.8
0 20 40 60 80 100 120
Time (s)

Fig. 11.10 Flight test example of drag reduction.

From the test data in Table 11.1, the performance engineer must apply
the first set of corrections. The recorded pressure altitude (Hpi) and indi-
cated airspeed (KIAS) are corrected to provide the proper pressure altitude,
the calibrated airspeed, and the true airspeed shown in Table 11.2. (Data
were collected during the airspeed calibration phase of flight testing, not
provided for this example; note that the recorded temperature was the
total temperature for the test.) Then the total thrust is computed using
the engine manufacturer’s thrust model for the installed engines. The per-
formance engineer also verifies that the test weight and CG were acceptable
as recorded or adjust the value (see Table 11.3). Here the CG (not shown)
was acceptable, and the test weight was corrected based on pre- and
postflight weight.

Table 11.1 Extracting Data from the Flight

Start N1 N1 Left Right

Test time Hpi Airspeed Weight Temp. L R FF FF
point (s) (ft) (KIAS) (lb) 8C (%) (%) (pph) (pph)
Seg 1 53 10,150 133.4 92,010 12 66 66 3210 3235
Seg 2 26 10,165 133 92,160 11.9 66 66 3212 3237
 CHAPTER 11
Table 11.2 Applying the First Corrections

Test Altitude Airspeed Thrust RoC ½DT climb Accel ½DT accel Final Drag

Cruise and Drag Performance Testing

Point (ft) (KCAS) KTAS (lb) (fpm) (lb) (kt/s) (lb) (lb)
1 10,130 132.8 158.3 9312 -31.0345 -178.302 0.037931 183.2815 9307.0
2 10,145 132.4 157.9 9315 -46.3158 -267.206 0.054386 263.2191 9319.0

483
484 Operational Aircraft Performance and Flight Test Practices

Table 11.3 Computing the Aerodynamic Coefficients

Test Point Weight (lb) KEAS CL CD CL/CD

1 92100 132.4 1.293 0.13069 9.90
2 92250 132.0 1.303 0.131652 9.90

For both segments 1 and 2, the airplane lost some altitude, and a geometric
rate of climb (negative for a loss of altitude)

is computed (see Table 11.2).
The equivalent thrust deficit ½DT climb for excess thrust for climb) is
computed. In both segments, the airplane accelerated slightly, and an accelera-


tion is computed (in kt/s) from which an excess thrust is computed ½DT accel .
The final drag for each test condition is computed as follows:
D ¼ TEngineDeck  ½DT climb ½DT accel (11:8)

The next step consists of reducing this collected drag information into
coefficients. For this, we compute the equivalent airspeed for each test point,
then compute a lift coefficient CL and a drag coefficient CD . Note that, for
this example, the reference wing area is 1200 ft2 . We also assumed an
average load factor of 1.0 for the maneuver because it was essentially level flight.
W
CL ¼ (11:9)
1
r V2 S
2 SL EAS
D
CD ¼ (11:10)
1
r V2 S
2 SL EAS
The next step is to collect enough test conditions to establish a relation-
ship between CL and CD , the drag polar for the configuration tested. There is
no preset number of test conditions to be flown; one must simply collect the
information, expecting test scatter and secondary effects (discussed later in
the chapter) over a range of lift coefficients of interest (e.g., enough to
cover the takeoff performance of the airplane). Once the performance engin-
eer believes he or she has collected and analyzed enough data, he or she can
create a chart of CD vs CL 2 (see Fig. 11.11). If the data fit well, a linear trend
can be extrapolated from them. In this case, the slope of the linear trend is the
induced drag factor K, and the intercept corresponds to the parasite drag
coefficient CDo .
Notice that the data may deviate away from the linear fit on either side of
a given range. The performance engineer must therefore set the proper oper-
ating range (the lift coefficient range where the data will be used) to fit the
intent of the model. For a takeoff performance model (takeoff flaps), as we
will see in Chapter 15, the expected speed range of interest will be near
1.13VSR (1.2VS ); then the lift coefficient range tested should be such that
 CHAPTER 11 Cruise and Drag Performance Testing 485

0.07

0.06

0.05

0.04
y = 0.0407x + 0.0163
CD

0.03

0.02
Operating range
0.01

0
0 0.2 0.4 0.6 0.8 1
CL2

Fig. 11.11 Fitting of flight test drag data, CDo ¼ 0.0163, K ¼ 0.0407.

one gets enough airspeed coverage, say from 1.08VSR to 1.4VSR , to properly
predict the performance of the airplane. Note that the drag model created
cannot typically be used in isolation; it is an excess thrust model that is
paired with a specific engine model.
Figure 11.12 shows an example of a certification drag model for various
flaps and gear positions. Note that when the data fit well, an average line is

0.2500
y = 0.0332x + 0.1357
F0 Gear Up
F30 GD
F08 Gear Up
F15 Gear Up
0.2000 F15 Gear Down
F08 Gear Down y = 0.0348x + 0.1005
F30 Gear Down F15 GD

y = 0.033x + 0.0908
F08 GD y = 0.0457x + 0.0528
0.1500 F15 GU
CD

1.3Vs

y = 0.0438x + 0.037
0.1000 F08 GU

1.3Vs
0.0500
y = 0.0447x + 0.0253
F0 GU

0.0000
0 0.5 1 1.5 2 2.5 3 3.5 4
CL2

Fig. 11.12 Flight test data fit.

486 Operational Aircraft Performance and Flight Test Practices

drawn through the test points. In the case of the flaps at 30 deg and the gear-
down condition, the performance engineer elected to draw a conservative
(higher drag) fit. This is usually done in the presence of large data scatter
or because of uncertainty during the testing and follow-on verification.
From a certification point of view, the drag of the airplane takes on the
form as identified by the equations of Fig. 11.12.

Data Reduction: Power Effects

As the drag curve is produced, the performance engineer may notice that
the fit may not be optimal. The drag model may or may not include a thrust
vector (the previous model does not), but a thrust vector may not be enough
to make the data fit the drag polar model. The exercise to create the final low-
speed drag model started with a drag model and a thrust model derived from
preflight test analysis. The models may have been adjusted by wind tunnel
testing and/or CFD, and the engine model may even have flown on a
flying test bed.
If there is such a disagreement, one can start comparing not only the CD
to CL relationship (the drag polar), but also the a-CL -CD relationship as a
function of thrust input (or power input). We define a thrust coefficient as
T
CT ¼ (11:11)
1
r sV 2 S
2 SL
Wing-mounted engines, those with propellers in particular, will impact
the production of lift more than a simple thrust vector component, especially
with flaps deployed. For a given lift coefficient CL based on Eq. (11.9), one will
note a shift in the plane’s angle of attack at different values of CT (Fig. 11.13).
There will be an impact on the drag coefficient; in addition, for example, the
fuselage will be presented at a different angle to the airflow for a given CL .

CT1
CL CT3 > CT2 > CT1
V
CT3
CT2
CT1 CT2
V

CT3
V

Fig. 11.13 Effects of thrust generation on the aerodynamics of the airplane.

 CHAPTER 11 Cruise and Drag Performance Testing 487

9.75
(32)

7.06
(27.8)

22.40 28.42
(73.5) (93.25)
Dimensions in m (ft)

Fig. 11.14 NASA QSRA.

If the impact on the drag polar is strong enough, the performance engineer
may want to include such effects in the modeling of the airplane’s drag.
One airplane where this effect was notable was the NASA Quiet
Short-Haul Research Aircraft (QSRA; see Fig. 11.14). The airplane was a
de Havilland DHC-5 Buffalo with a modified wing and four turbofan
engines located on top of the wing to provide a Coanda effect using upper
surface blowing (USB).
Another factor that can influence the drag model with thrust production
is the location of the thrust line vertically and its impact on the airplane trim.
The further away from the center of gravity, the greater the impact on the
trim drag.
In general, this author has observed that aft fuselage–mounted podded
engines have the least impact on the overall aerodynamics of the airplane,
whereas wing-mounted propellers had the greatest impact. In the end, the
performance engineers decide how best to integrate the thrust model into
the drag model with an overall goal of best predicting the excess thrust capa-
bility of the airplane.

Testing for High-Speed Drag

Business jets and jetliners regularly cruise near or into the transonic
Mach region where significant Mach effects (compressibility and drag rise)
are present. Testing for high-speed drag is not fundamentally different
from the low-speed drag testing methodology described earlier. But if the
applicant wants to limit the scope of testing to be done to a reasonable
level, a slightly different approach must be used.
488 Operational Aircraft Performance and Flight Test Practices

The following is an alternative approach to doing drag testing to address

the issue of the drag rise. As discussed in Chapter 8, when the airplane passes
the critical Mach number (where the airflow near the airplane’s surface first
becomes supersonic), small shock waves will start to form on the airplane. As
the air flows through a shock wave, there is a sudden increase in pressure and
a reduction in the airspeed to a subsonic Mach number. This results in a loss
of energy in the flow and often leads to flow separation behind the shock. For
a given angle of attack, as the Mach number increases and the shock waves
become stronger, the result will be a loss of lift coefficient and a sharp
increase in the drag coefficient.
If one were to fly at a constant lift coefficient while increasing the Mach
number, it could be noticed that there would exist a Mach number where the
drag coefficient would start to rise. The baseline drag model states that, for a
given lift coefficient, the drag coefficient is constant

CD ¼ CDo þ K CL2 (11:12)

The Mach number at which this occurs is the Mach drag rise. Some compa-
nies define the Mach drag rise as the point where the drag coefficient has
increased by approximately 20 counts over the incompressible drag coeffi-
cient. If one were to repeat the test at a different lift coefficient, it could be
noticed that the Mach for drag rise would be different (see Fig. 8.10 in
Chapter 8). In fact, it decreases with increasing lift coefficient.
It can be quite difficult to try to fit an equation through the data collected
during high-speed testing (in the same way that was done for the low-speed
testing) that would cover all expected combinations of weight, altitude, and
Mach number an airplane is expected to see in service. The drag coefficient
is usually well behaved (predictable variation) for a given lift coefficient, so
instead of coming up with an equation, it may be easier to come up with a
series of lookup tables of CD vs CL for a given Mach number and to interp-
olate between the different tables.
Again, due to the enormous number of combinations of weight, altitude,
and Mach number an airplane is expected to operate at, without a proper test
methodology, the cost of developing a good cruise model for a manufacturer
could be prohibitive. Knowing that the drag coefficient is relatively well
behaved as a function of Mach number for a constant lift coefficient, one
approach that has been used to build a high-speed drag model is to fly at con-
stant W/d values.
W 1
¼ g pSL M 2 S CL (11:13)
d 2
The manufacturer may then select typical W/d that the airplane is
expected to fly at and build a drag model around these points. Note that
this is still a long process, although reduced. Figure 11.15 shows an
example in which a given airplane has possibly three zones of interest for
 W/␦
45,000
320K MMO = 0.82
40,000 280K 40,000

240K
35,000 Long range 35,000 LR
200K
30,000 180K 30,000

CHAPTER 11
160K CAS
(kts)
100

Pressure altitude (ft)

Altitude (ft)

25,000 140K Short/medium range 25,000

SR / MR
150
20,000 120K 20,000 200
250

Cruise and Drag Performance Testing

100K 300
15,000 Holding 15,000 H 350

80K
10,000 10,000
VMO = 320 KCAS

5000 5000
60K

0 0
50,000 60,000 70,000 80,000 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Max weight at altitude (lb) Mach Number

Fig. 11.15 Cruise test zone optimization.

489
490 Operational Aircraft Performance and Flight Test Practices

high-speed drag. One is the long-range cruise condition where the airplane
typically flies higher and the cruise speed may be Mach driven (compressibil-
ity effects). Then there is a short-to-medium-range envelope where the air-
plane may cruise at a lower altitude and the cruise speed is driven by
calibrated airspeed. Finally, there could be a holding envelope where the air-
plane flies near minimum drag conditions. One can optimize the test speed
condition for a given W/d range.
For a new model airplane, this author likes to test a minimum of 5 W/d in
the long-range cruise condition and extend this to one or two more W/d to
include the short/medium-range conditions. The holding zone is typically
flown at lower altitudes and near minimum drag, so a simple check/adjust-
ment of the low-speed drag model may be necessary. For the long-range
segment of the cruise testing, we recommend adding smaller speed/Mach
increments in the region of expected drag rise to better define the rapid
change in drag. For a derivative airplane where a previous cruise model
had been established and verified by flight test, this author believes that the
testing can be brought down to 3 W/d in the long-range cruise section.
Based on the results of the flight test, the corrections identified can be
applied to the rest of the envelope.
With the test scope identified, the testing is executed, and the data
reduced in the same way as was done for the low-speed drag model discussed
earlier in this chapter. If the manufacturer reduces the data for test points
done at constant W/d and expresses the results in terms of D/d, then one
can expect to obtain typical drag curves, as shown in Fig. 11.16.

D 1
¼ g pSL M2 S CD (11:14)
d 2

We must add that these curves are obtained while flying at a constant CG.
If the CG position cannot be controlled in flight, the result will be an increase

(a) (b)
D/␦ CL Low speed model
M1
M2
M3
M4
W/␦
M5
M6

Increasing mach
Test result
Model

M CD

Fig. 11.16 Curve fitting flight test data.

 CHAPTER 11 Cruise and Drag Performance Testing 491

CD
CL8 CL7 CL6 CL5
Increasing CL
CL4

CL3

Normal operating zone CL2

CL1

Fig. 11.17 Final verification of high-speed drag model.

in the data scatter, especially in the drag rise region where fractions of angle
of attack can introduce significant drag changes.
If we look at Fig. 11.16a, the circles represent the test conditions collected
for which we have performed our data reduction (the grouping being per
W/d). Then the performance engineer creates a drag model by fitting
curves through the data. The curves that extend below the lowest Mach or
higher than the highest Mach tested are extrapolations; the test airplane is
not necessarily able to reach all corners, limited either by capability (e.g.,
test day too warm) or by test scope (cost of execution). Two of the lines
have no test conditions; they may have been derived from the low-speed
drag model. From these curves (D/d vs Mach for given W/d), the manufac-
turer can then establish curves of CL vs CD for constant Mach numbers
(Fig. 11.16b) and verify that the data still look reasonable. Are the curves
behaving as expected? If the curves turn out to be too wavy, adjust the
curve fitting model to account for the test conditions. There is natural test
scatter in the test results; if the curve on the D/d vs Mach curves deviates
too much from the analysis, is it worth performing more testing? Were the
test points unstable, which may explain deviation from the curve fit? Were
enough data collected in the drag rise region to best define the curves?
The goal of the exercise is to have a model that can also be represented by
curves of CD vs Mach for a constant lift coefficient CL (see Fig. 11.17). That
model must have a variation that is as expected (i.e., no waviness or disconti-
nuities, a behavior similar to Fig. 8.10 in Chapter 8). Again, carefully look at
the behavior of the drag rise compared to the data collected and compared to
the expectation—more testing may be needed. The focus should be on the
normal expected operating zone; deviation outside this zone may not justify
492 Operational Aircraft Performance and Flight Test Practices

the additional flight test cost. Once satisfied with the model, it can then be
broken down into lookup tables and used with a specific engine model for
cruise performance analysis.

Test Monitoring
Just like the testers did for low-speed drag testing, some acceptable test
tolerances are used for high-speed drag testing. The following test limits
have usually provided a good compromise between execution time and
data quality for high-speed drag testing:
• Maximum airspeed variation during the test: + 2 KIAS (ideally within 1 kt).
• Mach number: +0.005.
• Maximum altitude variation: + 50 ft.
• W/d at the start of the test condition: +1% of the target.
• Maximum temperature variation: + 18C.
• Minimum test time: 2 min or one phugoid period.
• The pilot should never touch the throttle during the test. We need the
engine thermodynamics to remain stable during the execution of the test.
• Desired propeller settings if applicable (blade pitch, engine torque, etc.).
• Center of gravity: +1% MAC of target.
The flight crew may find it more difficult to maintain these tight tolerances
while the aircraft is flying near the minimum drag conditions. For this con-
dition, a shorter period of time could be considered, with the risk of increas-
ing the scatter during data reduction.
The minimum expected instrumentation this author sees for the purpose
of performing dedicated high-speed drag testing is similar to that required for
low-speed drag testing.

Data Reduction: Reynolds Number Effects

For Part 25 and Part 23 airplanes, low-speed drag testing is typically done
between 5000 ft and 15,000 ft. The impact of the different altitude/tempera-
ture combinations (impact on boundary layer transition and laminar vs tur-
bulent flow coverage of the airframe) are low enough that the model remains
valid, within the typical test scatter for all altitudes. Testing is typically done
within one altitude test zone with temperature of the day. This helps to
reduce the scope of the test (flight test costs).
Commercial airliners tend to fly over a large range of altitudes (20,000
to 41,000 ft), and business jets can extend this to 51,000 ft. This large altitude
range will have a greater impact on the cruise performance of the airplane
and will introduce further scatter into the drag model. One can compare
the test condition Reynolds number [Eq. (11.15)] against a reference
(design cruise condition) Reynolds number to apply corrections (see
Fig. 11.18). The reference length for the Reynolds number can be the mean
 CHAPTER 11 Cruise and Drag Performance Testing 493

0.008
⌬CD

0.007

0.006

0.005 (Re)ref Re
Full
y tu
rbul
0.004 ent
Cf

bou
nda
ry la
0.003 yer
one
0.002 Lam nz
inar si tio
bou
nda T ran
0.001 ry la
yer
0
1.00E+05 1.00E+06 1.00E+07 1.00E+08
Re

Fig. 11.18 Reynolds number correction for high-speed drag.

aerodynamic chord (MAC) of the airplane, and the reference condition could
be, for a commercial airliner, Mach 0.78 at 37,000 ft under standard con-
ditions, for example.
rVl
Re ¼ (11:15)
m
To introduce numbers into the discussion, say a business jet with a MAC
of 12 ft can fly at Mach 0.80 from 29,000 ft to 51,000 ft. One can then expect
that the Reynolds number would vary from approximately 1.1  107 to about
2.8  107 . With this expected variation in Re, the manufacturer can establish
a preflight test CFD or wind tunnel analysis expected correction curve (see
Fig. 11.18). To validate correction, the test team can fly the same W/d at
two different weights (light weight, high altitude and heavy weight, lower
altitude).

Testing for SAR

Testing for specific air range (SAR) implies not only excess thrust testing
(matching airframe drag to engine thrust models), but also energy use vs time
and distance. The ultimate goal in cruise performance is to be able to predict
the aircraft fuel burn (energy use) over a given mission. We have already
established our low- and high-speed drag models (that are paired with an
engine model), and we are now required to be able to match the fuel burn
with these models.
Remember that the approach used is to match the engine manufacturer’s
engine deck prediction in thrust and fuel flow with our aircraft drag model so
494 Operational Aircraft Performance and Flight Test Practices

as to come up with an overall model that is representative of the aircraft’s

capabilities. The SAR is the ability of the airplane to cover a given air distance
(no effects of winds) for a given amount of fuel.

dXa TAS TAS

SAR ¼ ¼ ¼ (11:16)
dWf Ẇ f SFCT T

The SAR can therefore be measured easily in flight by simply recording

the true airspeed at a point and the corresponding fuel flow. In order to
reduce the scatter during the data reduction, the same approach as was
used for high-speed drag is advised (flying W/d) for test tolerances.
Each test point should be a minimum of 2 min or at least one phugoid
period, whichever is longer. A shorter period of time could also be considered
while flying near minimum drag speeds, with again the risk of increasing the
scatter during data reduction. The engine should be thermally stable, and
manufacturers often start test conditions for a given W/d at the higher
Mach number (more thrust), and then slow down. This may not make a
big difference if the testing starts soon after the climb where engine thrust
was higher.
During the “stabilized” point, there will inevitably be a small variation in
altitude and airspeed/ground speed. This excess/deficit in energy as com-
pared to a stable airspeed and altitude condition will require more/less
energy from the engines and therefore a higher/lower fuel consumption.
As we did for drag testing, one should correct the measured drag of the
test point to a steady condition [Eq. (11.8)]. Then, one can use the engine
manufacturer’s engine deck to determine the fuel flow correction for the
corrected thrust.
The quantity of fuel burn (as shown to the crew by the fuel flow meter or
the fuel quantity indicator) is usually a measured fuel volume or fuel volu-
metric flow converted to a fuel weight by using a standard fuel density (or
a fuel density corrected for temperature if a fuel computer is installed). It
is therefore advisable to measure the density of the fuel loaded on the aircraft
prior to the cruise performance flight test. The usual units of measurement
for fuel density are the degrees API (8API), where API is the acronym for
American Petroleum Institute:
141:5
RDð608FÞ ¼ (11:17)
8API þ 131:5
RD(608F) is the relative density (specific gravity) of the fuel compared to
that of water at 608F (water has a density of 0.9990 g/cm3 or 8.3378 lb/U.S.
gal at 608F ). As an example, if one were to measure the 8API of water
(specific gravity of 1.0), then you would get the following answer:
141:5
8API ¼  131:5 ¼ 108API
RDð608FÞ
 CHAPTER 11 Cruise and Drag Performance Testing 495

A liquid with a higher 8API than 10 has a lower density than water. Jet A
fuel has a specific gravity of 0.815 (about 428API); the range of acceptable
values for Jet A (and Jet A-1) falls between 37 and 51. Jet A/A-1 fuels, com-
monly used in North America and Europe, have a typical density of 6.76 lb
per U.S. gal at 608F (158C) (see Fig. 5.17 in Chapter 5).
It is also advisable to measure the weight of the fueled airplane prior to
the flight and reweigh the airplane upon engine shutdown. In this way, one
is actually measuring the actual fuel weight burned as compared to the com-
puted fuel burn from fuel flow meters. This total fuel burn can be compared
with the computed fuel burn by the instrumentation, and then the ratio can
be used to correct the fuel flow indication at a given point in the mission.
An engine manufacturer will guarantee a minimum installed thrust
specific fuel consumption (SFCT ) for a given altitude, temperature, Mach
number, and N1/thrust. This SFCT is based on the use of a fuel that has
at least a minimum low heat value (LHV). The LHV is the energy content
of the fuel; the usual units are BTU/lb or MJ/kg. The LHV of Jet A fuel typi-
cally ranges from 18,400 to 18,650 BTU/lb. Prior to a flight, a sample of the
fuel should be collected and sent to a lab for determination of its LHV.
An LHV that is lower than the LHV used by the engine manufacturer will
result in a higher fuel consumption than guaranteed. The measured fuel flow
should then be corrected as follows:
LHVmeasured
Ẇ f ,ref ¼ Ẇ f ,measured (11:18)
LVHref
As an example, consider a batch of fuel used for a test that had 18,400
BTU/lb instead of a reference condition used by the engine manufacturer
to “guarantee” an SFCT of 18,550 BTU/lb. This would mean that the airplane
would burn 0.8% more fuel (ratio of 18,550 to 18,400 BTU/lb) to produce the
same thrust.

Data Reduction: Secondary Effects, Power Leakage

Testing for SAR means quantifying the use of onboard energy during the
flight, not just the production of thrust. When the test team sets up for a test
condition, all sources of power drains must be known. What is the engine
bleed setting? (How much bleed air is extracted from the engine, which trans-
lates into additional fuel burn to maintain thrust?) How much electrical
power is extracted to support the systems on the airplane? (Experimental
equipment and flight test engineer stations may extract more power than
normal flight.) How much hydraulic power is used? Is it representative of
in-service use? Is the APU on? (This burns fuel even if not providing
power to the airplane.) That APU may have an inlet door that opens
(creates more drag) into the airflow. These power drains all represent
leakage of the energy available for the airplane (fuel weight) and must be
quantified and compared to the operational scenario expected.
496 Operational Aircraft Performance and Flight Test Practices

These energy drains may represent a small fraction of the energy expen-
diture of a conventional jet fuel/gasoline-powered airplane because electric
generators are very efficient, and the electric power extraction may have
little impact on the engine’s ability to produce the required thrust for
flight. It is not necessarily the case for electrically powered airplanes,
however, where the powertrain is much more efficient and the same power
extraction to feed the onboard systems has a bigger percentage impact on
the total energy use. The test team must carefully map out the source of
energy drains to properly define the airplane’s cruise performance.

Flight Testing for SAR on Electric Airplanes

On conventional airplanes, the energy source is from fuel/gasoline;
measuring the fuel flow gives you a good indication of the energy being
used and where it is going, the engine. The power extraction from the
engines for the airplane’s systems is relatively small and has relatively little
impact on engine efficiency. It still needs to be measured, however, to
compare to the expected use in service.
For battery-powered airplanes, the efficiency of the overall system is
much higher, and the power extracted from the system (see Fig. 11.19) to
feed the airplane’s systems will be relatively higher. The cruise performance
model depends on understanding the efficiency of the powertrain system
from battery to propeller, but the efficiency of the battery, and its effective
capacity, varies depending on the power extraction. The higher the power
extraction, the greater the impact on efficiency and effective capacity. To
create a good performance model, one must map all power extraction
carefully.

Motor model =
Power to
f(torque, rpm,
airplane
voltage, and
systems
current)
Measure Propeller
power 85% eff.

Battery pack DC Controller AC Electric motor

0.30 kWh/kg 99.5% eff. 97% eff.

Wire power Wire power

drop model drop model
Propeller model =
f (torque, rpm,
blade pitch)

Fig. 11.19 Mapping the power draw on the electrical system.

 CHAPTER 11 Cruise and Drag Performance Testing 497

1
⌬SAR (%)

–1

–2

–3

–4

–5
0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90
Mach

Fig. 11.20 Integrated model check.

Integrated Model Verification

The data collected in-flight must be compared against the integrated
model (airframe drag-engine thrust-engine fuel burn) values to verify its accu-
racy. One clear way to perform this exercise is to compare the measured (and
corrected) SAR to the model SAR for the same flight condition (see Fig. 11.20).
SARtest  SARmodel
DSAR ¼ (11:19)
SARtest
A model that falls within 1% accuracy is typically a good one, given the
variability of flight test data collection. When representing the data in this
fashion, some flight conditions (such as the region of drag rise) that may
not fit as well will stand out. The performance engineer may then decide
to add additional correction factors, if required.

Validation Flight
The ultimate demonstration of the validity of a thrust-drag-fuel burn
integrated model is to compare the output to a few validation flights. Vali-
dation flights consist of executing a mission from engine start to engine shut-
down along representative long-range (stepped cruise, if possible) flights.
The manufacturer will select a few typical cruise missions, such as the
design maximum range with the design payload, with specific climb and
descent schedule, design cruise Mach number, and land with (hopefully)
the design reserve fuel. The analysis of the flight can then reveal the actual
fuel burn for various conditions of thrust, airspeed, altitude, temperature,
and winds (see Fig. 11.21). The mission is broken down into small sections,
 498
Altitude (ft) (103)

Operational Aircraft Performance and Flight Test Practices

SAT (˚C)
50 40
Cruise
(ft) (103) 40 20
Altitude

30 0

SAT
(˚C)
20 –20

10 –40

0 –60

KCAS (kt) Mach

350 0.9
300 0.8
0.7
250
0.6
KCAS (kt)

200

Mach
0.5
150 0.4
Descent 0.3
100
0.2
50 Taxi
0.1
Taxi
0 0
Fuel (lb) Engines FF (pph)
28 14
12

Shutdown
26 10

Engines FF
(pph) (103)
(lb) (103)

Engine start

8
Fuel

24
Climb 6
thrust 4
22
2
20 0
0 1000 2000 3000 4000 5000 Land
TO Time (s)

Fig. 11.21 Flight test example of a validation flight.

 CHAPTER 11 Cruise and Drag Performance Testing 499

Altitude

TOW Time
Weight

Actual

Model
LW

Fig. 11.22 Example of a validation flight.

and each section compares the specific performance. Then the sections are
assembled to see the actual fuel burn vs the model (see Fig. 11.22). It may
be necessary to add an adjustment factor to the model to best fit the data.
So as not to adversely affect the model, the adjustment factor (possibly a
fuel flow factor) should be derived by more than one validation flight (to
account for scatter in the data).

Instrumentation for SAR Testing

The instrumentation required for SAR testing is similar to that required
for drag testing plus anything required to measure the energy use of the air-
plane. One must carefully record fuel flow and/or battery power extraction,
the airplane’s system power use, bleed extraction, and so forth. One must
capture all leakages.

Testing for Zero Fuel Level

There are no certification requirements for cruise performance (fuel
burn), only operational requirements on fuel management and fuel reserves.
To satisfy the operational requirements, one needs to be able to reliably
predict remaining available fuel quantity. There is always a certain amount
of fuel in the fuel tanks that will not be reliably available. The certification
requirements of §25.959 define the unusable fuel supply:
500 Operational Aircraft Performance and Flight Test Practices

§25.959 Unusable fuel supply: The unusable fuel quantity for each fuel tank
and its fuel system components must be established at not less than the
quantity at which the first evidence of engine malfunction occurs under
the most adverse fuel feed condition for all intended operations and flight
maneuvers involving fuel feeding from that tank. Fuel system component
failures need not be considered.

The manufacturer must determine the condition(s) of most adverse fuel

supply to the engines and perform flight testing to validate. The guidance
material of FAA AC 25-7D suggests:
• A fuel tank that is not designed to feed the engines under all flight
conditions needs to be tested only for the flight regime for which it is
designed (e.g., cruise conditions).
• The fuel system and tank geometry should be analyzed to determine the
critical conditions for the specific tanks being considered.
• After the most adverse fuel feed condition and the critical flight attitude
have been determined for the specific fuel tanks being considered, the
appropriate flight tests should be conducted.
W Steady state sideslips anticipated during operation with the airplane in

both the approach and landing configurations should be determined.

W A go-around condition at maximum acceleration and maximum rotation

rate to the maximum pitch attitude should be considered.

W Effects of turbulence on unusable fuel quantity should be considered.

The testing involves performing maneuvers until one engine on the air-
plane goes into starvation. This does make the testing interesting because
many Part 25 airplanes feed the engines from the respective side wing (left
engine fed by left wing fuel and right engine by right wing fuel). Performing
the testing may lead to a large lateral fuel imbalance if the fuel system is not
properly configured for the testing (possibly some special modification that
would still make the fuel feed representative of the production plane). The
testing will most likely lead to one engine flaming out and thus a possible
large thrust asymmetry. The testing should be done with the same care as
one would take during the execution of minimum control test for §25.149.
Once the highest level of unusable fuel is validated by test, the zero fuel
indication is set to this value in the fuel quantity system. Any fuel below
this value may not be available for continued flight. Any fuel in the tanks
below the zero fuel line cannot be used to plan the mission and becomes
part of the airplane’s empty weight.

Testing for Zero Energy on Battery-Powered Airplanes

The aviation industry is familiar with the use of fuel/gasoline as a source
of energy. Pilots and ground maintenance personnel deal with liquids,
volume, and weight as a means to understand the amount of energy left
for the flight. With batteries (we will refer to the lithium-ion batteries at
 CHAPTER 11 Cruise and Drag Performance Testing 501

Discharge curve Discharge rate Temperature impact

4.0 4.0 4.0

+55°C
Cell voltage (V)

Voltage (V)

Voltage (V)
+20°C
3.5 3.5 3.5
–20°C
Increasing rate
3.0 3.0 3.0

20 40 60 80 100 Capacity (Ah) Discharge time (h)

Percent capacity discharged

Fig. 11.23 Battery performance characteristics.

this stage as being the current state of the art), one needs to consider the
operating temperature of the batteries and the discharge rate to get an under-
standing of the available energy (the effective capacity of the battery). The
typical discharge curve of a lithium-ion battery is similar to the chart in
Fig. 11.23a. The cells will have a relatively flat change in voltage (a means
to compute energy available) vs capacity, so one cannot rely only on
measured voltage to understand the capacity left—one must also keep
track of the power used [P ¼ I V, Eq. (8.77)]. Temperature impacts the
internal resistance of the batteries, and lithium-ion batteries suffer plating
of the anode at very cold temperatures, which leads to a permanent reduction
in capacity (see Fig. 11.23b). The rate at which batteries are discharged also
impacts the effective capacity, with higher discharge rates leading to
lower capacity.
The performance engineer will need to fully characterize the batteries
used in the airplane to be capable of reliably predicting the amount of
energy available to complete the mission with the proper reserves. The
testing to determine the zero energy available capacity will most likely
involve a high power extraction near the end of the capacity curve and
seeing when the zero is declared based on the ability to provide useful
power to the engine. (As the voltage decreases, the current must increase
to provide the same power, leading to an increasing discharge rate.) Once
this is well understood, a clear indication must be provided to the crew
(similar to a fuel level).
Another factor impacting the batteries is the depth of discharge (how
close you get to zero energy). At a given discharge rate and temperature,
the amount of active chemicals transformed in the battery with each
charge will be proportional to the depth of discharge (see Fig. 11.24). To
be useful, a battery must maintain an effective capacity that will allow for a
mission to be planned. If one were to always use the batteries to the 100%
discharge point, the batteries would have a very short life (number of
cycles) and would need to be replaced (maintenance cost). The less the
depth of discharge per flight, the greater the battery life, so limiting the dis-
charge to no more than 80% on average would allow for a reasonable life
502 Operational Aircraft Performance and Flight Test Practices

Cycle life Depth of discharge vs cycle life

5.0 1,00,000

Expected cycles
10,000
Capacity (Ah)

4.0

3.0 1000

2.0 100

0 1000 2000 Depth of discharge (%)

Number of cycles

Fig. 11.24 Battery life vs average discharge.

(1000 or so cycles). The positive aspect to this is that most early all-electric
airplanes will have a capability of only a little over one hour. Combining
this one operational hour with the reserves of the operating rules (addition
of alternate and 30 to 45 min air time beyond the time required to get to des-
tination) effectively means the batteries on these planes should not normally
be used to more than 80% of capacity.
The aviation industry will need to define new standards for energy avail-
able and state of charge that the pilot community will readily understand and
be able to use to plan flights in the same manner they can now with fuel.

Presenting the Information to the Flight Crew

For the purpose of planning a long and segmented mission (all phases of
flight with fuel/energy management, weight, and CG variation), it is this
author’s belief that a table format is easier for the flight crew to navigate
and reduces the risk of reading mistakes. A chart format is also good for
trends and rapid viewing of certain parameters. Using tables may end up esti-
mating a little more fuel required than the minimum required for the mission
and it may cause some payload to be removed, but it is always safer to have
more fuel than required. If the crew actually uses a mission computer
program, a more exact estimate of fuel required for the planned mission
can be obtained, but who hasn’t had a weather surprise en route?
For flight planning purposes, the tables that are provided to the flight
crew should include at a minimum:
• Temperature at altitude
• Weight and CG for the condition
• Power setting and resulting fuel flow (or energy consumption)
• Airspeed/Mach for the condition
Complimentary to the table, the manufacturer must provide a means to
compute KTAS from airspeed/Mach (or it can be included in the table, as
 CHAPTER 11 Cruise and Drag Performance Testing 503

is often the case), and provide a means for the operator to correct the com-
puted fuel flow for any deviation in CG from the one on the chart. (This could
be in the form of a percent change in fuel burn.)
In the following sections, we offer examples of industry-preferred table
and chart formats to allow crews to plan for a flight with the proper fuel
and fuel reserves.

Cruise Altitude Selection

In order to determine the quantity of fuel needed to perform a mission,
the pilot must select the desired cruise altitude. For most planes, as discussed
in Chapter 10, this is often the highest altitude that can be sustained in level
flight at the desired cruise airspeed/Mach. One way to present the general
capability of the airplane is to use charts as shown in Fig. 11.25. This chart
provides trends that give the pilot a chance to understand the airplane’s capa-
bility at altitude at the desired cruise speed. It also provides a clear under-
standing of the effects of thrust and weight in selecting a cruise altitude.
The chart can provide the cruise altitude for best fuel economy (best SAR),
the thrust limited altitude (equivalent to a given rate of climb), and the alti-
tude where the airplane is expected to have a given maneuver margin (typi-
cally 1.3 g, see Fig. 14.31 in Chapter 14) at the discretion of the manufacturer.
That information is provided in the chart comments section. The pilot must
then select an allowable altitude based on cruise altitude restrictions.

SAR Charts
Once the cruise altitude is selected, the crew must be provided with
enough information to plan the fuel burn during the cruise. This can, of

Chart 4–5: Cruise ceiling – Mach 0.78

49,000
MACH 0.78
48,000 MCT Setting
25% C.G.
47,000 WAI OFF Cert limit
46,000 ISA
– 10
45,000 °C
Th 1.3
44,000 ISA ru
st gt
Be lim ob
Cruise altitude (ft)

43,000 st it ( uff
fue et
42,000 ISA le 20
+1
0°C co 0f
41,000 no pm
my )
40,000

39,000

38,000

37,000 ISA
+ 20°C
36,000

35,000
9500 10,000 10,500 11,00011,500 12,000 12,500 13,000 13,500 Weight
Cruise weight (lb)

Fig. 11.25 Selecting the cruise altitude.

 504
Operational Aircraft Performance and Flight Test Practices
Constant Speed Cruise – 41,000 FT TWIN ENGINE CRUISE
0.85 MT ISA
CRUISE WEIGHT KTAS GROSS WEIGHT – 1000 LB
ALT OAT (°C) 78 74 70 66 62 58 54 50 46 42
CRUISE 9000 10,000 11,000 12,000 13,000 13,500
45,000 MT ***** ***** ***** ***** ***** ***** ***** ***** 0.85 0.85
MACH TEMP TAS FF SAR FF SAR FF SAR FF SAR FF SAR FF SAR
487.5 KCAS ***** ***** ***** ***** ***** ***** ***** ***** 231 231
KIAS (°C) (KTS) (PPH) (NM/LB) (PPH) (NM/LB) (PPH) (NM/LB) (PPH) (NM/LB) (PPH) (NM/LB) (PPH) (NM/LB) FF ***** ***** ***** ***** ***** ***** ***** ***** 2844 2628
ISA – 10°C 392 585 0.671 610 0.643 638 0.615 663 0.592 693 0.566 706 0.555 –56.5 NAM / LB ***** ***** ***** ***** ***** ***** ***** ***** 0.1714 0.1855
43,000 MT ***** ***** ***** ***** ***** ***** ***** 0.85 0.85 0.85
0.70 ISA 401 601 0.668 627 0.640 656 0.612 682 0.589 714 0.562 727 0.552
487.5 KCAS ***** ***** ***** ***** ***** ***** ***** 242 242 242
204 ISA + 10°C 411 618 0.664 645 0.637 675 0.608 702 0.585 735 0.559 749 0.548 FF ***** ***** ***** ***** ***** ***** ***** 3060 2853 2788
ISA + 20°C 420 636 0.660 663 0.633 694 0.605 721 0.582 –56.5 NAM / LB ***** ***** ***** ***** ***** ***** ***** 0.1593 0.1709 0.1749
41,000 MT ***** ***** ***** ***** ***** 0.85 0.85 0.85 0.85 0.85
ISA – 10°C 403 614 0.656 635 0.635 663 0.609 689 0.585 718 0.562 733 0.551
487.5 KCAS ***** ***** ***** ***** ***** 253 253 253 253 253
0.72 ISA 413 632 0.654 653 0.632 682 0.605 710 0.582 740 0.558 755 0.547 FF ***** ***** ***** ***** ***** 3528 3272 3093 3024 2964
211 ISA + 10°C 422 650 0.650 672 0.629 702 0.602 731 0.578 762 0.554 777 0.543 –56.5 NAM / LB ***** ***** ***** ***** ***** 0.1382 0.149 0.1576 0.1612 0.1645
39,000 MT ***** ***** ***** 0.85 0.85 0.85 0.85 0.85 0.85 0.85
ISA + 20°C 432 668 0.647 690 0.626 721 0.599
487.5 KCAS ***** ***** ***** 265 265 265 265 265 265 265
ISA – 10°C 415 645 0.643 664 0.624 689 0.601 717 0.578 742 0.559 756 0.548 FF ***** ***** ***** 3986 3713 3492 3344 3276 3216 3176
0.74 ISA 424 663 0.640 684 0.621 710 0.598 739 0.575 765 0.555 780 0.544 –56.5 NAM / LB ***** ***** ***** 0.1223 0.1313 0.1396 0.1458 0.1488 0.1516 0.1535
37,000 MT ***** ***** 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85
217 ISA + 10°C 434 683 0.636 704 0.617 731 0.594 761 0.571 788 0.551 802 0.541
487.5 KCAS ***** ***** 278 278 278 278 278 278 278 278
ISA + 20°C 444 701 0.633 723 0.614 FF ***** ***** 4153 3925 3730 3619 3556 3510 3455 3407
ISA – 10°C 426 679 0.627 699 0.609 720 0.591 748 0.569 776 0.548 789 0.540 –56.5 NAM / LB ***** ***** 0.1174 0.1242 0.1307 0.1347 0.1371 0.1389 0.1411 0.1431
35,000 MT 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85
0.76 ISA 436 700 0.623 721 0.605 742 0.587 772 0.565 801 0.544 814 0.535
490 KCAS 291 291 291 291 291 291 291 291 291 291
224 ISA + 10°C 446 720 0.619 742 0.601 764 0.584 794 0.562 FF 4631 4382 4173 4006 3935 3873 3831 3778 3729 3687
ISA + 20°C 456 739 0.616 –54.3 NAM / LB 0.1058 0.1118 0.1174 0.1223 0.1254 0.1265 0.1279 0.1297 0.1314 0.1329
33,000 MT 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85
0.78 ISA – 10°C 437 735 0.594 754 0.580 774 0.565 798 0.574 830 0.526 849 0.514
494.4 KCAS 304 304 304 304 304 304 304 304 304 304
230 ISA 447 758 0.590 777 0.576 798 0.560 824 0.543 857 0.522 876 0.511 FF 4655 4442 4352 4284 4218 4161 4109 4076 4032 3993
ISA + 10°C 458 780 0.587 800 0.572 821 0.557 845 0.542 –50.4 NAM / LB 0.1062 0.1113 0.1136 0.1154 0.1172 0.1188 0.1203 0.1213 0.1226 0.1238
31,000 MT 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85
0.80 ISA – 10°C 448 792 0.566 810 0.553 833 0.538 857 0.523 899 0.499 922 0.486 498.7 KCAS 318 318 318 318 318 318 318 318 318 318
237 ISA 459 817 0.562 836 0.549 859 0.534 884 0.519 FF 4791 4791 4653 4588 4530 4477 4429 4402 4390 4326
ISA + 10°C 469 840 0.559 860 0.546
–46.4 NAM / LB 0.1041 0.1057 0.1072 0.1087 0.1101 0.1114 0.1126 0.1133 0.1144 0.1153
29,000 MT 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85
0.82 ISA – 10°C 459 876 0.524 903 0.509 935 0.491 971 0.473
503.1 KCAS 332 332 332 332 332 332 332 332 332 332
243 ISA 470 904 0.520 931 0.505 FF 5107 5041 4976 4927 4875 4828 4782 4742 4724 4693
0.83/247 ISA – 10°C 465 931 0.500 968 0.480
–42.5 NAM / LB 0.0985 0.0998 0.1011 0.1021 0.1032 0.1042 0.1052 0.1061 0.1065 0.1072

Fig. 11.26 Examples of SAR tables.

 CHAPTER 11 Cruise and Drag Performance Testing 505

MAXIMUM CRUISE POWER

2000 RPM
ISA + 10°C

AIRSPEED KNOTS
PRESSURE IOAT ENGINE FUEL FLOW TOTAL
11,000 LBS 10,000 LBS 9000 LBS
ALTITUDE TORQUE PER ENG FUEL FLOW
(4990 KG) (4536 KG) (4082 KG)
FEET °C °F LB • FT LBS/HR LBS/HR TAS IAS TAS IAS TAS IAS
0 32 89 1558 477 954 262 260 262 260 262 260
5000 23 73 1678 452 904 282 260 282 260 282 260
10,000 14 57 1793 430 860 303 260 303 260 303 260
15,000 6 42 1906 413 826 327 260 327 260 327 260
20,000 –3 27 2010 401 802 353 260 353 260 353 260
23,000 –7 19 2089 401 802 370 260 370 260 370 260
25,000 –11 13 2154 403 806 382 260 382 260 382 260
27,000 –15 6 2017 379 758 381 250 384 252 387 254
28,000 –17 2 1929 364 728 379 244 382 246 385 248
29,000 –19 –2 1841 350 700 376 238 379 240 383 243
31,000 –23 –10 1671 322 644 369 225 374 229 378 231
33,000 –28 –18 1506 295 590 361 212 367 216 372 219
35,000 –33 –27 1344 268 536 351 198 358 202 364 206
37,000 –36 –32 1178 240 480 338 183 346 187 353 191
39,000 –37 –34 1000 210 420 – – 329 169 338 174
41,000 – – – – – – – – – – –

Fig. 11.27 Example of cruise chart for a propeller-driven airplane.

course, include airspeed, possibly Mach, fuel flow, and SAR. Figure 11.26 pro-
vides two examples of cruise tables found in the aviation industry; these are
for jet-driven airplanes.
Propeller-driven airplanes have engines rated in terms of power, as dis-
cussed previously, and the cruise tables generally take on a slightly different
format. Propeller RPM and engine torque are usually the reference variables
leading to a given fuel flow and cruise true airspeed. The information still
provides a means of estimating the fuel usage for a given distance to be tra-
veled. Figure 11.27 provides such an example.

Reserve Fuel Planning

Reserve fuel, as discussed in Chapter 10, must allow the airplane to reach
alternate airports and also spend additional time in the air compared to the
fuel and time required to get to the primary destination. The performance
engineer can provide charts that define standard reserves (following the
 506
Operational Aircraft Performance and Flight Test Practices
Reserve Fuel and Enroute Time to Alternate – Twin Engine Alternate Flight Plan Fuel
30 min Holding at 160KIAS/5000 FT, 9000 lbs ZFW Distance To Flight Enroute Time TAS Mach CAS Fuel Qty
Alternate (NM) level (hr:min) (kt) No. (kt) (lb)
FUEL USED (LBS) 100 230 00:25 303 0.499 214 2730
ENR, TIME (HH:MM) 120 250 00:28 317 0.526 216 2880
140 310 00:31 333 0.568 206 3010
RANGE CRUISE ALT CRUISE SPD HEADWIND (KT) 160 350 00:33 342 0.593 197 3130
(NM) (FT) (KIAS/MACH) 0 50 100 150 180 370 00:36 352 0.618 195 3240
200 410 00:38 384 0.669 194 3360
608 675 780 974 220 430 00:41 408 0.712 198 3470
100 20,000 200 / 0.440
00:26 00:32 00:41 00:59 240 450 00:44 420 0.732 195 3570
712 793 914 1123 260 450 00:47 420 0.733 196 3660
150 25,000 200 / 0.487 280 450 00:50 420 0.733 196 3760
00:37 00:44 00:55 01:15
300 450 00:52 421 0.734 196 3850
200 35,000 200 / 0.603 785 861 969 1133 320 450 00:55 421 0.734 196 3950
00:45 00:53 01:04 01:20 340 450 00:58 422 0.735 196 4050
250 35,000 200 / 0.603 872 964 1093 1290 360 450 01:01 422 0.735 196 4140
00:54 01:03 01:16 01:35 380 450 01:04 422 0.735 196 4240
936 1021 1133 1283 400 450 01:07 422 0.736 196 4340
300 45,000 200 / 0.748 420 450 01:10 422 0.736 196 4440
01:00 01:08 01:19 01:33
440 450 01:12 423 0.737 197 4530
1011 1106 1230 1402 460 450 01:15 423 0.737 197 4630
350 45,000 200 / 0.748
01:07 01:16 01:28 01:44 480 450 01:18 423 0.738 197 4730
400 45,000 200 / 0.748 1086 1190 1330 1519 500 450 01:20 423 0.738 197 4830
01:14 01:24 01:37 01:55 520 450 01:23 424 0.739 197 4920
1161 1278 1428 1636 540 450 01:26 424 0.739 197 5020
450 45,000 200 / 0.748 560 450 01:29 424 0.74 197 5120
01:21 01:32 01:46 02:05
580 450 01:32 424 0.74 197 5220
500 45,000 200 / 0.748 1237 1362 1529 1756 600 450 01:35 424 0.74 197 5320
01:28 01:40 01:55 02:16 Less Than ISA Temperature:
NOTE: For tailwind cases, use zero wind reserve fuel data. Increase Enroute Time by 2% for each 10° C below ISA
Decrease Fuel Quantity by 0.5% for each 10° C below ISA
NOTE: Increase fuel reserve by 25 pounds per 10°C above ISA.
Greater Than ISA Temperature:
NOTE: Add 50 pounds per 100 nm for a ZFW CG more forward than 25% MAC. Decrease Enroute Time by 2% for each 10° C above ISA

Fig. 11.28 Examples of table formats for reserve fuel planning.

 CHAPTER 11 Cruise and Drag Performance Testing 507

flight format of Fig. 10.1, for example; see Fig. 11.28) or provide enough data
to the pilot to define each segment of the reserve flight profile.
One of the segments is holding. Holding fuel can be planned by setting
the holding speed (typically minimum drag) and simply becomes a function
of holding time. Thus, the performance engineer will typically provide tables
of fuel flow for a defined airspeed and airplane weight (see Fig. 11.29).

Flight Planning: Simplified vs Detailed

The performance engineer must present a clear methodology for the crew
to be able to plan the mission fuel from engine start to end of mission and
engine shutdown. One approach is to provide a simplified fuel planning
chart (see Fig. 11.30). The scenario to develop these charts must include
precise flight profiles (climb schedule, target altitude, cruise speed, descent
profile, expected average temperature deviation from standard). The pilot
must have the expected landing weight (airplane empty weight, crew
weight, passengers plus cargo weight, fuel reserves) and the expected
ground distance to be covered. With that information in hand, the pilot
can then determine the required takeoff weight to perform the mission; the
fuel required for the mission is then simply the takeoff weight minus the
landing weight.
The advantage of this approach for the crew is that it gives them a quick
estimate of the required fuel and the takeoff weight required to execute the
mission. This takeoff weight is not checked against the takeoff requirements
for the airport from which the airplane is dispatched; that is another step of
the analysis. The charts of Fig. 11.30 may have some conservatism in them
because the conditions will not always be of constant winds and temperature
deviation from ISA throughout the flight; it is a simplified flight planning tool
with which a pilot can get used to the margins it provides in accomplishing
a mission.
If the pilot wants to be able to plan the mission in more detail and possibly
stretch the range a little, then some means of estimating the mission in a
more detailed way can also be provided to the crew. It typically consists of
segmenting the mission and estimating the flight conditions to be adopted
and the atmospheric conditions aloft (see Fig. 11.31). This sample table
allows the crew to fill the mission profiles with data collected from other sec-
tions of the flight planning manual.

Fuel (Energy) Management and Conservation

An airplane is designed to perform an optimum mission. For long-range
airplanes, the mission is carrying a given payload for the least cost and fuel
burn. The way an airplane is operated and maintained will often result in
the fuel consumption being higher than the optimum an airplane can
achieve. Fuel conservation is done by carefully managing the operation and
 508
Operational Aircraft Performance and Flight Test Practices
Holding fuel flow (PPH) – 160 KIAS
Weight Holding altitude (feet)
Temp (lbs) 0 1500 5000 10,000 15,000 19,000 21,000 23,000 25,000 27,000 29,000 31,000
8000 597 584 557 522 498 479 471 463 455 446 437 428
9000 620 607 578 546 522 503 495 487 478 468 459 450
ISA 10,000 645 631 605 574 549 530 552 513 504 494 486 477
–20°C 11,000 671 658 633 603 578 559 550 541 532 523 515 508 10
12,000 702 689 666 636 610 591 581 572 563 555 547 541
13,000 734 722 699 669 643 623 612 601 591 581 573 567

FT.
8000 611 598 570 536 510 492 484 476 467 459 449 440
8
9000 634 621 593 561 535 517 508 500 491 481 472 463
de– 00
titu 30,0 00

Holding time – HRS

ISA 10,000 660 646 619 588 562 544 535 527 518 509 500 491
al 25,0,000
–10°C 11,000 687 674 648 618 592 574 565 556 547 538 530 523 re
ssu 20 0
Pre
6 0
12,000 719 706 681 652 626 607 597 687 579 570 563 557 15,0 00
13,000 752 739 715 686 660 639 628 618 607 598 590 583 10,00
8000 627 613 583 550 523 504 496 488 480 471 462 453 500L.
S.
9000 650 636 606 574 548 530 522 513 504 495 486 476
4
ISA 10,000 676 663 633 603 577 558 550 541 532 523 514 505
11,000 704 691 661 633 607 589 580 570 561 553 545 538
12,000 737 724 697 669 642 622 612 603 594 586 578 573
13,000 771 759 733 703 676 655 645 634 624 614 606 600 2
8000 641 627 597 561 536 518 509 502 494 484 474 465
9000 665 652 621 587 562 543 535 527 518 508 498 489
ISA 10,000 692 678 649 617 591 573 564 555 546 537 528 519
0
+10°C 11,000 721 707 679 649 622 604 594 685 576 568 559 552
0 1000 2000 3000
12,000 754 740 714 685 658 639 628 619 610 602 594 589
13,000 788 776 750 720 693 672 661 651 640 631 623 617
Holding fuel – LBS
8000 657 643 612 575 549 530 522 514 505 498 486 477
9000 681 667 636 601 575 557 549 540 531 521 511 501
ISA 10,000 709 695 664 632 605 587 579 670 560 550 541 532 0 200 400 600 800 1000 1200 1400
+20°C 11,000 738 724 695 664 637 619 610 600 591 582 574 567
12,000 772 758 731 701 673 654 644 634 626 617 610 604
Holding fuel – kilograms
13,000 806 792 768 737 709 689 677 667 657 647 639 633

Fig. 11.29 Examples of data provided for holding.

 39,000 FT, MCT
150 Twin engine flight planning for Mach = 0.80 at ISA +10°C

HEADWIND (KTS)
36

100 32

Trip fuel burned – 1000 LB (Does not include reserves)

0 NM

M
200 NM

NM

NM
0N

NM
28

600 N
50

00
140
100

00
1

00
LB

18

26
22

NM
00
24 3,9

00
=7

30
TAILWIND (KTS)

0 GW
TO
20 M ax
LB 90
00 3
–50 74,3 al FL
= nit i
16 ht

CHAPTER 11
ig I 0
e 41
pw

ine
2 FL
am ial

el
z r i t
12 In
Ma

nc
e
0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 0

fer
TRIP TIME (HRS) 43
l FL

Re
8 n i tia
I
TEMP ISA (°C) 0
45
–10 ISA +10 l FL

Cruise and Drag Performance Testing

4 n i tia
MTOW 13,950 LBS 14,000 I
66 64 62 60 58 56 54 52 50 48 46 44
25 0 5 6
3 Landing WT (Including reserves) – 1000 LB
12
,7 ,00 0
12 4 7
0 13,000
,00
11

d
0 6

in
,00 0

lw
10 10 0

TAKEOFF WEIGHT (LBS)

ai
00

st
5
90 12,000

ot
5 nd

Kn
Range – 1000 NM
wi
REF. LINE

0
No 50

15
S)
LB

4 100
T(
GH

11,000
EI

ind
W

3
dw
G

ea
IN

ts h
ND

2 no
0K
LA

10,000 15
1
9000 0
1 3 5 7 9 11
Time - Hours

Fig. 11.30 Two examples of simplified flight planning charts.

509
510 Operational Aircraft Performance and Flight Test Practices

Block Fuel Correction Factor

Zero Fuel Weight (lb) ZFW CG Range
Center of Gravity (% MAC) (%MAC) <1000 n miles >100 n miles
Fuel Load (lb) 14–19.9% –3% –5%
20–24.9% –1.5% –2.5%
Ref (25%) 0 0

Weight Time Distance Fuel Altitude Speed/Mach Temperature Wind

(lb) (min) (n miles) (lb) (ft) (KIAS / M) (⌬ISA) (KTAS)
Ramp Weight
Engine Start, Taxi, Takeoff
Start of Climb
Climb
End of Climb
Accel to Cruise
Start of Cruise
Cruise
End of Cruise
Zoom Climb
Accel to Cruise
Start of Cruise
Cruise
End of Cruise
Zoom Climb
Accel to Cruise
Start of Cruise
Cruise
End of Cruise
Start of Descent
Descent
End of Descent
Holding
Approach
Landing
Reserve Fuel
Zero Fuel Weight

Fig. 11.31 Example of detailed flight planning table.

condition of an aircraft. Although fuel is not the only cost of operation of an

aircraft, it has a significant impact on the bottom line, and its impact becomes
more important as the cost per gallon increases.
For electric planes, which are much more efficient, direct energy cost will
be a lesser impact, but the depth of discharge of a battery pack for a given
mission will shorten the life of the battery. Proper energy management will
reduce long-term, indirect costs to the operator by increasing the number
of available cycles before the battery pack must be replaced.

Fuel Conservation: Ground Operation

A small portion of the fuel used in a mission is burned for ground oper-
ation. Typical measures used to conserve fuel on the ground include:
• Start engine as late as possible.
• Minimize APU use on the ground.
• Take the shortest taxi route.
• Use minimum thrust and braking during taxi.
On multiengine airplanes, taxiing with one or more engines shut down is
possible. But one must balance operation and safety considerations with
 CHAPTER 11 Cruise and Drag Performance Testing 511

fuel conservation. Can the engines that are off be started quickly enough to
not delay other airplanes taxiing? Are there any systems that would not work
properly if one engine was left inoperative?

Aerodynamic Cleanliness: Design and Maintenance

Missing seals, steps and gaps outside tolerance, misrigged flight controls,
misrigged doors, paint peeling, overfilled butt joint gaps—these all contribute
to increasing the drag of the airplane. There is a tradeoff between the impact
on drag (increased operating cost due to additional fuel burn) and the loss of
revenue due to the airplane being down for repair. The manufacturer will
allow a certain amount of degradation to the baseline aerodynamics of the
airplane for which safety will not be impacted (stall speed degradation negli-
gible, increase in drag from a climb capability and obstacle clearance point of
view also negligible). Beyond these limits, the airplane will not be authorized
to fly.
Chapter 12 Climb
Performance

Chapter Objective
In this chapter, we expand on the concept of excess thrust discussed in pre-
vious chapters and review the theory behind the ability to climb or descend.
The crew must understand the airplane’s ability to change altitude, including
the time to climb, distance, and fuel to climb. We discuss basic test method-
ologies to validate the models created, and finally we show how the infor-
mation can be presented to the flight crew to plan a mission. This chapter
focuses on the all engines operating (AEO) scenario.

Developing a Climb Performance Model

T
o gain or lose altitude, the airplane flight path vector must move
away from the horizontal. If we first look at a climb, we can identify
the basic forces acting on the airplane necessary to build a climb
model. The axis system we will use for the analysis of climb performance
will be the wind vectors. Figure 12.1 shows the forces, speed vector, and
angles we need to consider for the climb model.
For a climb, the velocity vector will be above the local horizon. The angle
between the velocity vector and the horizon is the flight path angle g. This is
the direction the airplane is flying to. The lift L and drag D forces are, by defi-
nition, perpendicular and parallel to the velocity vector, respectively. The
weight W will act downward along the local vertical. The airplane’s angle
of attack a is the angle between a reference axis on the plane (typically the
longitudinal axis) and the velocity vector. The thrust vector is often not
aligned with the airplane reference axis, so we introduce the thrust angle
of attack angle aT as the angle between the velocity vector and the thrust
vector. Finally, the airplane pitch angle, what the pilot sees, is the angle
between the local horizon and the airplane reference axis. Mathematically,
u¼aþg (12:1)
We will select the wind axes to create our performance models for climb.
Using the wind axes, the sum of forces parallel to the velocity vector are
W dV
T cos (aT )  D  W sin (g) ¼ (12:2)
g dt

513
514 Operational Aircraft Performance and Flight Test Practices

L
T
V


 T
 V

Fig. 12.1 Forces and angles for a basic climb performance model.

where (dV/dt) is the acceleration along the flight path vector. The forces per-
pendicular to the velocity vector are
W V2
L  W cosðgÞ ¼ (12:3)
g r
where (V 2 /r) represents the acceleration that curves the flight path, r being
the radius of the curved path. To simplify our analysis, we will start with a
steady state (dV/dt ¼ 0), straight path (r  1). Then Eqs. (12.2) and (12.3)
reduce to
T cosðaT Þ  D  W sinðgÞ ¼ 0 (12:4)

L  W cosðgÞ ¼ 0 (12:5)
For most flight conditions, the thrust angle of attack will be less than
10 deg. Then the cos(aT )  0.985 or approximately equal to 1.0. This
further simplifies Eq. (12.4), so we get Eq. (12.6), per Fig. 12.2.
T  D  W sinðgÞ ¼ 0 (12:6)
From Eq. (12.6) we note that the flight path angle is proportional to the
specific excess thrust (thrust minus drag divided by the total airplane weight)
T D
sinðgÞ ¼ (12:7)
W
 CHAPTER 12 Climb Performance 515

or
 
T D
g ¼ sin1 (12:8)
W
The specific excess thrust is simply that excess thrust above what is
required to sustain level flight for a given flight condition and airplane
configuration.
The airplane load factor is defined as the ratio of the lift generated to the
airplane weight. Looking at Eq. (12.5), we see that the load factor in a steady
state climb is equal to
L
n¼ ¼ cos (g) (12:9)
W
The load factor in a climb is therefore less than 1.0, the value required for
level flight.
The drag in the climb will also be a little less than the drag in level flight at
the same weight/altitude/temperature and airplane configuration as the drag
due to the lift being smaller. Writing the drag equation in full, using the lift
value from Eq. (12.5),
1 K ½W cosðgÞ2
D ¼ rSL s V 2 S CD0 þ 1 2
(12:10)
2 2 rSL s V S
Gamma is typically small, and the cos(g) is typically near 1.0 in value. The
design T/W ratio of an aircraft is often a good indication of the maximum
climb angle capability of an aircraft in the clean configuration (flaps retracted
and gear up). We can develop the following equation by combining
Eqs. (12.5) and (12.6):
T D T cosðgÞ
sinðgÞ ¼  ¼  (12:11)
W W W E
where E is the lift to drag ratio. To get a better feel for the climb capability of
an airplane, consider the data from Table 12.1.

L
V
T


Fig. 12.2 Forces in a climb, simplified model.

 516
Operational Aircraft Performance and Flight Test Practices
Table 12.1 Typical T/W and Em Values

Aircraft Type Typical T/W (static thrust) Typical Max Em

Two-engine jetliners 0.27–0.36 17–22

Four-engine jetliners 0.23–0.35 17–19

Business jets 0.30–0. 40 14–18

Military jet trainers 0.32–0.95 9–11

Fighters heavily loaded 0.50–0.70 6–8

Fighters combat weight 1.0–1.2 7–10

 CHAPTER 12 Climb Performance 517

Example 12.1
What would be the approximate climb angle of a two-engine jetliner with
average values of T/W and Em from Table 12.1?

Solution: The average maximum T/W for this example would be 0.315, and
the average value of Em would be 19.5. (Note: These average maximums typi-
cally do not occur at the same airspeed; they are simply used to determine a
maximum value of climb angle.) Using Eq. (12.11) and assuming a cos(g) of
1.0 for initial iteration, we get a maximum climb angle for this type of airplane
of the order of
T cosðgÞ 1
sinðgÞ ¼  ¼ 0:315  ¼ 0:264
W E 19:5
g ¼ 15:3 deg
For this climb angle, the cos(g) equals 0.965. Inserting this value back into
Eq. (12.11), the corrected maximum climb angle becomes 15.4 deg, a very
small impact for assuming a value of 1.0 for cos(g) for most climb cases.
Figure 12.3 shows the impact of excess thrust on the climb capability of the
airplane. For most airplane categories in Table 12.1, the maximum climb angle
would be in the range of 15 to 30 deg. The exception is the last category, where
the thrust to weight ratio is greater than 1.0; these airplanes can sustain ver-
tical climbs and possibly even accelerate while climbing at 90 deg.

90

80

70
Flight path angle (deg)

60

50

40

30

20

10

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Specific excess thrust

Fig. 12.3 Flight path angle vs specific excess thrust.

518 Operational Aircraft Performance and Flight Test Practices

Climb Gradient and Flight Path Angle

Another way to express the climb capability of the airplane is to use the
climb gradient. The climb gradient (grad) is defined as the ratio of the height
gained (dh) over the horizontal distance covered (dX ).
dh
grad ¼ (12:12)
dX
By observation of Fig. 12.4, one can see the gradient is equal to the
tangent of the climb angle.
grad ¼ tanðgÞ (12:13)
For small climb angles,
grad  g (12:14)
When the flight path angle is measured in radians, there is a difference
of about 1% at 10-deg flight path angle (a gradient of 0.176 or 17.6%) when
using Eq. (12.14). Figure 12.5 illustrates the relationship between the climb
gradient and flight path angle. Two climb gradients of interest are also
shown; they will be explained later in the chapter.

Maximum Climb Angle

The airplane will reach its maximum steady state climb angle when the
specific excess thrust is maximized. The climb angle, Eq. (12.8), can be
rewritten as follows:
T cosðgÞ
sinðgÞ ¼  (12:15)
W E
where E is the lift-to-drag (L/D) ratio of the plane, its aerodynamic efficiency
at the current speed, weight, and configuration. For most Part 25 transport
category airplanes, cos(g)  1 (relatively small climb angle capability), and

dh

dX

Fig. 12.4 Climb gradient vs flight path angle.

 CHAPTER 12 Climb Performance 519

18%
17%
16%
15%
14%
13%
12%
Climb gradient

11%
10%
9%
8%
7%
6%
5%
4% 3.2%
3%
2% 2.4%
1%
0%
0 1 2 3 4 5 6 7 8 9 10
Flight path angle (deg)

Fig. 12.5 Climb gradient to flight path angle relationship.

Eq. (12.15) can be simplified as

T 1
sinðgÞ ¼  (12:16)
W E
To maximize Eq. (12.16), one must maximize the thrust to weight while
minimizing the drag (maximize 1/E). The drag component is minimized by
flying at the speed for minimum drag (Dmin ). As for maximizing the thrust,
the condition of maximum excess thrust may be close to minimum drag
condition or at a speed just below (depending on the thrust lapse rate with
airspeed; see Fig. 12.6).
T 1
sinðgmax Þ   (12:17)
W Em
Then, the airspeed for maximum climb angle is the minimum drag air-
speed as derived in Chapter 8; the equation is rewritten here for convenience:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffisffiffiffiffiffiffiffiffi
W 4 K
Vgmax ¼ VDmin ¼ 1 (12:18)
2 rSL s S
CDo

Equation (12.18) holds true for airplanes equipped with turbojet or low
bypass turbofan engines that have a low, near-zero, thrust lapse rate with air-
speed [i.e., T = f (V )]. For airplanes with medium to high bypass turbofans,
turboprops, piston props, or electric-motor props with a larger thrust lapse
520 Operational Aircraft Performance and Flight Test Practices

12,000
T
10,000
Max excess thrust
Drag, thrust (lb)

8000
D
6000

4000
Min drag
2000

0
0 50 100 150 200 250 300 350 400
Airspeed (kt)

Fig. 12.6 Speed for maximum excess thrust.

rate, the maximum excess thrust will be located at a speed smaller than VDmin .
It is important to take into consideration not flying below minimum drag air-
speed when selecting an operational climb speed. We will elaborate further
on this later in this chapter.
When the lapse rate is larger, one can obtain the speed for maximum
climb gradient graphically. Figure 12.7 shows one such example with a rela-
tively large thrust lapse rate. One curve represents the thrust available T,
another represents the drag for a given configuration D, and the last one is
the excess thrust (T 2 D). Note that the airspeed for maximum climb
angle (maximum excess thrust) can be found by sliding the thrust curve
down until it becomes tangent to the drag curve.
For Part 23 airplanes, up to Amendment 45 (1993), the speed for best
climb angle was expressed as Vx under paragraph 23.65. An online search
for some pilot operating handbooks (POHs) for two Part 23 airplanes (see
Fig. 12.8) allowed us to extract the following information.
For the Cessna 172N, the POH states Vx is the best angle-of-climb speed,
which is the speed that results in the greatest gain of altitude in a given hori-
zontal distance (flaps up). From the POH, at sea level, Vx is equal to 59 KIAS
(61 KCAS once corrected), and at 10,000 ft it is equal to 61 KIAS (63 KCAS).
This airplane has a fairly small weight variation in flight, so the manufacturer
has opted to not list Vx as a function of weight. The maximum takeoff
weight for the 172N is 2300 lb. At that weight, for flaps up, the stall speed Vs
is 53 KCAS (most forward CG) and the minimum drag speed VDmin is
approximately 66 KCAS. Some quick computations of speed ratios show
that Vx /Vs ¼ 1.15. We can also find the ratio VDmin /Vs ¼ 1.26.
For the Beechcraft Bonanza V-35B, the flaps up and gear up best climb
angle airspeed 72 KIAS (at 2650 lb, stall speed under these conditions is
 CHAPTER 12 Climb Performance 521

20,000

18,000

T
16,000

14,000

12,000
Thrust, drag (lb)

T-D

10,000

8000
D
6000

4000
V max
2000
VDmin

0
100 150 200 250
Airspeed (kt)

Fig. 12.7 Finding the maximum excess thrust.

58 KIAS, power off), Vx /Vs ¼ 1.24. The glide speed is 105 KIAS, which leads
to a VDmin /Vs ¼ 1.81, for a maximum takeoff weight (MTOW) of 2650 lb.
Paragraph 23.65 was modified under Amendment 50 to define a
minimum stall speed ratio of not less than 1.2Vs , where the minimum
climb gradient was then defined.

Cessna 172 Beechcraft bonanza V-35

Fig. 12.8 Cessna 172 and Bonanza V-35.

522 Operational Aircraft Performance and Flight Test Practices

20,000

18,000

T
16,000
Back Side
power
14,000
Curve
12,000
Thrust, drag (lb)

T-D

10,000

8000
D
6000

4000

V max
2000
VDmin

0
100 150 200 250
Airspeed (kt)

Fig. 12.9 Flying on the back side of the power curve.

The airspeed for maximum climb angle, especially for the clean (gear and
flaps up) configuration, will usually be lower than the speed for minimum
drag. As seen previously (Chapter 8) and in Fig. 12.9, this is the back side
of the power curve, where a change in throttle position (thrust) changes
the altitude, and the elevator controls speed via a change in angle of attack
(increased induced drag). For large bypass turbofans and even more for
propeller-equipped aircraft, Vgmax may be near or even below Vsw . Therefore,
operationally, to improve airplane handling, reduce the pilot workload, and
provide some additional margin to stall, the speed for best climb angle
should essentially be the speed for minimum drag (max L/D).

Rate of Climb (ROC)

An airplane’s rate of climb (ROC) is the rate of change in altitude vs time
(dh/dt). It can be expressed as the airspeed multiplied by the sine of the flight
 CHAPTER 12 Climb Performance 523

path angle.

dh
¼ V sinðgÞ (12:19)
dt
Combining Eqs. (12.19) and (12.7), we get a ROC equal to

ðT  DÞ TV  DV
ROC ¼ V ¼ (12:20)
W W
The rate of climb is therefore proportional to the airplane’s excess power.
The power required for a given airspeed is the drag multiplied by the true
airspeed. It can be expanded to show

1 K W 2 cos2 ðgÞ
DV ¼ rSL s V 3 S CDo þ 1 (12:21)
2 2 rSL s V S

We thus note that the contribution of the parasitic drag to the power
required increases proportional to the cube of the airspeed, as discussed in
Chapter 8. To see the impact of this power-required effect on the rate of
climb, we look at Fig. 12.10, which shows a typical variation vs the two
extreme conditions of power available from Chapter 8—one where the
assumption of thrust is independent of airspeed and the other where
power available is independent of airspeed.

1.2E+07

Ta V
1.0E+07
ion
pt DV
m
su
as
st
8.0E+06
hru
n tt
Power (ft lb/s)

nsta
Constant power available Co Pa
6.0E+06 assumption

4.0E+06

2.0E+06
Minimum drag
0.0E+00
0 50 100 150 200 250 300 350 400 450
True airspeed (KTAS)

Fig. 12.10 Variation of power available and power required.

524 Operational Aircraft Performance and Flight Test Practices

We clearly note the rapidly increasing power required (DV) with increas-
ing airspeed in Fig. 12.10, as expected. We also note the impact of the power-
available assumption on the chart (constant power available or constant
thrust available models); thus, thrust lapse rate (change with airspeed) will
have a significant impact on the rate of climb. We will expand further on
this subject later in this chapter.
We also added the location of the minimum drag for the model airplane
used on the chart. Graphically, that condition is determined by drawing a line
from the origin to a point tangential to the power required curve (DV),
as shown.
From the previous model used to create Fig. 12.10, we computed the rate
of climb of the same airplane (same drag model) to create Fig. 12.11, which
represents the rate of climb capability of that plane.
We can see how the selection of the powerplant could drastically impact
the airplane’s performance in climb. This is not typically a design choice, but
one manufacturer (Dornier) successfully converted an airplane that was orig-
inally designed as a turboprop (Do328) into a turbofan-equipped (Do328JET)
regional airplane (see Fig. 12.12).
The former was equipped with two Pratt & Whitney Canada PW119B
2179-shaft-horsepower (SHP) engines; the other was equipped with two
Pratt & Whitney Canada PW306B 6000-lb-thrust turbofans. The turboprop
version typically cruised at 25,000 ft for a 500-n-mile mission whereas the
turbojet version would cruise at 35,000 ft.

Rate of climb comparison

3000

2000
Co sum
as
ns pt

1000
ta on
Rate of climb (fpm)

nt
th
i
ru

0
st
Co ump
ass
nst tio
an n

–1000
tp
ow
er

–2000
ava
ilab

–3000
le

Minimum drag
–4000
0 50 100 150 200 250 300 350 400 450
True airspeed (KTAS)

Fig. 12.11 Rate of climb comparison.

 CHAPTER 12 Climb Performance 525

Fig. 12.12 Dornier Do328 and Do328JET.

Maximum Rate of Climb

We note from Fig. 12.10 that the peak rate of climb for the constant
power available model was below the minimum drag condition whereas
the one for the constant thrust model was above. We will now investigate
the specific excess power for these two extreme conditions to define the air-
speed that will yield the maximum rate of climb for the airplane; that is the
highest gain in altitude vs time.
From Eq. (12.20), we know that maximum rate of climb will occur when
the excess power (T V 2 D V) is maximum. For the constant thrust model,
we can find the maximum rate of climb (subscript fastest climb, FC) by
setting the derivative of the rate of climb from Eq. (12.20) with respect to
airspeed equal to zero

     
@ROC @ T V  DV @D
¼ ¼ T DV ¼0 (12:22)
@V FC @V W FC @V FC

We then find that the dynamic pressure for fastest climb is

" sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#
ðT=SÞ 12 CDo K 1 2
qFC ¼ 1þ 1þ 2
¼ rSL s VFC (12:23)
6 CDo ðT=W Þ 2

Again, this is for the condition in which thrust is independent of airspeed.

We can also solve this equation graphically by finding the airspeed where the
slope of the power required curve is equal to the slope of the power available
curve (see Fig. 12.13). This is equivalent to sliding down the power available
curve until a tangent point is found. For the cases where the thrust lapse rate
is nonzero and Eq. (12.23) does not work, the maximum ROC can still be
determined graphically.
For the case where the available power is independent of airspeed (i.e.,
T V is replaced by Pa ), as is often the case for propeller-driven airplanes, the
derivative of Eq. (12.20) with respect to airspeed will yield the condition for
minimum power required (subscript MP) flight. As shown in Chapter 8 and
526 Operational Aircraft Performance and Flight Test Practices

1.2E+07

Ta V
1.0E+07
DV
ion
pt
m
8.0E+06 assu
u st
Power (ft lb/s)

hr
ntt
6.0E+06 nsta
Co

4.0E+06

2.0E+06

Minimum drag Max ROC

0.0E+00
0 50 100 150 200 250 300 350 400 450
True airspeed (KTAS)

Fig. 12.13 Graphical solution to the max ROC for constant thrust assumption.

repeated here, these conditions are

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffisffiffiffiffiffiffiffiffiffiffiffi
W 4 K
VMP ¼ 1 (12:24)
2 rSL s S
3CDo

pffiffiffi
CLMP ¼ 3 CLEm (12:25)

CDMP ¼ 4 CDo (12:26)

qffiffiffiffiffiffi
EMP ¼ 3= Em (12:27)
4

Equation (12.25) shows that the lift coefficient under conditions of

minimum power required is greater than that for minimum drag, pointing
to a flight condition at a slower speed than minimum drag. This was graphi-
cally confirmed by observation of Figs. 12.9 and 12.10.
We should again note that flying below minimum drag may result in
increased flight crew workload and that the operational airspeed to be used
should be pushed to the minimum drag airspeed instead of the minimum
power required speed. One should also consider verifying that the climb
speed has a minimum margin to the stall speed to provide a reasonable man-
euver margin of about 1.3-g under all-engine-operating (AEO) conditions.
 CHAPTER 12 Climb Performance 527

Factors Influencing Climb Performance

To better understand the impact of various parameters on rate of climb, it
can be useful to expand the equation into its smaller components.
 
T D
ROC ¼ V
W
0   1
1r sV3 C K W= cos2 ðgÞ
T S
¼ @ V  2 SL  D0  1
A (12:28)
W W= 2 rSL s V
S
The impact of airspeed is both explicit [all three components of
Eq. (12.28) have an airspeed variable] and implicit (the thrust and the drag
coefficients may change with changing airspeed). The impact of the airspeed
on the climb performance of the airplane was shown previously for constant
drag coefficients and two extreme thrust models (constant thrust model and
constant power model).

Weight
The airplane’s rate of climb capability at any airspeed decreases with
increasing aircraft weight W (see Fig. 12.14). A quick glance at Eq. (12.28)
indicates that a higher weight will:
• Decrease the thrust-to-weight ratio (most significant impact on climb
performance, as seen previously)

12,000

10,000
Rate of climb (fpm)

8000

6000

Increasing
4000
weight

2000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Mach

Fig. 12.14 Impact of weight on rate of climb.

528 Operational Aircraft Performance and Flight Test Practices

• Decrease the percentage contribution of the parasite drag to the decrease

of rate of climb
• Increase the percentage contribution of the induced drag
The airplane’s wing loading (W/S), as seen in Eq. (12.28), is used mostly
as a design parameter to size the wing. The variation of W with fuel burn
during the mission is usually the term used for performance analysis of a
production aircraft.
On traditional airplanes, one should expect the weight of the plane to
decrease as the climb proceeds and fuel/gasoline is burned. This weight
reduction will partially offset the loss of climb capability as the airplane
climbs. For electric-powered airplanes, one should expect a fixed weight
throughout the flight, and thus no benefit during the climb. For hybrid-
electric airplanes, the weight variation during the climb will depend on the
power system used during the climb. (If all electric, there is no change in
weight.)

Thrust
Looking at Eq. (12.28), we see that thrust has a direct impact on the rate
of climb. Thrust is first and foremost controlled by the flight crew by the
movement of the thrust lever (or throttle) from a minimum (idle) to a
maximum available value under the flight condition (speed, altitude, tempera-
ture); we will expand on altitude a little later in this chapter. Figure 12.10
showed the impact of the thrust lapse rate on the rate of climb capability of
an airplane for two extreme thrust models. Thrust models were discussed in
Chapter 8.

Drag
Drag will, of course, have a significant impact on the climb performance.
The rate of climb is proportional to the excess power, which itself is pro-
portional to the excess thrust (T 2 D). Parasite drag, as identified in
Eq. (12.28) by the grouping that includes CD0 , will have the most impact at
the higher speeds. Induced drag, as represented by the grouping containing
K, will have a larger impact at the slower speeds and lower climb angles.
Because the airspeed for best rate of climb on jet airplanes is faster than
the minimum drag airspeed (see Fig. 12.10), and because of the flight path
angle influence on the induced drag term, a given percentage change in
the parasite drag coefficient usually has a more significant impact on the
rate of climb than an equivalent percentage change in the induced drag
parameter K.
For propeller-driven airplanes, the impact of a percentage change to CD0
and K has about the same impact as the airspeed for maximum rate of climb
near minimum drag airspeed (Fig. 12.10).
 CHAPTER 12 Climb Performance 529

Altitude
An altitude change will have the most impact on the available thrust part
of the rate of climb equation. Air-breathing engines, as seen in Chapter 8,
depend on an air mass flow to produce thrust or power; we have created
models for thrust (and power) that changed with density altitude to represent
this impact. The rate of climb is directly proportional to the excess power
(available thrust times true airspeed), so a reduction in thrust with altitude
will result in a decrease in rate of climb at all airspeeds with altitude. An
increase in altitude will also have an impact on the drag (through q), which
will tend to increase the airspeed for maximum rate of climb on jet-driven
airplanes (see Fig. 12.15).
We also offer Fig. 12.16 to show the more detailed impact of the
density altitude on the climb performance. This chart represents the
climb capability of a small turboprop airplane with an engine that has a
flat rating (constant power available below a given temperature at a
given altitude). On the left side of the chart, we see the engine equivalent
rating with a distinct corner at given altitudes (358C at sea level, 258C at
10,000 ft). On the right side of the corner, the engine is thermodynamically
limited, and the available power drops rapidly with temperature (density
altitude change). On the left side of the corner, the engine is flat rated
(constant power available), and the small increase in rate of climb
comes from the decreasing true airspeed with decreasing temperature

2500

2000

Sea level
1500
Rate of climb (fpm)

1000
10,000 ft
500
Increasing
20,000 ft KTAS
0

–500

–1000
0 50 100 150 200 250 300 350 400 450
True airspeed (KTAS)

Fig. 12.15 Effect of altitude on rate of climb.

530 Operational Aircraft Performance and Flight Test Practices

1800

Reference line
Maximum continuous power
1900 propeller RPM
Landing gear retracted 1600
Climb speed 125 KIAS

1400
ISA

Rate of climb - fpm

1200

0
1000
2000

-ft
4000

de
6000 ltitu
re a 800
8000
ssu

10,000 600
Pre

400
–40 –20 0 20 40 60 11,000 10,000 9000 8000 7000 6000
Outside air temp - °C Aircraft weight - LB

1400

1350

1300
Rate of climb - (fpm)

1250

1200

1150

1100
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10,000
Altitude (ft)

Fig. 12.16 Example of the impact of altitude for a turboprop airplane.

while keeping the climb speed constant at 125 KIAS. Overall, there is still
a decreasing rate of climb with altitude increase for a given temperature
deviation from standard.
 CHAPTER 12 Climb Performance 531

6000

5000

4000
Rate of climb (fpm)

36,089 ft
3000

2000

1000

0
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000
Altitude (ft)

Fig. 12.17 Rate of climb vs altitude, thrust reducing with altitude.

We now go one step further and look at an example of rate of climb

change vs altitude with the limiting case of zero rate of climb (Fig. 12.17).
The chart was created assuming a jet-driven airplane with thrust varying
with altitude [Eq. (8.76)] and fixed drag coefficients. The model was run
under best rate of climb conditions and standard atmosphere. As one can
see, the rate of climb steadily decreases with altitude. We note a change in
slope above 36,089 ft. As altitude increases from sea level to 36,089 ft, both
the temperature and the pressure decrease, but above 36,089 ft, the tempera-
ture remains constant while the pressure continues to decrease. This leads to
a faster decrease in air density above the tropopause than below and thus the
larger impact on the rate of climb.
Airplanes with turbocharged engines [see Eq. (8.54) in Chapter 8] may
be able to maintain the power available to a higher altitude, so the
impact on rate of climb capability will be reduced below that transition alti-
tude. Electrically driven airplanes may be capable of maintaining the power
available at all altitudes if there is sufficient cooling for the engine, because
the power generation does not rely on air density. Constant power available
does not mean constant rate of climb, because power required will increase
with increasing altitude while flying at a constant calibrated airspeed (main-
taining best climb condition), which leads to increasing true airspeed with
altitude. One will see a shallower reduction in maximum rate of climb
with altitude.
532 Operational Aircraft Performance and Flight Test Practices

The time required to climb is proportional to the inverse of the rate of

climb
ð H2
dh
t¼ (12:29)
H1 RC

The time to climb being proportional to the area below the curve (1/RC),
it becomes rapidly evident that the time to climb will become significant as
the aircraft’s rate of climb becomes very small (Fig. 12.18). In theory, the
time to climb tends toward infinity as the aircraft gets near a rate of climb
of zero (its ceiling altitude). So the absolute ceiling is not a practical
maximum altitude for an airplane.

Climb Ceilings
An aircraft ceiling is the maximum altitude it can reach given some flight
condition:
• The absolute ceiling is the absolute maximum altitude an aircraft can reach
and maintain altitude (dh/dt ¼ 0) at any speed. If a steady state climb is
used, the aircraft will, in theory, take an infinite time to climb to
this altitude.
• The operational ceiling of an aircraft is the altitude at which the aircraft is
capable of at least 100 fpm. Operationally, this ceiling often gets to be for
rates of climb of 300 fpm.

6000 60

5000 50
1/RC
RC
4000 40
Rate of climb (fpm)

36,089 ft

1/RC (s/100 ft)

3000 30

2000 20

1000 10

0 0
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000
Altitude (ft)

Fig. 12.18 Time to climb vs altitude.

 CHAPTER 12 Climb Performance 533

1000 25
RC Time
900

800 20

Time to climb next 1000 ft (min)

700
Rate of climb (fpm)

600 15
Combat ceiling 200 fpm
500
ceiling

36,089 ft
400 10

300
200 fpm ceiling
200 5
100 fpm ceiling Combat ceiling
100

0 0
30,000 32,000 34,000 36,000 38 000 40 000 42 000
Altitude (ft)

Fig. 12.19 Time to climb next 1000 ft.

• The combat ceiling of fighters is the altitude at which the aircraft is still
capable of a rate of climb of 500 fpm.

With these definitions in mind, and with the example shown in Fig. 12.18,
we created Fig. 12.19, which represents the time required to climb the next
1000 ft once a climb ceiling is achieved. For the combat ceiling value, once
the airplane reaches an altitude where the rate of climb is 500 fpm (just
below 37,000 ft), the pilot should expect approximately 3 min more to
climb the next 1000 ft. When the airplane reaches the altitude where the
rate of climb is 200 fpm (just above 39,000 ft), the pilot should expect that
the next 1000 ft would require over 12 min. The 100 fpm is often considered
the absolute ceiling of the airplane that would be provided in flight
planning manuals.
The previous discussion focused on the excess thrust capability of the air-
plane to select a maximum altitude, but one must also consider the airplane’s
ability to generate lift. For Part 25 (and Part 23) airplanes, one should expect
the airplane to be capable of a minimum maneuverability margin at the
higher altitude (discussion in Chapter 14). The airplane’s manufacturer will
often recommend a maximum altitude for the airplane to be capable of
about 30- to 40-deg bank (1.15 to 1.3 g capability) without encountering
low-speed or high-speed buffet [see Chapter 14 EASA AMC for
25.251(e)(2)]. The chart shown in Fig. 12.20 was found on the Internet and
shows the varying buffet limits during the climb of a Lockheed U-2.
534 Operational Aircraft Performance and Flight Test Practices

75

70 Buffet
region
Altitude -1000 ft

65

Climb speed
60

55
Buffet
margin

50
0.60 0.65 0.70 0.75 0.80
Mach number

Fig. 12.20 Lockheed U-2 climb profile vs the buffet region. Source: Flight Manual Models U-2C
and U-2F aircraft, Figure 6–2, 196.

Acceleration
In the previous analysis, we assumed that the airplane was climbing at a
constant true airspeed (constant TAS, zero acceleration) with respect to a
stable air mass with no wind gradient (constant ground speed). We also
noted that the airspeed for best rate of climb does increase with increasing
altitude; thus, if the pilot wants to maintain the best rate-of-climb airspeed,
the airplane will need to accelerate. If one accounts for the acceleration in
the climb (not at a constant TAS), the rate of climb equation takes on the
following form:
ðT  DÞV dh V dV dh V dV dh
¼ þ ¼ þ (12:30)
W dt g dt dt g dh dt
We find the following form of the rate-of-climb equation:
ðT  DÞV
dh ðdh=dt Þaccel¼0
¼ W ¼  (12:31)
dt V dV V dV
1þ 1þ
g dh g dh
 CHAPTER 12 Climb Performance 535

Table 12.2 Approximate Acceleration Factor in Climb

Condition H , 36,089 ft H . 36,089 ft

Constant M –0.133 M2 0
Constant VEAS 0.567 M 2 0.7 M 2

The term V/g dV/dh in the denominator is

called the acceleration factor. It takes on the Note: The correction factor
approximate form shown in Table 12.2. is based on geometric
altitude change, not the
pressure altitude change
Winds measured. These
approximations result in a
The previous climb modeling was for a no wind small error that can be
neglected. Figure 12.21
condition; it represented a climb capability with shows an example of the
respect to the air mass. A steady horizontal wind impact of climbing at a
(no change in wind speed and direction with alti- constant CAS instead of a
tude) will not impact the climb angle of the airplane constant TAS. There is a
reduction in the maximum
with respect to the air mass (the flight path angle g); rate of climb (due to the
that angle is proportional to the specific excess acceleration, of course), and
thrust per Eq. (12.8). A steady horizontal wind there is a reduction in the
will impact the climb angle with respect to the airspeed for maximum rate
of climb.
ground. A headwind will increase the apparent

9000

8000

7000
Constant TAS
Rate of climb (fpm)

6000

5000
Constant CAS
4000

3000

2000

1000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Mach

Fig.12.21 Example of impact of acceleration on rate of climb.

536 Operational Aircraft Performance and Flight Test Practices

Vw
d
a dwin
 still air  he  still Vv
 Headwind air
Vh

Fig. 12.22 Steady wind impact on climb angle.

climb angle of the airplane with respect to a ground obstacle.

 
1 V sinðgÞ
ggr ¼ tan (12:32)
V cosðgÞ + Vw

The negative sign is used for a headwind component. It can be seen in

both Fig. 12.22 and Eq. (12.32) that a headwind component will effectively
increase the airplane climb gradient with respect to a ground obstacle; that
is, it will be easier to clear a ground obstacle with a headwind component
than with a tailwind component. Let’s define a variable that we will call the
wind fraction vw equal to the ratio of the horizontal wind speed component
along the flight path Vw to the airplane true airspeed V.

Vw
vw ¼ (12:33)
V

Then, Eq. (12.32) becomes

 
sinðgÞ
ggr ¼ tan1 (12:34)
cosðgÞ + vw

Another factor that will have a significant impact on the rate of climb is the
wind gradient in the climb (see Fig. 12.23). Climbing with an increasing head-
wind, for example, will reduce the aircraft’s true ground speed even if the pilot
maintains a constant true airspeed. Figure 12.24 shows the true airspeed in a
climb vs the corresponding ground speed for the data shown in Fig. 12.23;
there is an increasing headwind with altitude increase. This slowing down of
the aircraft will result in an increased rate of climb (i.e., trading kinetic
energy for potential energy!). As the airplane crosses 31,000 ft while changing
heading, there is a large increase in ground speed while the true airspeed
remains essentially constant; this would result in a decreased rate of climb.
The impact of the wind gradient can be analyzed by using Eq. (12.31) as
described for the acceleration effects in a climb; the gradient of the ground
speed vs the change of altitude is the acceleration to consider.

dV dðKTGSÞ
¼ (12:35)
dh dh
 CHAPTER 12 Climb Performance 537

150 60

Airplane heading (deg), wind direction (deg)

100 50

50 40
Wind speed

Wind speed (kt)

0 30
Airplane
heading
–50 20

–100 10
Wind direction
–150 0
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000
Pressure altitude (ft)

Fig. 12.23 Wind variation with altitude during a climb.

500

480
Constant KCAS Constant Mach
460

440

420
KTAS, KTGS

KTAS
400

380
KTGS
360

340

320

300
10,000 15,000 20,000 25,000 30,000 35,000 40,000
Altitude (ft)

Fig. 12.24 Speed gradient vs altitude.

538 Operational Aircraft Performance and Flight Test Practices

45,000

40,000

35,000

30,000
Best rate of climb
schedule
Altitude (ft)

25,000

250 KCAS/Mach 0.74

20,000
Climb schedule

15,000

10,000

5000

0
0.3 0.4 0.5 0.6 0.7 0.8
Mach

Fig. 12.25 Example of best climb schedule.

Climb Schedules
The best rate of climb (ROC) airspeed is a function of multiple factors, as
discussed previously. It may not be practical for the crew to adjust the air-
speed with increasing altitude and decreasing weight, make corrections for
nonstandard days, and so forth. The OEM can find an easy-to-follow climb
speed schedule (usually a combination of KCAS/Mach) that will result in
the best approximation of the best ROC for distance/fuel/time and so on.
Figure 12.25 shows one such example.
That speed schedule is also easily followed by an automatic flight control
system (autopilot). It also has the advantage of predefining climb profiles that
the manufacturer can optimize for time/fuel/distance to climb and then
provide data in the flight planning manual for the crew to use.

Descending Flight
The analysis for a descending flight is similar to that for a climb. We again
use the flight path vector as the basis for the orientation of the major axes (see
Fig. 12.26). The equation for the force equilibrium along the flight path in a
 CHAPTER 12 Climb Performance 539

L
D

T
V W

Fig. 12.26 Force equilibrium in a descent.

constant airspeed descent is the same as the one used for climb in Eq. (12.4),
and the flight path angle is as defined by Eq. (12.8).
To get a negative angle of climb g, required for a descent, one can see that
a thrust deficit (drag exceeding thrust) is required. This can be a problem
when designing very sleek airplanes to optimize cruise performance and
reduce fuel burn. These airplanes may have a harder time descending
without the use of drag devices (spoilers, speed brakes, etc.). We reserve
this discussion of the steep approach to land requirements until Chapter 20.
If acceleration is present during the descent, the same correction factors
as used in the climb equation will apply. The manufacturer also defines stan-
dard descent airspeed schedule, often similar to that defined for climb.

Load Factor and Flight Path Variation

To change the flight path angle from zero for level flight (n ¼ 1.0) to a
climb or descent, the pilot will have to initiate a pitching maneuver. On con-
ventional airplanes, this means an elevator displacement. As a starting point,
let’s consider a pitch-up maneuver (see Fig. 12.27). The airplane is initially in

L>W r
L=W

W W

Fig. 12.27 Initiation of a pitch-up maneuver.

540 Operational Aircraft Performance and Flight Test Practices

straight level flight where lift equals weight (n ¼ L/W ¼ 1.0). To start a climb
(flight path greater than zero), the flight path will need to be curved up. To do
so, the airplane will need to generate an increased lift over weight ratio (load
factor greater than 1.0). The curved flight path will have a radius r.
The sum of the forces along the vertical axis at the point of the start of the
pullup will be
W V2
LW ¼ (12:36)
g r
where:
V ¼ Airplane true airspeed
r ¼ Radius of curvature of the flight path
Equation (12.36) can be rewritten as follows:
   
LW L V2
¼ 1 ¼ (12:37)
W W gr
or,
V2
n¼ þ1 (12:38)
gr
The radius of curvature of the flight path at the start of the pullup can
therefore be expressed as a function of the load factor.
V2
r¼ (12:39)
g ðn  1Þ

KTAS (kt) Pressure altitude (ft)

265 13.0
Pressure altitude

260 12.8
KTAS (kt)

255
(ft) (103)

12.6
250
245 12.4
240 12.2
235 12.0
Pitch (deg) FPA (deg)
12
10
Pitch (deg)

8
FPA (deg)

6
4
2
0
–2
Nz (g) Elevator (deg)
1.15 0
1.10 –0.5
Elevator
Nz (g)

(deg)

1.05 –1.0
1.00 –1.5
0.95 –2.0
40 50 60 70 80
Time (s)

Fig. 12.28 Example of start of climb.

 CHAPTER 12 Climb Performance 541

Along the same lines, one can determine that the rate of change of the
flight path angle is proportional to the true airspeed divided by the radius
of curvature,
V g ðn  1Þ
ġ ¼ ¼ (12:40)
r V
Figure 12.28 shows an example of a start-of-climb maneuver. There is an
elevator input at time 56 s (bottom trace) that leads to a load factor (Nz)
increase. The flight path angle and pitch angle start to increase, and altitude
is gained.

0 Elevator position

–0.5

–1.0

–1.5

–2.0
Pitch attitude Angle of attack
12
10
8
6
4
2
0

Airspeed
230

225

220

215

210
Load factor
1.4

1.3

1.2

1.1

1.0

Fig. 12.29 Example of end of descent.

542 Operational Aircraft Performance and Flight Test Practices

To generate the required increase in load factor Dn to initiate a pullup,

the lift coefficient must be increased by a given amount above that required
for the steady flight condition. A lift coefficient increase is done through
an increase in the angle of attack Da. Then the increase in load factor
becomes proportional to the airplane’s lift curve times the increase in
angle of attack.

Dn ¼ CLa Da (12:41)

Once the target load factor has been reached, the rate of change in flight
path becomes directly proportional to the pitch rate (see the middle trace of
Fig. 12.28). Figure 12.29 is another time history of an end of descent followed
by a start of climb (as noted by the pitch attitude starting smaller than the
angle of attack and then going above the angle of attack).
The pullup is initiated by an elevator input by the pilot. This increases the
angle of attack above that required for steady level flight. We note that in this
case, only a 1-deg angle of attack was required to generate a 0.3-g load factor
increase.
This material applies to any change in flight path angle with the
only difference being the initial condition and the general direction of the
flight path change. For example, Fig. 12.30 shows the flight conditions at
the start of a descent from level flight. At time 22 s, the pilot applies a
pitch input to lower the flight path angle, the resulting pitch rate
being approximately 0.3 deg/s nose down and the average load factor
reduction from level flight being about –0.07 g. By time 30 s, the nose-down
input is essentially all removed, and the flight path angle settles to

Altitude (feet)
41.0
(feet) (103)

40.8
Altitude

40.6
40.4
40.2
Nz (g)
0.02
0
–0.02
Nz (g)

–0.04
–0.06
–0.08
–0.10
–0.12
–0.14
Pitch rate (deg/s) FPA (deg)
Pitch rate (deg/s)

0.5
0
FPA (deg)

–0.5
–1.0
–1.5
–2.0
0 10 20 30 40 50 60 70
Time (s)

Fig. 12.30 Start of descent.

 CHAPTER 12 Climb Performance 543

Elevator (% max)
Elevator
(% max) 0
–50

Nz (g)
0.3
0.2
Nz (g)

0.1
0
–0.1
Pitch (deg) Flight path angle (deg)
10
Angle
(deg)

0
–10
Airspeed (KTAS)
Airspeed
(KTAS)

135
130
125
0 2 4 6 8 10 12 14
Time (s)

Fig. 12.31 End of descent.

approximately –1.5 deg. For this example, the true airspeed was approxi-
mately 440 KTAS.
The next example, Fig. 12.31, is the end of a steep descent at almost con-
stant airspeed. Note again how the pitch angle and the flight path angle move
in tandem.

Example 12.2
An airplane climbing through 20,000 ft at 390 KTAS and a flight path angle of
2 deg needs to level off by 21,000 ft. The pilot pushes the nose over so as to
reduce the climb load factor by about 0.1 g (average) until the altitude is cap-
tured. What is the airplane rate of climb prior to the pushover? At what alti-
tude should the pilot start the pushover so as capture 21,000 ft exactly? What
is the climb load factor?

Solution: The airplane rate of climb is

6076 ft=n miles ft ft
V sinðgÞ ¼ 390 KTAS sinð2 degÞ ¼ 22:97 ¼ 1378
3600 s=hr s min

(Continued)
544 Operational Aircraft Performance and Flight Test Practices

Example 12.2 (Continued)

The load factor in the climb is equal to cos(2 deg) or 0.999 (essentially 1.0). So
average load factor in the pushover would be 0.90. The time it would take to
turn the flight path vector from þ2 deg in the climb to 0 deg for level flight
can be calculated as follows:
 
V g ðn  1Þ 32:17 sft2 ð0:9  1:0Þ rad deg
ġ ¼ ¼ ¼ ft
¼ 0:00489 ¼ 0:28
r V 658:2 s s s

So it would take a push of approximately 7.1 s to level off. While climbing

at 22.97 ft/s, the pilot would need to start the pushover approximately 163 ft
below the level-off altitude. But this is a curved path with reducing rate of
climb during the pushover. The radius of curvature is
V2
r¼ ¼ 134,586 ft
g ðn  1Þ
The actual altitude the pushover should start at is ½r(1  cosðgÞ) ¼ 82 ft
below the target altitude.
Pressure altitude (ft)
21.2
Pressure altitude

21.0
20.8
(ft) (103)

20.6
20.4
20.2
20.0
Body normal acceleration (g)
0.10
Normal acceleration

0.05
0.00
(g)

–0.05
–0.10
–0.15
–0.20
Flight patch angle (deg)
4
Flight patch angle

3
2
(deg)

1
0
–1
–2
0 10 20 30 40 50 60
Time (s)

Energy Height
An airplane flying at a given altitude and airspeed will possess a given
level of energy. This energy can be broken down, for the purpose of
airplane performance analysis, into a component of potential energy
(function of mass and altitude) and kinetic energy (function of mass and
speed). We define an airplane’s specific energy He as its total energy
divided by its weight.

m g h þ 12 m V 2 V2
He ¼ ¼hþ (12:42)
W 2g
 CHAPTER 12 Climb Performance 545

The specific energy has units of altitude and is sometimes referred to as

energy height. The height h and the speed V are inertial values. The height h is
therefore the geometric height of the airplane. The speed V is the airplane
ground speed. Therefore, local winds will impact the airplane ground
speed for a given airspeed and must be accounted for in the analysis. To sim-
plify the conversation, we will only use the term speed in our discussion of
specific energy.
Figure 12.32 shows lines of constant energy height. Note the quadratic
behavior of the airspeed vs the linear behavior of altitude, per Eq. (12.42). We
note, for example, that an airplane at a height of 10,000 ft and zero speed
would have as much energy as one flying at a height of 0 ft (sea level) and a
speed of about 800 ft/s (802.1 ft/s). If the airplane climbed to 50,000 ft and
zero speed, it would have as much energy as one at sea level and 1794 ft/s.
Note that Eq. (12.42) does not include any of the airplane’s characteristics
(drag, thrust available, weight, etc.). The rate of change per unit time of
energy height is called specific power Ps .
d He dh V dV
¼ Ps ¼ þ (12:43)
dt dt g dt
from which we can extract the rate of climb
dh PS
¼  (12:44)
dt V dV
1þ
g dh

50,000

45,000

40,000

35,000

30,000
Height (ft)

25,000

20,000

15,000

10,000

5000

0
0 200 400 600 800 1000 1200 1400 1600 1800
Speed (ft/s)

Fig. 12.32 Lines of constant energy height.

546 Operational Aircraft Performance and Flight Test Practices

Note the similarity between Eq. (12.44) and Eq. (12.31). The specific
power defined previously (airplane independent) can be tied to the airplane’s
specific excess power [Eq. (12.20)] to map the airplane’s capability to climb
and/or accelerate. A search of the Internet will reveal several examples of
specific energy map of airplanes (see Fig. 12.33), with supersonic fighters

F-104G 1-g specific excess power

Combat weight-18037 pounds
Clean max power - no maneuver flap
70
G

60
0
100
C D A F
50 200
300
Altitude, h, FT x 103

40
E B 400
500
30

Stall
20 200
100
10

0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Mach
VJ*,M
–C
m
1k

300
=0
H

200
m
5k

1 km
km
8
km
K=0 11 km
100 12 km
13 5 km
8 km

0.5 1.0 1.5 2.0 M

Fig. 12.33 Example of specific height capability. Source: Flight Manual F104G.
 CHAPTER 12 Climb Performance 547

offering maps that clearly show the impact of the drag rise near Mach 1.0
(large dip in the specific power).

Certification Requirements
The Part 25 certification requirements for all-engine-operating (AEO)
climb performance are few, the focus being on the climb capability with
one engine inoperative (discussed in the next chapter). The FAA AEO
climb requirements (25.117 and 25.119), Part 25 Transport Category
Airplanes, at Amendment level 140 (year 2015) were:
§25.117 Climb: general.
Compliance with the requirements of §§25.119 and 25.121 must be shown
at each weight, altitude, and ambient temperature within the operational
limits established for the airplane and with the most unfavorable center of
gravity for each configuration.

These requirements ask that the manufacturer be able to predict climb per-
formance under various conditions of weight, altitude, and temperatures (see
§25.119, author bold and italics). It also asks that the most adverse center of
gravity (the one having the most impact on the climb performance) be used
so as not to let the operator determine that condition. On traditional (tail aft)
airplanes, this will correspond to a forward center of gravity.
§25.119 Landing climb: All-engines-operating.
In the landing configuration, the steady gradient of climb may not be less
than 3.2 percent, with the engines at the power or thrust that is available 8
seconds after initiation of movement of the power or thrust controls from the
minimum flight idle to the go-around power or thrust setting—
(a) In non-icing conditions, with a climb speed of VREF determined in
accordance with §25.125(b)(2)(i); and
(b) In icing conditions with the landing ice accretion defined in appendix C,
and with a climb speed of VREF determined in accordance with
§25.125(b)(2)(ii).

This requirement states that at the proper approach speed in the landing con-
figuration (landing flaps and gear down), the manufacturer must provide the
operator with the combination of weight, altitude, and temperature that will
just meet 3.2% climb gradient capability. Because this is to provide a limiting
weight for landing go-around, one must expect that the engine will need to
increase thrust from that required for the approach to one that will allow
at least 3.2% climb capability. Larger engines tend to accelerate more
slowly; to put an upper bound on the acceleration time, the airplane manu-
facturer must determine what is the available thrust no later than 8 s after the
pilot input to increase thrust during the go-around. The airplane manufac-
turer should find the most adverse (slowest) acceleration condition (engine
bleed, power extraction, altitude, temperature, etc.) and demonstrate that
condition in flight. Figure 12.34 shows one condition where the engine
548 Operational Aircraft Performance and Flight Test Practices

100

90
Percent maximum go-around thrust
80

ion
dit
70

on
%c
60

i on
100

it
nd
50

co
%
40

80
30

§25.119
20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12
Time from TLA movement (s)

Fig. 12.34 Examples of thrust after 8 s.

reached 100% of maximum go-around thrust within the time limit and
another where the engine reached 80%; the difference in time to reach
100% was only about 1.5 s.
That 1.5 s means the airplane manufacturer cannot use more than 80% of
the maximum available go-around thrust to compute the maximum allow-
able landing weight AEO. This can be a big impact on the airplane’s ability
to bring a payload at a given airport.
For small airplanes, FAA Part 23 has many more requirements for AEO
climb performance because some airplanes operating under Part 23 have
only a single engine. The requirements are as follow, pre-Amendment 64.
(The author underlined the climb gradients; note that there were no specific
requirements for electric airplanes at that time):
§23.63 Climb: General.
(a) Compliance with the requirements of §§23.65, 23.66, 23.67, 23.69, and
23.77 must be shown—
(1) Out of ground effect; and
(2) At speeds that are not less than those at which compliance with the
powerplant cooling requirements of §§23.1041 to 23.1047 has been
demonstrated; and
(3) Unless otherwise specified, with one engine inoperative, at a bank
angle not exceeding 5 degrees.
(b) For normal, utility, and acrobatic category reciprocating engine-powered
airplanes of 6,000 pounds or less maximum weight, compliance must be
shown with §23.65(a), §23.67(a), where appropriate, and §23.77(a) at
maximum takeoff or landing weight, as appropriate, in a
standard atmosphere.
 CHAPTER 12 Climb Performance 549

(c) For reciprocating engine-powered airplanes of more than 6,000 pounds

maximum weight, single-engine turbines, and multiengine turbine
airplanes of 6,000 pounds or less maximum weight in the normal, utility,
and acrobatic category, compliance must be shown at weights as a
function of airport altitude and ambient temperature, within the
operational limits established for takeoff and landing, respectively,
with—
(1) Sections 23.65(b) and 23.67(b) (1) and (2), where appropriate, for
takeoff, and
(2) Section 23.67(b)(2), where appropriate, and §23.77(b), for landing.
(d) For multiengine turbine airplanes over 6,000 pounds maximum weight
in the normal, utility, and acrobatic category and commuter category
airplanes, compliance must be shown at weights as a function of airport
altitude and ambient temperature within the operational limits
established for takeoff and landing, respectively, with—
(1) Sections 23.67(c)(1), 23.67(c)(2), and 23.67(c)(3) for takeoff; and
(2) Sections 23.67(c)(3), 23.67(c)(4), and 23.77(c) for landing.

§23.65 Climb: All engines operating.

(a) Each normal, utility, and acrobatic category reciprocating engine-
powered airplane of 6,000 pounds or less maximum weight must have
a steady climb gradient at sea level of at least 8.3 percent for landplanes
or 6.7 percent for seaplanes and amphibians with—
(1) Not more than maximum continuous power on each engine;
(2) The landing gear retracted;
(3) The wing flaps in the takeoff position(s); and
(4) A climb speed not less than the greater of 1.1 VMC and 1.2 VS1 for
multiengine airplanes and not less than 1.2 VS1 for single—
engine airplanes.
(b) Each normal, utility, and acrobatic category reciprocating engine-
powered airplane of more than 6,000 pounds maximum weight,
single-engine turbine, and multiengine turbine airplanes of 6,000 pounds
or less maximum weight in the normal, utility, and acrobatic category
must have a steady gradient of climb after takeoff of at least 4 percent
with
(1) Takeoff power on each engine;
(2) The landing gear extended, except that if the landing gear can be
retracted in not more than seven seconds, the test may be conducted
with the gear retracted;
(3) The wing flaps in the takeoff position(s); and
(4) A climb speed as specified in §23.65(a)(4).

§23.77 Balked landing.

(a) Each normal, utility, and acrobatic category reciprocating engine-
powered airplane at 6,000 pounds or less maximum weight must be
able to maintain a steady gradient of climb at sea level of at least 3.3
percent with—
(1) Takeoff power on each engine;
(2) The landing gear extended;
(3) The wing flaps in the landing position, except that if the flaps may
safely be retracted in two seconds or less without loss of altitude and
without sudden changes of angle of attack, they may be retracted; and
(4) A climb speed equal to VREF , as defined in §23.73(a).
550 Operational Aircraft Performance and Flight Test Practices

(b) Each normal, utility, and acrobatic category reciprocating engine-

powered and single engine turbine powered airplane of more than
6,000 pounds maximum weight, and multiengine turbine
engine-powered airplane of 6,000 pounds or less maximum weight in the
normal, utility, and acrobatic category must be able to maintain a steady
gradient of climb of at least 2.5 percent with—
(1) Not more than the power that is available on each engine eight
seconds after initiation of movement of the power controls from
minimum flight-idle position;
(2) The landing gear extended;
(3) The wing flaps in the landing position; and
(4) A climb speed equal to VREF , as defined in §23.73(b).
(c) Each normal, utility, and acrobatic multiengine turbine powered airplane
over 6,000 pounds maximum weight and each commuter category
airplane must be able to maintain a steady gradient of climb of at least 3.2
percent with—
(1) Not more than the power that is available on each engine eight
seconds after initiation of movement of the power controls from the
minimum flight idle position;
(2) Landing gear extended;
(3) Wing flaps in the landing position; and
(4) A climb speed equal to VREF , as defined in §23.73(c).

For FAA Part 23 at Amendment 64 and beyond, the requirements have

moved to a performance-based description, and the powerplant type has
been dropped (again, climb gradients underlined by author):

§23.2120 Climb requirements.

The design must comply with the following minimum climb performance
out of ground effect:
(a) With all engines operating and in the initial climb configuration—
(1) For levels 1 and 2 low-speed airplanes, a climb gradient of 8.3
percent for landplanes and 6.7 percent for seaplanes and
amphibians; and
(2) For levels 1 and 2 high-speed airplanes, all level 3 airplanes,
and level 4 single-engines a climb gradient after takeoff of
4 percent.
(b) After a critical loss of thrust on multiengine airplanes—
(1) For levels 1 and 2 low-speed airplanes that do not meet single-engine
crashworthiness requirements, a climb gradient of 1.5 percent at a
pressure altitude of 5,000 feet (1,524 meters) in the cruise
configuration(s);
(2) For levels 1 and 2 high-speed airplanes, and level 3 low-speed
airplanes, a 1 percent climb gradient at 400 feet (122 meters) above
the takeoff surface with the landing gear retracted and flaps in the
takeoff configuration(s); and
(3) For level 3 high-speed airplanes and all level 4 airplanes, a 2 percent
climb gradient at 400 feet (122 meters) above the takeoff surface
with the landing gear retracted and flaps in the approach
configuration(s).
(c) For a balked landing, a climb gradient of 3 percent without creating
undue pilot workload with the landing gear extended and flaps in the
landing configuration(s).
 CHAPTER 12 Climb Performance 551

Flight Test and Data Reduction

Testing for climb performance is, by definition, not a steady state because
at a minimum, atmospheric conditions will change (pressure, temperature,
and possibly winds). The tester can nevertheless control the test by providing
some control variables. The following test tolerances usually result in good
data collection:
• Maintain target airspeed within +3 KCAS or within +0.005 Mach.
• Some variation is unavoidable, especially for long climbs, but the small
deviation in the speed/Mach from the target airspeed will reduce the
test scatter.
• Testing should be done in a fixed airplane configuration (flaps, gear, etc.).
• Testing should be done with a fixed throttle setting. The general idea is to
get a predictable thrust variation with altitude, removing the pilot
variability. This is feasible for airplanes where engines are controlled by
computers (FADECs) with thrust schedules that are a function of pressure
altitude, temperature, and airspeed.
• For non-FADEC-controlled engines, the pilot should maintain the
defined climb power manually, which will add to the test scatter.
Figure 12.35 shows an example of an airplane where the pilot must
verify that maximum climb power is not exceeded in the climb.
• Testing should be done with wings level.
• If turns are required to avoid traffic or at ATC’s direction, then the
bank angle should be limited to less than 15 deg. This will minimize
the impact on the load factor and therefore the variation in airplane
drag during the climb.
Figure 12.36 shows an example of a long climb (from 10,000 ft to
41,000 ft) following a climb schedule of 250 KCAS/Mach 0.70. We can see
some of the variability in maintaining the airspeed/Mach during the climb.
These small speed variations have an impact on the instantaneous rate of
climb (Fig. 12.37a), but tend to average out over long climbs if the speed tol-
erance specified previously is maintained.
Figure 12.37 illustrates some of the elements that can affect climb per-
formance for the climb shown in Fig. 12.36. The temperature gradient may
not be per a standard atmosphere; the winds may change as the airplane
gains altitude [as seen by the change in the spread between the ground
speed (KTGS) and the airplane true airspeed (KTAS)]. For long climbs, the
airplane may need to change direction (bank angle) as directed by ATC.
With this variability in mind, we will define some climb test procedures to
collect flight test data and validate the airplane’s climb performance.

Sawtooth Climb
A sawtooth climb test is a series of sequential test climbs followed by
repositioning descents to collect climb performance information. The
552 Operational Aircraft Performance and Flight Test Practices

Conditions:
1900 RPM
Inertial separator normal
Pressure Climb Rate of climb - feet per minute (FPM)
Weight
altitude speed
pounds –40 °C –20 °C 0 °C 20 °C 40 °C
feet KIAS
Sea level 104 975 960 940 920 705
4000 104 940 915 890 785 450
8000 104 890 865 780 515 215
8750 12,000 101 840 710 510 275 ---
16,000 95 610 445 260 50 ---
20,000 87 345 195 20 --- ---
24,000 78 95 --- --- --- ---
Note
1. Torque set at 1865 foot-pounds or lesser value must not exceed
maximum climb ITT of 765 °C or Ng of 101.6%

Torque lime – 1865 foot-pounds

1900

1800

2
4
1700 6

Sea
8

1600

leve
10

l
12

1500
Torque – foot-pounds

1400
14

ure

1300
16

rat
18

pe

1200
tem
20

g
tin

1100 Pr
22

es
era

su
24

op

re
1000 a lti
um

tu
de
xim

900 -1
Ma

00
0f
ee
800 t

700
–60 –50 –40 –30 –20 –10 0 10 20 30 40 50 60
Outside air temperature –°C

Fig. 12.35 Example of climb performance with pilot setting climb variable power.

name of the test methodology comes from the typical shape of the altitude vs
time data traces, as shown in Figs. 12.38 and 12.39.
Each climb is performed at a constant airspeed (calibrated) or Mach
number over an altitude band appropriate for the test conditions. For each
 CHAPTER 12 Climb Performance 553

speed tested, the climb is typically done in two opposite headings to average
out wind and wind gradients. If the testing is done in the same air mass, one
can average out the two climbs. Wind effects can also be minimized by flying
at 90 deg to the wind; however, winds will typically change direction and

45,000

40,000

35,000

30,000
Pressure altitude (ft)

25,000

20,000

15,000

10,000

5000

0
0 100 200 300 400 500 600 700 800 900 1000
Time (s)

260 0.75

250 0.7
Calibrated airspeed (KCAS)

240 0.65
Mach

230 0.6

220 0.55

210 0.5

200 0.45
0 100 200 300 400 500 600 700 800 900 1000
Time (s)

Fig. 12.36 Steady 250 KCAS/Mach 0.70 climb to cruise altitude.

 554
Operational Aircraft Performance and Flight Test Practices
45,000 45,000

40,000 40,000 ISA

35,000 35,000

Pressure altitude (ft)

Pressure altitude (ft)

30,000 30,000

25,000 25,000

20,000 20,000

15,000 15,000

10,000 10,000

5000 5000

0 0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 –60 –50 –40 –30 –20 –10 0
Rate of climb (fpm) Static air temperature (deg C)

45,000 45,000

40,000 40,000

35,000 35,000
KTGS
Pressure altitude (ft)

Pressure altitude (ft)

30,000 30,000
KTAS
25,000 25,000

20,000 20,000

15,000 15,000

10,000 10,000

5000 5000

0 0
250 270 290 310 330 350 370 390 410 430 –15 –10 –5 0 5 10 15
Speed (kt) Bank angle (deg)

Fig. 12.37 Some variables impacting climb performance flight test.

 CHAPTER 12 Climb Performance 555

Fig. 12.38 Sawtooth climb profiles.

13,000 150
Altitude
12,500
Test condition 1 Test condition 2
12,000 100

11,500
Pressure altitude (ft)

Heading (deg)
11000 50

10,500

10,000 0

9500

9000 –50
Heading
8500

8000 –100
0 100 200 300 400 500 600 700 800
Time (s)

Fig. 12.39 Flight test data of performance climb test with reciprocal heading.

magnitude as an aircraft climbs, so only part of the climb could be done at

truly 90 deg to the wind (wind gradient effects minimized, not eliminated).
To execute this test, the flight crew first selects the altitude band to be
covered for the test (with the target altitude for the rate of climb being the
midpoint in the band). The pilot then trims the aircraft at the desired test air-
speed and a much lower altitude than the start altitude. This is followed by a
thrust increase to the desired thrust setting for the test and a pitch up to
maintain the airspeed as the thrust goes up. The timer for the climb test is
started when the start altitude is crossed (Fig. 12.40) and stopped when the
end altitude is crossed.
The rate of climb for the midpoint altitude is then simply the altitude
gained divided by the time. The test is repeated in the opposite direction
to correct for winds.
556 Operational Aircraft Performance and Flight Test Practices

End altitude

e
t im
mb
Cli

Start altitude
Establish constant
airspeed & thrust climb
prior to reaching start altitude

Fig. 12.40 Sawtooth climb altitude band.

Example 12.3
Figure 12.41 shows the flight trace for an airplane that weighs 57,200 lb at test
time under standard atmospheric conditions with essentially calm steady
winds. Calculate the excess thrust. What is the rate of climb for this flight con-
dition? What is the climb gradient?

12,500 139

138

12,000 137
Altitude
Calibrated airspeed (KCAS)

136
Pressure altitude (ft)

11,500 135

134

11,000 133

132
Airspeed
10,500 131

130

10,000 129
400 450 500 550 600 650 700 750
Time (s)

Fig. 12.41 Climb flight condition.

(Continued)
 CHAPTER 12 Climb Performance 557

Example 12.3 (Continued)

12,500 139
165 s 138

12,000 137
Altitude

Calibrated airspeed (KCAS)

136
Pressure altitude (ft)

11,500 1000 ft 135

134

11,000 133

132
Airspeed
10,500 131
130.6 KCAS
130

10,000 129
400 450 500 550 600 650 700 750
Time (s)

Fig. 12.42 Average rate of climb.

Solution: The first task to analyze this climb is to find the stable portion of
the climb. By observation, we selected the climb portion between 11,000 ft
and 12,000 ft where the calibrated airspeed displayed to the pilot was stable
to within about +0.5 kt; typically the smallest resolution provided is 1 kt. It
took 165 s to cover this 1000-ft climb (Fig. 12.42), so the resulting average
rate of climb for this test is as shown below and in Fig. 12.43.
dh 1000 ft
¼ ¼ 6:06 ft=s ¼ 363:6 ft=min
dt 165 s
The chart in Fig. 12.43 shows the rate of climb, as computed by the change
in altitude recorded over time (computed every second in this example). The
signal is noisier, as expected, but one can still extract an average rate of climb
of about 6 ft/s, similar to what was computed previously for the
simplified method.
This averaging method is good over a small range of altitude (say 1000 ft at
reasonable excess thrust condition), but is not as good over a much larger
altitude band.
The average airspeed for this test is 130.6 KCAS. At the midpoint altitude
of 11,500 ft under standard atmospheric conditions, 130.6 KCAS equals 155.1

(Continued)
558 Operational Aircraft Performance and Flight Test Practices

Example 12.3 (Continued)

12,500 25
165 s

12,000 Altitude 20
Pressure altitude (ft)

Rate of climb (fps)

11,500 1000 ft 15

11,000 10

6 fps
10,500 5

Rate of climb
10,000 0
400 450 500 550 600 650 700 750
Time (s)

Fig. 12.43 Rate of climb.

KTAS. Using this true airspeed value as the average speed for the test, the
horizontal distance covered over 165 s is
dX ¼ 155:1 KTAS  165 s ¼ 43,193 ft or 7:1 n miles
The average climb gradient for this test is
dh 1000 ft
grad ¼ ¼ ¼ 0:0231 ¼ 2:31%
dX 43,193 ft
During the climb at nearly constant KCAS, the true airspeed increased
by 2.5 kt over the 1000-ft climb. The specific excess thrust (the zero accelera-
tion equivalent climb gradient) is then computed from Eq. (12.30) to be (after
adjusting the units)
 
ðT  DÞ ðdh=dt Þ V dV
¼ 1þ
W V g dh
 
ð363:6 fpmÞ 155:1 KTAS
¼ 1þ 0:0025 kt=ft ¼ 0:023945
155:1 KTAS 32:17 ft=s2
The airplane excess thrust is then the airplane weight multiplied by the
specific excess thrust calculated earlier, or
ðT  DÞ ¼ 1370 lb
 CHAPTER 12 Climb Performance 559

Level Flight Acceleration

The level acceleration method consists of establishing the aircraft at a low
airspeed in the desired configuration to be tested, and then applying climb
thrust (fixed throttle position) and accelerating while maintaining a constant
altitude until the aircraft reaches its maximum airspeed (VMO /MMO or small
acceleration, less than 1 kt/s, for example).
To ensure that the engines are thermally stable and provide the computed
thrust at the low airspeeds at the beginning of the acceleration, the pilot may
climb from a lower altitude at the low airspeed side of the acceleration with
the desired thrust setting. The acceleration could be started by leveling
off the aircraft (see Fig. 12.44). For a level acceleration (dh/dt ¼ 0), the

15,000

14,500
Pressure altitude (ft)

14,000

13,500

13,000

12,500

12,000
340
dt
320

300 dV
Airspeed (KTAS)

280

260

240

220

200
200 220 240 260 280 300 320 340 360 380 400

Fig. 12.44 Level acceleration test example.

560 Operational Aircraft Performance and Flight Test Practices

Level acceleration (Target pressure altitude:)

Engine
Time Airspeed/Mach Ground speed OAT Weight Speed Inlet pressure Inlet temp. Fuel flow
(HH:MM:SS) (KCAS/M) (KTGS) (°C) (lb) (RPM/N1) (psi) (°C) (PPH)

Fig. 12.45 Example of data collection table for level flight acceleration.

excess thrust is
W dh W dV W dV
ðT  DÞ ¼ þ  (12:45)
V dt g dt g dt
This test technique allows for a direct measurement of the aircraft excess
thrust at a given altitude. Testing at constant altitude will also result in an
essentially constant temperature. One should nevertheless record the temp-
erature on a regular basis. If the testing is done by using the indicated
pressure altitude as a reference, care should be taken to correct this altitude
to an equivalent tapeline altitude because a change in geometric altitude
creates an acceleration/deceleration force that will impact the measured
acceleration.
One should record the airspeed on a regular basis (every 15 to 30 s) as
well as the engines’ parameters and fuel remaining (for weight estimate).
The time interval for this test can be very small (only a few minutes), but
the weight variation can be important enough to have an impact on the
climb model developed. Ground speed [inertial navigation system (INS) or
Global Positioning System (GPS)] should also be recorded to see if a wind
gradient (increasing/decreasing difference between TAS and GS) exists
along the acceleration path. Figure 12.45 provides an example of a data
collection table.

Validation Climb
A validation climb test consists of performing a climb from a low altitude
to near climb ceiling in an uninterrupted way using a defined climb pro-
cedure and climb speed schedule. The altitude range covered should be suf-
ficiently large (i.e., 10,000–35,000 ft, for example, for an aircraft with a service
ceiling of 37,000–39,000 ft) to confirm the effects of fuel burn (weight
reduction) on the overall climb performance. The validation climbs should
also cover a range of initial climb weights and climb speed schedules.
The data collected are then used to validate the climb performance model
developed (a combination of aerodynamic coefficients, thrust model, and fuel
flow model for weight variation during the climb) against expected climb
schedules. Figures 12.36 (earlier in the chapter) and 12.46 are examples of
validation climbs. The test should maintain the tolerances specified earlier.
 CHAPTER 12 Climb Performance 561

As specified previously, sources of scatter include:

• Wind speed and direction variation vs altitude

• Temperature variation with altitude and temperature at a specific altitude
• Turbulence
• Airplane handling for ease of speed tracking

40,000 155,500

35,000 155,000
Altitude
30,000
154,500
Pressure altitude (ft)

25,000
154,000

Weight (lb)
20,000
Weight 153,500
15,000
153,000
10,000

5000 152,500

0 152,000
0 200 400 600 800 1000 1200
Time (s)

40,000 3500

35,000 3000
Altitude
30,000 2500
Pressure altitude (ft)

Rate of climb (fpm)

25,000 2000

20,000 1500

15,000 1000

10,000 500
Rate of
climb
5000 0

0 –500
0 200 400 600 800 1000 1200
Time (s)

Fig. 12.46 Example of a validation climb.

562 Operational Aircraft Performance and Flight Test Practices

Climb schedule:
Engine
Time Altitude OAT Ground speed Speed Inlet pressure Inlet temp. Fuel flow
(HH:MM:SS) (ft) (°C) (KTGS) (RPM) (psi) (°C) (PPH)

Fig. 12.47 Example of data recording table for climb test.

We recommend that data be recorded at regular time intervals, every 30 s or

1 min, rather than at altitude intervals. (This will make it easier for data col-
lection, especially at the initial high rate of climb.) The ground speed during
climb (INS or GPS required) should be recorded to compute distance to
climb and adjust for winds (by comparing TAS to GS). Also record tempera-
ture as the aircraft climbs so as to compute TAS from known Mach/CAS (the
climb speed schedule selected), record engine parameters (typically speed,
inlet pressure and temperature, fuel flow, propeller torque, and RPM) so as
to be able to compute the thrust generated, and record start weight and
compute the test weight from the recorded fuel flow. Figure 12.47 shows
an example of a table that could be used to document climb performance
testing.

Required Instrumentation to Support Climb Performance Flight Testing

It is desirable to have the climb parameters recorded on a continuous
basis because the testing usually goes by very quickly, and a significant
amount of data has to be collected. As we reviewed some test techniques,
some important required parameters were listed. As a summary, they are:

• Time
• Airspeed (KCAS, Mach, KTAS, and KTGS, if available)
• It is possible to do without KTGS if repeat accelerations or sawtooth
climbs are done in opposite headings to account for wind gradients.
• Pressure altitude
• Geometric altitude, if possible
• Static air temperature
• Airplane weight
• Airplane configuration should at least be hand recorded by the crew if
not by instrumentation.
• Pitch attitude and flight path angle
• Bank angle and load factor are also useful for long climbs where heading
change may be required.
• Engine parameters (those required to compute thrust, usually the
following: engine RPM, inlet pressure and temperature, power extraction
and bleed air flow, etc.) and fuel flow
 CHAPTER 12 Climb Performance 563

Data Analysis and Corrections

The aim of the flight test program for climb performance is to be able to
model the airplane’s specific excess thrust capability and be able to reliably
predict performance under varying atmospheric and flight conditions. The
test conditions may have more variability than when performing the test
for drag in level flight. One must plan for this variability and apply the
proper corrections.
While testing for climb performance, expect that larger variation in air-
plane weight and the high pitch attitude may render the calibration of the
fuel weight less reliable, so ensure the airplane stabilizes for a minute or so
prior to a climb to get a fix. Postflight, correct the weights as was done for
the cruise drag testing.

Geometric vs Pressure Altitude

The preferred instrument for measuring altitude is the barometric alti-
meter, which provides an adjusted pressure altitude (using the barometric
setting, see Chapter 2). If one were to fly in the equivalent of a standard
atmosphere (temperature and pressure vs altitude), then the equation can
be used for climb performance. If not, the rate of climb derived from a
pressure altitude measurement (ROC) must be converted to geometric (tape-
line) altitude rate of climb (ROCg ) for climb performance analysis. This is
done in the following way:
 
T
ROCg ¼ ROC (12:46)
TISA
where T is the temperature at the test-measured pressure altitude, and TISA is
the standard atmosphere temperature at the same pressure altitude. By
observation, a warmer temperature at altitude will correct the pressure alti-
tude rate of climb to a higher geometric altitude rate of climb. The geometric
climb gradient is then
ROCg
ggr ¼ (12:47)
VGS
where VGS is the true ground speed. Alternatively, using a GPS or an inertial
reference unit (IRU) as a source of altitude change will provide a direct
measurement of the true geometric altitude gain, which eliminates the
need to apply a correction to the measured pressure altitude gain.
Using these systems for the measurement of climb performance can help
reduce test scope. They offer the opportunity to not do a reciprocal heading
because they provide the wind component; therefore, it is possible to reduce
testing by half. I have yet to see a performance engineer refuse test data,
however, so one may still want to do two climbs in the same configur-
ation/altitude band to confirm the data (reduce scatter) in the models.
564 Operational Aircraft Performance and Flight Test Practices

Acceleration in a Climb and Winds

We discussed the impact of acceleration on climb performance earlier in
this chapter. Acceleration is expected during a climb because one must
account for the aircraft acceleration when flying at a constant CAS or
Mach (not at a constant TAS) as well as the normal test variation. The test
needs to characterize the airplane’s specific excess thrust. We can rewrite
Eq. (12.30) as
 
ðT  DÞ ðdh=dt Þ V dV
¼ 1þ (12:48)
W V g dh
The acceleration factor [(V/g) (dV/dh)] will be of the order of magnitude
as shown in Table 12.2. The actual flight test data should now be used to
correct the climb performance. We applied this correction in Example
12.2, shown earlier in the chapter.

Presenting the Information to a Flight Crew

There are a lot of variables to account for in a climb, temperature aloft
being a major one, especially for long climbs. These conditions will not be
constant throughout the climb. (The deviation of the temperature from stan-
dard will not be constant on most climbs and descents.) The performance
engineer must provide sufficient information for the crew to be able to deter-
mine the proposed top of climb to accomplish a mission and what would be
the expected time to climb, fuel to climb, and distance to climb. We consider
the AEO information here. The charts presented in this section are meant to
complement the cruise charts shown in Chapter 11.
The first part of the flight planning manual will be the airplane’s ability to
climb. The information can be provided in a table format, as in Fig. 12.48, or a
chart format, as in Fig. 12.49. The aim of these charts and tables is to provide
a quick reference to the airplane’s ability to achieve a given rate of climb at
altitude. Some pilots, aiming to fly as high as possible, may elect to climb
until the rate of climb is as low at 100 fpm; others are more comfortable
with slightly higher rates of climb at top of climb (like 300 fpm).
The climb conditions (speed schedule, weight, temperature, altitude,
power and propeller settings, airplane configuration) should be clearly speci-
fied. One of the advantages of the chart format in providing a summary of the
climb capability of an airplane is to clearly show the impact of multiple climb
speeds in a climb schedule (Fig. 12.49b). On that chart, as the airplane goes
from a climb speed of 250 KIAS to 310 KIAS, the acceleration factor is higher,
and the overall rate of climb reduces. When the airplane transitions to con-
stant climb Mach number (0.73 M), the rate of climb increases (as expected)
and allows the airplane to achieve a higher cruise altitude before it reaches
the lower values (e.g., 300 fpm) that would be used to judge maximum
altitude.
 CHAPTER 12 Climb Performance 565

All engines operating–climb rate for 3935 LBS (maximum continuous)

Temp °C 10 15 20 25 30 35 40 45 50
Press. alt. (ft) Rate of climb (ft/min)
0 1137 1119 1101 1083 1065 1048 1030 1013 995
1000 1137 1119 1101 1083 1065 1048 1030 1013 995
2000 1137 1119 1101 1083 1065 1048 1030 1013 995
3000 1137 1119 1101 1083 1065 1048 1030 1013 995
3500 1137 1119 1101 1083 1065 1048 1030 1013 995
4000 1100 1082 1064 1046 1028 1011 993 975 958
5000 1027 1009 990 972 954 936 918 901 883
6000 954 935 917 898 880 862 844 826 808
7000 880 861 842 824 805 787 769 751 733
8000 806 787 768 749 731 712 694 676 658
9000 732 713 694 675 656 637 619 600 582
10000 658 639 619 600 581 562 544 525 507
11000 584 564 544 525 506 487 468 449 431
12000 509 489 470 450 431 412 393 374 355
13000 434 414 395 375 355 336 317 298 279
14000 360 339 319 300 280 261 241 222 204
15000 285 264 244 224 204 185 166 147 128
16000 209 189 169 149 129 109 90 71 52
17000 134 114 93 73 53 33 14 –5 –24
18000 59 38 18 –3 –23 –43 –62 –81 –101

Fig. 12.48 Example of rate of climb capability in table format.

Another format one can find in the industry is shown in Fig. 12.50. This
shows a carpet plot of equivalent engine power as a function of pressure alti-
tude and temperature (left side of the figure) that is converted to a rate of
climb (right side of the figure) if the airplane is in the right configuration

Fig. 12.49 Example of rate of climb capability in chart format.

566 Operational Aircraft Performance and Flight Test Practices

Maximum rate of climb – flaps 0°

(Standard units)
Altitude-ft Airspeed
0 130 KIAS Example:
5000 125 KIAS Altitude 5000 ft
Associated conditions: 15000 125 KIAS
OAT 22 °C
Maximum climb power 20000 120 KIAS
30000 120 KIAS
Aircraft weight 8000 LB
Landing gear retracted Rate of climb 1875 FPM
3000
Min OAT limit

Max OAT limit

Reference line
2500

2000

–40

Rate of climb - FPM

–30 1500

0
–20 5000 1000
10,000
–10 15,000 e-ft
d 500
0 20,000 ltitu
e a
+10 25,000 sur
ISA +20 s
–30,000 Pre 0
- °C +30

–500
–60 –40 –20 0 20 40 60 11,000 10,000 9000 8000 7000 6000
Outside air temperature - °C Aircraft weight - LB

Fig. 12.50 Example of rate of climb capability in chart format.

(engine power setting, gear retracted, flaps retracted) and following the
proper climb speed schedule.
From a flight planning point of view, the crew must be provided with the
ability to plan the time, distance, and fuel to climb. This also can be provided
in chart or table format. We first examine typical chart formats. Figure 12.51
shows an integrated chart (multiple scales) to provide the three components
(time, distance, fuel) for a given atmospheric condition (standard day for the
baseline chart with correction notes at the bottom left of the chart). For this
airplane, the climb schedule (left scale) is a decreasing calibrated airspeed
with increasing altitude. The chart is created assuming a climb from sea
level at a starting weight, and the curves include the effect of fuel burn
(weight reduction) during the climb. The highly curved path shown in
Fig. 12.51 is one of the reasons the validation climb testing provides a
benefit to check the integrated model.
Instead of using a chart format, the manufacturer can use a table format.
When using this approach, more pages must be provided in the flight plan-
ning manual because one cannot combine as many conditions on one page.
Figure 12.52 shows the climb capability of an airplane at various temperature
conditions, but for a single initial climb weight. This format must be repeated
several times to cover the appropriate weight range for the airplane.
 CHAPTER 12 Climb Performance 567

Standard day
Data as of: 1 Oct 1959 Military thrust climb profile Model: F-86F
Based on: Flight test Two 200-gallon drop tanks Engine: J47-27
CAS True Mach (1000 feet)
(knots) number altitude
B 0 LB
LB 0L 120
195 0.81 50 00 110
B 10 0 LB
0L 130
LB 90

sed
LB
80
0 1400

el u
215 0.80 45 LB
1500

B
f fu

0L
LB

70
0

Bo
160

0L
60
235 0.78 40

B L
500
260 0.77 35
Service celling

Optimum range altitude

400 LB
280 0.74 30
Combat ceiling
Cumb speed

300 0.71 25 Sea level

300 LB
gross weights
13,000 LB
14,000 LB
315 0.68 20 16,000 LB
18,000 LB
200 LB

20,000 LB
22,000 LB
335 0.66 15
100 LB

350 0.63 10

370 0.61 5

390 0.59 5L
Remarks: 0 20 40 60 80 100 120 140 160 180 200 220 240
1. All data computed from time best climb speed is obtained. Air distance-nautical miles
2. For each 10°C rise in air temperature above standard
day conditions:
(1) increase gross weight 1500 pounds before Time–Minutes
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
entering chart.
(2) or use actual weight and: increase time, distance
and fuel 15%, decrease ceilings 200 ft
3. For each 10°C below standard day conditions, apply F-86F-1-93-208B
correction in the opposite direction. Maintain Mach number as shown.

Fig. 12.51 Climb flight planning chart example. Source: Flight Manual F-86F, May 1960.

Weight ISA -10°C ISA ISA +10°C ISA +15°C ISA +20°C
15,000 LB Time Dist Fuel Time Dist Fuel Time Dist Fuel Time Dist Fuel Time Dist Fuel
Min. N.M. Lb Min. N.M. Lb Min. N.M. Lb Min. N.M. Lb Min. N.M. Lb
51 25.4 157.5 631.0 27.4 173.8 684.7
49 19.7 120.7 553.9 21.0 131.2 596.4 38.9 254.1 859.5
47 16.5 99.6 505.7 17.7 109.1 545.6 25.5 162.2 682.0 39.9 262.2 903.8
45 14.2 84.7 468.8 15.4 93.8 507.0 20.7 130.0 612.0 27.0 172.4 726.1
43 12.5 73.4 438.0 13.7 82.0 474.6 17.9 110.4 564.8 22.3 140.3 654.3 31.1 199.9 820.3
41 11.1 64.3 410.9 12.2 72.5 446.1 15.8 96.4 527.2 19.4 120.3 604.6 25.8 162.7 735.1
39 9.9 56.7 386.4 11.1 64.6 420.3 14.3 85.6 495.1 17.4 105.8 564.6 22.5 139.8 676.7
37 9.0 50.3 363.6 10.1 57.9 396.3 13.0 76.9 466.6 15.8 94.7 530.6 20.2 123.8 631.3
Pressure altitude –1000 feet

35 8.2 45.1 343.3 9.3 52.5 374.9 12.0 70.0 441.7 14.5 86.2 501.6 18.5 112.0 594.3
33 7.5 40.6 324.0 8.5 47.5 354.0 11.1 63.7 416.9 13.4 78.3 472.6 17.0 101.1 557.2
31 6.7 35.7 301.4 7.7 42.0 329.0 10.0 56.2 385.9 12.0 68.6 435.4 15.0 87.4 508.2
29 6.1 31.5 279.3 7.0 37.1 304.5 8.9 49.2 355.0 10.7 59.6 398.2 13.2 74.7 460.0
27 5.5 27.9 258.6 6.3 32.8 281.3 8.0 43.1 326.1 9.5 51.8 363.9 11.6 64.1 417.0
25 5.0 24.8 239.0 5.7 29.0 259.1 7.2 37.7 298.8 8.5 45.1 331.9 10.2 55.3 377.7
23 4.6 22.2 220.1 5.1 25.7 237.7 6.4 33.0 272.6 7.5 39.2 301.6 9.0 47.6 341.2
21 4.1 19.8 201.7 4.6 22.6 217.0 5.7 28.8 247.3 6.6 33.9 272.5 7.9 40.9 306.8
19 3.7 17.6 183.7 4.1 19.9 196.8 5.0 24.9 222.7 5.8 29.2 244.5 6.9 35.0 273.9
17 3.4 15.7 166.2 3.7 17.3 177.0 4.4 21.4 198.7 5.1 24.9 217.3 6.0 29.7 242.4
15 3.0 13.9 148.9 3.2 15.0 157.4 3.8 18.2 175.0 4.3 21.0 190.7 5.1 25.0 211.9
13 2.7 12.2 131.4 2.8 12.9 137.9 3.2 15.2 151.7 3.7 17.5 164.6 4.3 20.7 182.4
11 2.3 10.4 113.3 2.4 10.9 118.5 2.7 12.5 128.6 3.0 14.2 138.9 3.5 16.8 153.5
9 2.0 8.7 94.6 2.0 9.0 99.0 2.2 9.9 105.6 2.4 11.2 113.5 2.8 13.2 125.2
7 1.6 6.8 75.2 1.6 7.1 78.8 1.7 7.5 82.6 1.9 8.4 88.3 2.2 9.9 97.1
5 1.2 4.9 54.9 1.2 5.1 57.4 1.2 5.3 59.7 1.3 5.8 63.1 1.5 6.8 69.2
3 0.7 3.0 33.6 0.7 3.1 35.1 0.7 3.2 36.5 0.8 3.4 38.0 0.9 3.9 41.4
1 0.2 1.0 11.4 0.2 1.0 11.9 0.2 1.1 12.4 0.3 1.1 12.7 0.3 1.3 13.8
Climb speed: 250 KIAS up to 32,000 feet.
0.70 Mi above 32,000 feet.

Fig. 12.52 Example of climb planning chart, single weight.

568 Operational Aircraft Performance and Flight Test Practices

250/300KCAS/Mach0.75 climb at maximum climb power ISA

Altitude (ft) Initial climb weight (lb)
OAT (deg C) 95,000 93,000 91,000 89,000 87,000 85,000
45,000 Time (min) **** **** **** **** **** ****
–56.5 Fuel (lb) **** **** **** **** **** ****
Distance (nm) **** **** **** **** **** ****
43,000 Time (min) **** **** **** **** 25.9 23.7
–56.5 Fuel (lb) **** **** **** **** 2515 2355
Distance (nm) **** **** **** **** 174 158
41,000 Time (min) 25.6 23.8 22.4 21.2 20.1 19.2
–56.5 Fuel (lb) 2644 2499 2375 2270 2175 2085
Distance (nm) 170 158 148 139 132 126
39,000 Time (min) 20.5 19.6 18.8 18.1 17.4 16.7
–56.5 Fuel (lb) 2312 2228 2145 2070 1995 1925
Distance (nm) 134 128 122 117 113 108
37,000 Time (min) 17.9 17.2 16.6 16.1 15.5 15
–56.5 Fuel (lb) 2127 2057 1990 1925 1860 1800
Distance (nm) 115 111 107 103 99 96
35,000 Time (min) 16.1 15.6 15.1 14.6 14.1 13.7
–54.3 Fuel (lb) 1990 1929 1870 1810 1755 1695
Distance (nm) 102 99 96 93 89 86
33,000 Time (min) 14.8 14.3 13.9 13.4 13 12.6
–50.4 Fuel (lb) 1873 1817 1765 1710 1655 1605
Distance (nm) 92 90 87 84 81 79
31,000 Time (min) 13.6 13.2 12.8 12.4 12 11.6
–46.4 Fuel (lb) 1763 1712 1660 1615 1565 1515
Distance (nm) 84 81 79 76 74 71
25,000 Time (min) 10.1 9.8 9.5 9.3 9 8.7
–34.5 Fuel (lb) 1405 1365 1330 1290 1250 1215
Distance (nm) 58 57 55 53 52 50
20,000 Time (min) 7.4 7.2 7 6.8 6.7 6.5
–24.6 Fuel (lb) 1094 1065 1035 1005 980 950
Distance (nm) 40 39 38 37 36 34
15,000 Time (min) 5.2 5.1 5 4.8 4.7 4.6
–14.7 Fuel (lb) 809 790 765 745 725 705
Distance (nm) 26 25 24 24 23 22
10,000 Time (min) 3 2.9 2.9 2.8 2.7 2.6
–4.8 Fuel (lb) 492 485 465 455 440 430
Distance (nm) 13 13 13 12 12 12
5000 Time (min) 1.4 1.4 1.4 1.3 1.3 1.3
5.1 Fuel (lb) 247 240 235 230 220 215
Distance (nm) 6 6 6 6 6 5

Fig. 12.53 Example of climb planning chart, ISA conditions.

Another approach is to cover multiple weights in one table for a given

temperature condition. Figure 12.53 provides such an example for standard
day (ISA) climb conditions. Then this table must be repeated several times to
cover the temperature range.
The advantage of table format over chart format is, of course, that the
time, distance, and fuel values are clearly spelled out, the manufacturer’s per-
formance engineer having done the work. The data are more precise, but is
this precision really needed given the variability in predicting a climb from
low altitude to high altitude? The format is really at the manufacturer’s
discretion.
 CHAPTER 12 Climb Performance 569

Descent performance information can follow the same general formats as

the climb conditions. The fuel burn is much less; therefore, the weight will be
almost constant, so the presentation of the information could be simplified
down to a format similar to Fig. 12.54. Again, the importance of providing
the information lies in the planning of the flight capability of the airplane
from engine start to engine shutdown.

Time, fuel and distance to descend

Associated conditions Example
Power As required to Initial altitude 11,500 ft
maintain 500 ft/min Final altitude 3986 ft
Rate-of descent Time to descend (23–8) = 15 min
Landing gear up Fuel to descend (8.3–2.8) = 5.5 GAL
Flaps up Distance to descend (170–23 = 47 NM
Descent speed
170 KTS
(196 MPH)

16,000
15,000
14,000
13,000
12,000
11,000
Pressure altitude ~feet

10,000
9000
8000
7000
6000
5000
4000
3000
2000
1000
SL
0 5 10 15 20 25 30 35 40
Time to descend ~ minutes

0 1 2 3 4 5 6 7 8 9 10 11 12
Fuel to descend ~ gallons

0 10 20 30 40 50 60 70 80 90 100
Distance to descend ~ NM

Fig. 12.54 Simplified format for descent performance.

570 Operational Aircraft Performance and Flight Test Practices

Exercise
(1) A jetliner is climbing to its cruise altitude. You have recorded the
following information:
• Knots true airspeed (KTAS)
• Knots true ground speed (KTGS)
• Knots calibrated airspeed (KCAS)
• Pressure altitude and GPS height
• Static air temperature
• Pitch attitude and flight path angle
• Airplane weight
Compute the airplane climb performance (ROCg , excess thrust) between
event lines 1 and 2, and between 3 and 4.

Pressure altitude (ft) GPS height (ft) SAT (°C)

–20
34
Pressure altitude (ft)
GPS height (ft) (103)

–38.28783 28925.23 HGPS –25

–34.17133
32 27019.87
30 –30
HP

SAT (°c)
–29.12792
28 –35
23219.8 25294.07 28524.74 –40
26
24 –45
26597.7 –42.82727 SAT
22 24862.24
–50
22782.9
20 –55
KCAS (kt) KTAS (kt) KTGS (kt)

400 381.66
KTAS (kt), KTGS (kt)

371.7578 KTAS
380 351.3923 359.6848

360
KCAS (ft)

340 KTGS
320 344.1396
327.396 340.5786
300 321.7826
280 249.8468 247.7842 249.0078
249.6088 KCAS
260
240
Pitch (deg) FPA (deg) Weight (lb)

8
5.227052 5.082728 5.145094 100.4
4.088583
6
Pitch (deg)
FPA (deg)

(lb) (103)

100.2
Weight

Pitch
4 100.0
FPA
2 4.210521 4.047953 2.879553
4.2027
99.8
100150 99975 99925 99800 Weight
0 99.6
24 1

03 2

22 3

01 4
0. ne

0. ne

0. ne

0. ne

200 250 300 350 400 450 500

05

44

34

72
25 t li

30 t li

35 t li

40 t li
en

en

en

en
Ev

Ev

Ev

Ev

Time (s)
Chapter 13 Climb
Capability OEI

Chapter Objective
Is it a coincidence that the first time we discuss engine failure is in Chapter 13?
Engine failures are a fact of life. For Part 25 airplanes, the loss of one engine in
flight should not represent more than a major event (increased pilot workload)
with continued safe flight possible. This is exemplified by the introduction of
Extended-Range Twin-Engine Operational Performance Standards (ETOPS)
requirements in the late 1980s in which twin-engine jetliners were granted
(after careful design and demonstration of capabilities) long-range route plan-
ning with up to 330 min (5.5 h) on one engine! We discuss the impact of losing
thrust from one engine on the ability to climb and on the ability to avoid
obstacles. We also review driftdown (when the ability to maintain the cruise
altitude is impacted) and gliding flights (following the loss of all thrust).

Impact of One Engine Inoperative (OEI) on Climb Performance

C
ertification requirements are set to ensure a minimum safety level for
the airplane. The ability to continue a climb following an engine
failure for Part 25 airplanes and for commercial operation for Part
23–certified airplanes is part of the minimum safety requirements. It pro-
vides the crew the time required to plan the safe return to land.
The time after an engine failure can still be challenging for the crew, and
misleading information or miscommunication in the cockpit can still lead to
an accident, as shown in the TransAsia Airways ATR72 accident (see acci-
dent box). The records of Part 25 and Part 23 commuter category (now
Level 4) certified airplanes have shown that these events are extremely few
and far apart; most engine failures result in a “simple” loss of thrust and
safe return to land.

Aviation Safety Council Taiwan, Aviation Occurrence Report, no.

ASC-AOR-16-06-001: On 4 Feb. 2015, about 1054 Taipei Local Time, Trans-
Asia Airways (TNA) flight GE 235, an ATR-GIE Avions de Transport
Régional ATR72-212A (ATR72-600) aircraft, registered B-22816, experienced

571
572 Operational Aircraft Performance and Flight Test Practices

a loss of control during initial climb and impacted Keelung River, 3 n miles
east from its departing runway 10 of Taipei’s Songshan Airport; 43 occupants
were killed.
During the initial climb after takeoff, an intermittent discontinuity in
engine number 2’s auto feather unit (AFU) may have caused the automatic
takeoff power control system (ATPCS) sequence that resulted in the uncom-
manded autofeather of engine number 2 propellers. Following the uncom-
manded autofeather of engine number 2 propellers, the flight crew did not
perform the documented abnormal and emergency procedures to identify
the failure and implement the required corrective actions. This led the pilot
flying (PF) to retard power of the operative engine number 1 and ultimately
shut it down. The loss of thrust during the initial climb and inappropriate
flight control inputs by the PF generated a series of stall warnings, including
activation of the stick shaker and pusher. The crew did not respond to the
stall warnings in a timely and effective manner. The aircraft stalled and con-
tinued descent during the attempted engine restart.
Eng 1 shut
Eng 2 fail down

Continued
climb

Basic OEI Impact on Climb Performance

Most of the certification requirements for climb performance focuses on
takeoff and the continued safe flight following an engine failure at the critical
part of the takeoff run. This is not to say that other phases of flight are less
critical, but rather reflects the much increased workload vs the available
time to react and take action of the takeoff phase, including the possibility
of having to clear obstacles in the flight path. Part 25 requirements define
the takeoff path as follows (author’s underline emphasizes the need to
follow an approved procedure):

§25.111 Takeoff path

(a) The takeoff path extends from a standing start to a point in the takeoff at
which the airplane is 1,500 feet above the takeoff surface, or at which the
transition from the takeoff to the en route configuration is completed
and VFTO is reached, whichever point is higher. In addition—
(1) The takeoff path must be based on the procedures prescribed in
§25.101(f);
 CHAPTER 13 Climb Capability OEI 573

(2) The airplane must be accelerated on the ground to VEF, at which

point the critical engine must be made inoperative and remain
inoperative for the rest of the takeoff; and
(3) After reaching VEF, the airplane must be accelerated to V2.

We offer Fig. 13.1 as an example of a takeoff following an engine failure at a

critical time. This takeoff is executed using the manufacturer’s defined
takeoff procedure. The engine failure occurs at time 62 s, as indicated by
the rapid reduction in N1 (the engine fan speed) just a little before rotation
speed. The thrust decays rapidly, forcing the pilot to react using a rudder
deflection to maintain the runway heading. At the same time, the pilot
must focus on continuing the takeoff procedure (capturing the target OEI
takeoff pitch and speed). As the airplane lifts off, some roll activity will be
noticed because the gear ground reaction is removed and the rudder-induced
roll must be compensated by the ailerons. The bank angle close to the ground
must be kept low (near wings level) to avoid wing tips hitting the ground, and
some additional rudder deflection may be required to maintain runway
heading as the airplane climbs away from the ground. All of these activities
occur within a few seconds of each other. Good takeoff procedures will
reduce the pilot workload. We will cover the takeoff in more detail in Chap-
ters 15 and 16, and we will focus this chapter on the airplane’s ability to climb
away from the ground following an engine failure.
To analyze the airplane’s ability to climb following an engine failure, we
will break down the event into multiple components and discuss each one.

Engine Failures and Windmilling Drag

An engine failure will, most of the time, result in some form of shutdown
of the engine and the loss of thrust. The most extreme conditions will result
in a destructive failure of the engine (see Fig. 13.2), but these are extremely
rare cases; this is not accounted for in defining the climb capability of the air-
plane from a certification point of view.
Gas turbine engines will rarely seize up when they fail; rather, they will
windmill when they are shut down (fuel flow stopped). Windmilling
removes energy from the flow, creating a drag force (see Fig. 13.3). It will
also force the extra air around the engine (spilled air) and reduce the overall
energy of the airflow, which can alter the aircraft drag and even lift. Windmill-
ing drag Dw is typically proportional to the Mach number and the altitude.
Engine manufacturers are responsible for determining this drag (for a
given engine/nacelle combination) and can provide the results in table
format or a curve fit of the data in equation form. This drag must be added
to the aircraft all-engine-operating (AEO) drag when an engine failure occurs.
A reciprocating engine will tend to stop rotating quickly, freezing the pro-
peller in place. It will not windmill unless the airplane is at a high airspeed
and sufficient energy is applied to the propeller to make it turn.
Propeller disks offer a very large surface to the airstream and will create
relatively more drag. Variable pitch propellers will adopt a fine pitch position
(see Fig. 13.4) to improve thrust generation. Following an engine failure, this
 574
Operational Aircraft Performance and Flight Test Practices
KCAS (kt) Radalt (ft)
Engine failure
150 500
400

Radalt (ft)
KCAS (kt) 100 300

Lift off
200
50 100
0
0 –100
N1 (%) Rudder (%) Sideslip (deg)
100
4

Sideslip (deg)
50
Rudder (%)

2
N1 (%)

0 0
–50 Some sideslip –2
Part rudder deflection –4
–100
Aileron (%) Bank (deg)
50 2
40 1
Aileron (%)

Bank (deg)
30 0
20 –1
10 –2
0 Part aileron deflection –3
–10 –4

Pitch (deg) Heading (deg)

10 45

Heading (deg)
Pitch (deg)

5 40
Maintain heading
0 35
–5 30
40 50 60 70 80 90 100 110 120
Time (s)

Fig. 13.1 Initial climb during a takeoff following an engine failure.

 CHAPTER 13 Climb Capability OEI 575

Engine control CTAI duct Inner barrel splice plates

unit cooling tube
thermal shield 11 o’clock

Inner fan cowl

Containment shield

Inner barrel 5 o’clock

backskin
Boeing 737-700 N772SW, NTSB report DCA18MA142 Airbus A380 F-HPJE, BEA report June 2019

Fig. 13.2 Examples of severe failure of turbofans.

Lower energy
Spilled air V
V
Lower energy

Fig. 13.3 Windmilling drag.

Thrust
factor
Forward thrust

Coarse Fine
(high speed) (takeoff low speed)

Reverse
(after landing)

⌬tw ⌬tf
Feathered
Time
(power failure)
from VEF
Windmilling
drag

Fig. 13.4 Propeller windmilling.

576 Operational Aircraft Performance and Flight Test Practices

fine pitch will essentially be a flat plate at 90 deg to the airflow and generating
significant drag; this propeller will windmill (be rotated by the incoming
airflow) if the resistance of the engine is low (e.g., gas turbines). Many air-
planes are equipped with an automatic feathering system that, once the
engine failure is sensed, the propeller adopts a low drag position where the
blades are essentially aligned with the airflow. On some airplanes, this func-
tion is manual and must be done by the pilot.
Airplanes with electric motors introduce interesting scenarios. The
motor can degrade more gently, losing one of three phases that, depending
on the motor design, can result in little power loss. The motor could fail com-
pletely and provide some resistance to the propeller windmilling, or it could
fail in regenerative mode (battery charging mode) and extract more power
from the airflow. The airplane manufacturer must understand the failure
modes of the motor and determine the most adverse mode for the various
climb segments of the airplane.
Under the new Part 23 requirements, an acceptable means of compliance
for use is ASTM F3179, Standard Specification for Performance of Aircraft
states (underline added by author):
Loss of thrust – Thrust—For conventional aeroplanes (reciprocating or
turbine engine-powered), loss of thrust means one engine inoperative. For
other aeroplanes, the amount of thrust loss shall be proposed by the appli-
cant and accepted by the CAA.

Control Drag
An asymmetric thrust condition must be compensated for by the deflec-
tion of flight controls to restore the moments about all axes to zero. An
additional drag component results, in addition to the windmilling drag,
that also must be accounted for. This drag will be a function of the asym-
metric thrust [the difference between the operating engine(s) and the wind-
milling drag of the inoperative engine(s)].
The yawing moment generated by the asymmetric thrust is represented
by a moment coefficient CN defined as
yeng ðT  DW Þ
CN ¼ 1
(13:1)
2 rSL s V 2 S b
where
yeng ¼ Distance from center line of the failed engine and opposite engine
b ¼ Wing span

The equation uses only one yeng value because airplanes are typically built
to be symmetric with respect to the centerline; if that is not the case, a dis-
tance appropriate to each engine must be used. The equation also holds
true for a four-engine airplane, like the one shown in Fig. 13.5, or even for
an eight-engine airplane like the Boeing B-52, because typically, the
 CHAPTER 13 Climb Capability OEI 577

T2
T3 CDyaw
T1

Failed
T4

yeng
DW

CN

␦R

Fig. 13.5 Control drag as a function of yawing moment coefficient.

engines not impacted (like the two inboard engines in Fig. 13.5) have the
same thrust value and cancel out in yawing moment.
On propeller-driven airplanes, slipstream flowing around the wing will result
in the loss of lift on the wing with engine failure, which may require significant
aileron/spoileron deflection that will compound the control drag effects.
When testing is performed with an asymmetric engine condition, the test
engineer collects the asymmetric thrust condition (to compute the yawing
moment) and the drag of the airplane DOEI . Then, the yaw drag is determined
from flight testing by flying various yawing moments (various asymmetric
T/W ) with one engine shut down (windmilling). Even with only reduced
thrust on one engine this has to provide enough data to properly define
the yaw drag coefficient chart of Fig. 13.5. Then,

Dyaw ¼ DOEI  ðDAEO  DW Þ (13:2)

Dyaw
CDyaw ¼ (13:3)
1
r s V2 S
2 SL
The equation is meant to capture all the effects associated with the asym-
metric thrust condition, including the increased drag due to sideslip, the
reduced forward thrust due to sideslip [T cos(b)], and the impact of flight
controls deflection on drag (see Fig. 13.5). This can be particularly significant
if roll spoilers are used at the typical low bank angles required for an
engine-out condition.
578 Operational Aircraft Performance and Flight Test Practices

The Mitsubishi MU-2 is one example where roll control input impacts
drag because, by design, it has no ailerons. The airplane is equipped with
overwing spoilers instead of conventional ailerons for roll control (see
Fig. 13.6). This allowed Mitsubishi to use full-span double-slotted flaps on
the trailing edge of the wing to improve stall speeds.

Certification Regulations: Part 25 Transport Category Airplanes

The pilot of a multiengine airplane must continue the takeoff if an engine
fails beyond V1 (defined in Chapter 16) or risk exceeding other limits such as
running out of runway or exceeding brake energy limits. All Part 25 airplanes
and most Part 23 multi-engine airplanes are certified to meet regulations
designed to ensure an airplane will be able to continue climbing following
the loss of the most critical engine at the most critical time of the takeoff
run. There is no single-engine airplane certified to Part 25.
To ensure that the airplane will be able to climb with one engine inopera-
tive, the manufacturer must determine the maximum weight, altitude, and
temperature (WAT) combination that will allow the aircraft to meet
minimum climb gradients defined in Parts 23 and 25.
The climb gradients are grouped into specific segments (see Fig. 13.7),
defined as follows:
• First segment: Extending from liftoff to the point where the gear is
fully retracted
• Second segment: With gear up and takeoff flaps, climb to at least 400 ft
AGL
• Third segment: Flap retraction and acceleration to final takeoff climb
• Final segment: Climb flap retracted to 1500 ft AGL

Fig. 13.6 Mitsubishi MU-2 roll spoilers instead of ailerons.

 CHAPTER 13 Climb Capability OEI 579

Level off, retract Climb at VFTO

Climb at V2 flaps and
Liftoff Gear retracted accelerate to VFTO
First segment Second segment Third segment Final segment

Fig. 13.7 Segmented climb.

The following sections provide an overview of the various regulations of

Part 25 and Part 23 that define the minimum gradients expected for the air-
planes certified under these requirements. This will also show how the
maximum takeoff (and landing) weight based on minimum climb capability
is defined.

First Segment Climb

The first segment climb requirements are designed to bring the airplane
away from the ground and, if applicable, retract the gear. (Note: There are no
fixed-gear Part 25 airplanes at this time.) During part of this time (or all of it),
the airplane will be in ground effects, which is commonly accepted to be of a
height within one wingspan. (Also codified in §25.111: The airplane is con-
sidered to be out of the ground effect when it reaches a height equal to its
wing span.) The minimum climb requirements for Part 25 first segment
climb are defined by §25.121 (see Fig. 13.8) (author’s underline):

§25.111 Takeoff path

(c) During the takeoff path determination in accordance with paragraphs(a)
and (b) of this section—
(1) The slope of the airborne part of the takeoff path must be positive at
each point;
(2) The airplane must reach V2 before it is 35 feet above the takeoff
surface and must continue at a speed as close as practical to, but not
less than V2, until it is 400 feet above the takeoff surface;
(4) The airplane configuration may not be changed, except for gear
retraction and automatic propeller feathering, and no change in
power or thrust that requires action by the pilot may be made until
the airplane is 400 feet above the takeoff surface;

§25.121 Climb: One-engine-inoperative

(a) Takeoff; landing gear extended. In the critical takeoff configuration
existing along the flight path (between the points at which the airplane
reaches VLOF and at which the landing gear is fully retracted) and in the
configuration used in §25.111 but without ground effect, the steady
gradient of climb must be positive for two-engine airplanes, and not less
580 Operational Aircraft Performance and Flight Test Practices

than 0.3 percent for three-engine airplanes or 0.5 percent for four-engine
airplanes, at VLOF and with—
(1) The critical engine inoperative and the remaining engines at the
power or thrust available when retraction of the landing gear is
begun in accordance with §25.111 unless there is a more critical
power operating condition existing later along the flight path but
before the point at which the landing gear is fully retracted; and
(2) The weight equal to the weight existing when retraction of the
landing gear is begun, determined under §25.111.

Note how little climb capability is allowed by the requirement—for

twin-engine airplanes, anything greater than zero is acceptable. This is accep-
table because the gradient is defined as out of ground effects. The airplane is
expected to be able to start the climb when in ground effects due to the lower
induced drag of the airplane. (See the discussion in Chapter 19 on ground
effects.) This makes for an interesting problem for the manufacturers of air-
planes with wing spans of less than approximately 50 ft because an airplane
could, in theory, be “stuck” in ground effect while waiting for the landing gear
to retract. The takeoff distance, as we will see in Chapter 15, is based on the
ability of the airplane to reach 35 ft (Part 25) and V2 . Airplanes with small
wing spans that are first segment limited could have the takeoff distance dic-
tated by the time needed to retract the landing gear. A closer look at Fig. 13.8
shows how much distance can rapidly be covered at the minimum climb gra-
dients; the benefit of ground effects at 35 ft above ground (end of the takeoff
distance) for an airplane with a wing span of 50 ft or less has essentially van-
ished (see Fig. 19.10 in Chapter 19).

160

140

120
.5%
s, 0
100 ine
ng
Height (ft)

ure
Fo
80
.3 %
gi n es, 0
60 en
e e
Thr
40

20
Two engines, 0%
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Distance (n miles)

Fig. 13.8 Part 25 first segment climb requirements.

 CHAPTER 13 Climb Capability OEI 581

WARNING

The propeller on the inoperative engine must be

feathered, landing gear retracted and wing flaps up
or continued flight may be impossible.

Fig. 13.9 Warning statement of a 1970s light Part 23 twin-engine airplane.

Part 23 airplanes certified prior to 1987 had no formal requirements to

meet for first segment climb. In fact, some light twin-engine airplanes
would actually start losing altitude following a takeoff with one engine inop-
erative, as exemplified by the statement from a pilot operating manual of an
older Part 23 certified airplane shown in Fig. 13.9. In 1996, FAA Part 23 at
Amendment 23-50 introduced a requirement, §25.66, to at least document
the capability of the airplane following an engine failure. The statement
from the notice of proposed rule making (NPRM) indicated:
Proposed new Sec. 23.66 would require the determination of the one-engine-
inoperative climb capacity of all WAT limited reciprocating engine-powered
and turbine engine-powered airplanes immediately after takeoff. Since
most reciprocating engine-powered airplanes do not have autofeather, the
condition immediately after takeoff can be critical. There is not a minimum
climb requirement in this configuration, only the determination of the climb
or descent gradient. This information does not become a limitation; it is
provided to the pilot in the AFM (see Sec. 23.1587) to allow the pilot to make
informed judgments before takeoff.

§23.66 [Takeoff climb: one engine inoperative]

[For normal, utility, and acrobatic category reciprocating engine-powered

airplanes of more than 6,000 pounds maximum weight, and turbine
engine-powered airplanes in the normal, utility, and acrobatic category, the
steady gradient of climb or descent must be determined at each weight,
altitude, and ambient temperature within the operational limits established
by the applicant with—
(a) The critical engine inoperative and its propeller in the position it rapidly
and automatically assumes;
(b) The remaining engine(s) at takeoff power;
(c) The landing gear extended, except that if the landing gear can be
retracted in not more than seven seconds, the test may be conducted
with the gear retracted;
(d) The wing flaps in the takeoff position(s);
(e) The wings level; and
( f) A climb speed equal to that achieved at 50 feet in the demonstration of
Sec. 23.53.]

In 1987, FAA Part 23, Amendment 23-34, the FAA introduced the commuter
category for propeller-driven airplanes. This was to support the introduction
of turboprop airplanes designed to carry up to 19 passengers. This was
expanded to jet-driven airplanes in Amendment 23-62 in 2012. The commu-
ter category airplane was defined by §23.3 as:
582 Operational Aircraft Performance and Flight Test Practices

§23.3 Airplane categories. (amendment 23-34)

(d) The commuter category is limited to propeller-driven, multiengine air-
planes that have a seating configuration, excluding pilot seats, of 19 or
less, and a maximum certificated takeoff weight of 19,000 pounds or
less, intended for nonacrobatic operation as described in paragraph (a)
of this section.

§23.3 Airplane categories. (amendment 23-62)

(d) The commuter category is limited to multiengine airplanes that
have a seating configuration, excluding pilot seats, of 19 or less, and a
maximum certificated takeoff weight of 19,000 pounds or less. The com-
muter category operation is limited to any maneuver incident to normal
flying, stalls (except whip stalls), and steep turns, in which the angle of
bank is not more than 60 degrees.

For the commuter category airplanes, FAA Part 23 defined a requirement for
first segment climb that mimics the Part 25 requirements:
§23.67 Climb: One engine inoperative
(d) For jets over 6,000 pounds maximum weight in the normal, utility and
acrobatic category and commuter category airplanes, the following
apply:
(1) Takeoff; landing gear extended. The steady gradient of climb at the
altitude of the takeoff surface must be measurably positive for
two-engine airplanes, not less than 0.3 percent for three-engine
airplanes, or 0.5 percent for four-engine airplanes with—
(i) The critical engine inoperative and its propeller in the position it
rapidly and automatically assumes;
(ii) The remaining engine(s) at takeoff power;
(iii) The landing gear extended, and all landing gear doors open;
(iv) The wing flaps in the takeoff position(s);
(v) The wings level; and
(vi) A climb speed equal to V2.

As discussed earlier, Part 23 has evolved in Amendment 23-64 to become

nonprescriptive. There is no formal first segment climb requirements for
this new amendment, but one acceptable means of compliance for climb is
specified by ASTM F3179 sections 12, 14, and 15, which specify:

ASTM F3179 Section 12. Climb General

12.1 Compliance with the climb requirements (Sections 13– 16 and 20) shall
be shown:
12.1.1 Out-of-ground effect;
12.1.2 At speeds that are not less than those at which compliance with
the powerplant cooling requirements has been demonstrated.
12.1.3 Unless otherwise specified, with critical loss of thrust at a bank
angle not exceeding 58.

ASTM F3179 Section 14. Takeoff Climb—Partial Loss of Thrust

14.1 For Level 3 low-speed multiengine aeroplanes, the steady gradient of
climb or descent shall be determined at each weight, altitude, and
ambient temperature within the operational limits established by the
applicant with:
 CHAPTER 13 Climb Capability OEI 583

14.1.1 The critical loss of thrust, including any propulsive drag changes
that are rapidly and automatically assumed;
14.1.2 The remaining engine(s) at takeoff power;
14.1.3 The landing gear extended, except that if the landing gear can be
retracted in not more than 7 s, the test may be conducted with
the gear retracted;
14.1.4 The wing flaps in the takeoff position(s);
14.1.5 The wings level;
14.1.6 A climb speed equal to that achieved at 15 m [50 ft] in the
demonstration of Section 7.

ASTM Section 15. Climb after Partial Loss of Thrust

15.3 For Level 3 high-speed multiengine aeroplanes and all Level 4
multiengine aeroplanes, the following apply:
15.3.1 Takeoff—Landing Gear Extended—The steady gradient of climb
at the altitude of the takeoff surface shall be measurably positive
with:
15.3.1.1 The critical loss of thrust including any propulsive drag
changes that are rapidly and automatically assumed, if
applicable;
15.3.1.2 The remaining engine(s) at takeoff power;
15.3.1.3 The landing gear extended and all landing gear doors
open;
15.3.1.4 The wing flaps in the takeoff position(s);
15.3.1.5 The wings level;
15.3.1.6 A climb speed equal to V2.

These are similar to §23.66 and §23.67 for the commuter category pre–
Amendment 23-64 of FAA Part 23. Section 15 does apply for out-of-
ground-effects climb capability.

Second Segment Climb

Second segment climb represents a step change in drag with the
landing gear now retracted. The focus of the segment is to get the airplane
away from the ground to allow the crew to plan for the follow-on step of
safe return to land. During this segment, the flaps are still extended and
the pilot is following a prescribed airspeed. Again, all Part 25 airplanes
have retractable landing gear, but several Part 23 airplanes do not; this is
reflected in the requirements. For Part 25, they are (underline added by
author):

§25.111 Takeoff path

(c) During the takeoff path determination in accordance with paragraphs (a)
and (b) of this section—
(1) The slope of the airborne part of the takeoff path must be positive at
each point;
(2) The airplane must reach V2 before it is 35 feet above the takeoff
surface and must continue at a speed as close as practical to, but not
less than V2, until it is 400 feet above the takeoff surface;
584 Operational Aircraft Performance and Flight Test Practices

(4) The airplane configuration may not be changed, except for gear
retraction and automatic propeller feathering, and no change in
power or thrust that requires action by the pilot may be made until
the airplane is 400 feet above the takeoff surface;

§25.121 Climb: One-engine-inoperative

(b) Takeoff; landing gear retracted. In the takeoff configuration existing at

the point of the flight path at which the landing gear is fully retracted,
and in the configuration used in §25.111 but without ground effect:
(1) The steady gradient of climb may not be less than 2.4 percent for
two-engine airplanes, 2.7 percent for three-engine airplanes, and 3.0
percent for four-engine airplanes, at V2 with:
(i) The critical engine inoperative, the remaining engines at the
takeoff power or thrust available at the time the landing gear is
fully retracted, determined under §25.111, unless there is a more
critical power operating condition existing later along the flight
path but before the point where the airplane reaches a height of
400 feet above the takeoff surface; and
(ii) The weight equal to the weight existing when the airplane’s
landing gear is fully retracted, determined under §25.111.
(2) The requirements of paragraph (b)(1) of this section must be met:
(i) In non-icing conditions; and
(ii) In icing conditions with the most critical of the takeoff ice
accretion(s) defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g), if in the configuration
used to show compliance with §25.121(b) with this takeoff ice
accretion:
(A) The stall speed at maximum takeoff weight exceeds that in
non-icing conditions by more than the greater of 3 knots
CAS or 3 percent of VSR; or
(B) The degradation of the gradient of climb determined in
accordance with §25.121(b) is greater than one-half of the
applicable actual-to-net takeoff flight path gradient
reduction defined in §25.115(b).

These stated gradients are much larger than those for the first segment climb,
and the airplane will gain significantly more altitude than for the first
segment climb requirements; however, these are still low values. A 2.4%
climb gradient corresponds to a flight path angle of 1.375 deg. An airplane
having a takeoff speed of 120 KTAS would have a rate of climb of about
290 ft/min. It would take over 1 min, 20 s to reach an altitude of 400 ft
above ground, a very long time under a one engine inoperative condition
where the crew workload is increased. This is, of course, the minimum
climb gradient authorized; an airplane having a combination of weight, alti-
tude, and temperature (WAT) that would result in a climb gradient smaller
than this requirement would not be authorized to take off because the
assumption of engine failure at the critical time is always made for Part 25
airplanes. The distance covered during these low climb gradients (see
Fig. 13.10) also requires the airplane to not face any obstacles. We will
address this subject later in this chapter.
 CHAPTER 13 Climb Capability OEI 585

1000

900
%
800 3.0 7%
i n es, , 2.
e ng gines
700 ur n
Fo ree e %
Th 2.4
600 i n es,
Height (ft)

ng
oe
500 Tw
Minimum height above ground
400

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Distance (n miles)

Fig. 13.10 Part 25 second segment climb gradients.

For FAA Part 23, pre–Amendment 23-64, the second climb segment
requirements were emphasized with the introduction of the commuter cat-
egory. These are (author’s underline):
§23.67 Climb: One engine inoperative

(d) For jets over 6,000 pounds maximum weight in the normal, utility and
acrobatic category and commuter category airplanes, the following
apply:
(2) Takeoff, landing gear retracted. The steady gradient of climb at an
altitude of 400 feet above the takeoff surface must not be less than 2.0
percent for two-engine airplanes, 2.3 percent for three-engine
airplanes, and 2.6 percent for four-engine airplanes with—
(i) The critical engine inoperative and its propeller in the position it
rapidly and automatically assumes;
(ii) The remaining engine(s) at takeoff power;
(iii) The landing gear retracted;
(iv) The wing flaps in the takeoff position(s);
(v) A climb speed equal to V2.

With the new Part 23, similar requirements are stated in §23.2120, which are
similar to the previous Part 23 requirements (again, author’s underline):
§23.2120 Climb requirements
The design must comply with the following minimum climb performance
out of ground effect:

(b) After a critical loss of thrust on multiengine airplanes—

(2) For levels 1 and 2 high-speed airplanes, and level 3 low-speed
airplanes, a 1 percent climb gradient at 400 feet (122 meters) above
586 Operational Aircraft Performance and Flight Test Practices

the takeoff surface with the landing gear retracted and flaps in the
takeoff configuration(s); and
(3) For level 3 high-speed airplanes and all level 4 airplanes, a 2 percent
climb gradient at 400 feet (122 meters) above the takeoff surface with
the landing gear retracted and flaps in the approach configuration(s).

With an acceptable means of compliance per ASTM F3179:

ASTM F3179 Section 15. Climb after Partial Loss of Thrust
15.3.2 Takeoff—Landing Gear Retracted—The steady gradient of climb at
an altitude of 122 m [400 ft] above the takeoff surface shall not be less
than 2.0 % with:
15.3.2.1 The critical loss of thrust including any propulsive drag
changes that are rapidly and automatically assumed, if
applicable;
15.3.2.2 The remaining engine(s) at takeoff power;
15.3.2.3 The landing gear retracted;
15.3.2.4 The wing flaps in the takeoff position(s);
15.3.2.5 A climb speed equal to V2.

Again, second segment is meant to bring the airplane safely (minimum climb
gradient) to at least 400 ft above the runway.

Third and Final Segments Climb

For these segments, the airplane is meant to climb to a normal pattern
altitude while accelerating to higher airspeed and “cleaning up” (going to
flaps retracted). This phase of flight can start as low as 400 ft above
runway altitude. Part of this climb may actually be done in level flight
(while the airplane accelerates), so it does require an acceleration of a mag-
nitude equivalent to a climb gradient specified by the regulations of §§25.111
and 25.121 (author’s underline):
§25.111 Takeoff path
(c) During the takeoff path determination in accordance with paragraphs (a)
and (b) of this section—
(1) The slope of the airborne part of the takeoff path must be positive at
each point;
(2) The airplane must reach V2 before it is 35 feet above the takeoff
surface and must continue at a speed as close as practical to, but not
less than V2, until it is 400 feet above the takeoff surface;
(3) At each point along the takeoff path, starting at the point at which
the airplane reaches 400 feet above the takeoff surface, the available
gradient of climb may not be less than—
(i) 1.2 percent for two-engine airplanes;
(ii) 1.5 percent for three-engine airplanes; and
(iii) 1.7 percent for four-engine airplanes.
(4) The airplane configuration may not be changed, except for gear
retraction and automatic propeller feathering, and no change in
power or thrust that requires action by the pilot may be made until
the airplane is 400 feet above the takeoff surface; and
(5) If Sec. 25.105(a)(2) requires the takeoff path to be determined for
flight in icing conditions, the airborne part of the takeoff must be
based on the airplane drag:
 CHAPTER 13 Climb Capability OEI 587

(i) With the most critical of the takeoff ice accretion(s) defined in
Appendices C and O of this part, as applicable, in accordance
with § 25.21(g), from a height of 35 feet above the takeoff surface
up to the point where the airplane is 400 feet above the takeoff
surface; and
(ii) With the most critical of the final takeoff ice accretion(s) defined
in Appendices C and O of this part, as applicable, in accordance
with § 25.21(g), from the point where the airplane is 400 feet
above the takeoff surface to the end of the takeoff path.
§25.121 Climb: One-engine-inoperative

(c) Final takeoff. In the en route configuration at the end of the takeoff path
determined in accordance with §25.111:
(1) The steady gradient of climb may not be less than 1.2 percent for
two-engine airplanes, 1.5 percent for three-engine airplanes, and 1.7
percent for four-engine airplanes, at VFTO with
(i) The critical engine inoperative and the remaining engines at the
available maximum continuous power or thrust; and
(ii) The weight equal to the weight existing at the end of the takeoff
path, determined under §25.111.
(2) The requirements of paragraph (c)(1) of this section must be met:
(i) In non-icing conditions; and
(ii) In icing conditions with the most critical of the final takeoff ice
accretion(s) defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g), if in the configuration
used to show compliance with §25.121(b) with the takeoff ice
accretion used to show compliance with §25.111(c)(5)(i):
(A) The stall speed at maximum takeoff weight exceeds that in
non-icing conditions by more than the greater of 3 knots
CAS or 3 percent of VSR; or
(B) The degradation of the gradient of climb determined in
accordance with §25.121(b) is greater than one-half of the
applicable actual-to-net takeoff flight path gradient
reduction defined in §25.115(b).

A similar requirement exists for Part 23 pre–Amendment 23-64:

§23.67 Climb: One engine inoperative
(d) For jets over 6,000 pounds maximum weight in the normal, utility and
acrobatic category and commuter category airplanes, the following apply:
(3) Enroute. The steady gradient of climb at an altitude of 1,500 feet above
the takeoff or landing surface, as appropriate, must be not less than 1.2
percent for two-engine airplanes, 1.5 percent for three-engine
airplanes, and 1.7 percent for four-engine airplanes with—
(i) The critical engine inoperative and its propeller in the minimum
drag position;
(ii) The remaining engine(s) at not more than maximum
continuous power;
(iii) The landing gear retracted;
(iv) The wing flaps retracted; and
(v) A climb speed not less than 1.2 VS1.

For the new Part 23, the acceptable means of compliance with ASTM F3179
points to similar performance:
588 Operational Aircraft Performance and Flight Test Practices

ASTM F3179 Section 15. Climb after Partial Loss of Thrust

15.3.3 En Route—The steady gradient of climb at an altitude of 457 m
[1500 ft] above the takeoff or landing surface, as appropriate, shall be
not less than 1.2 % with:
15.3.3.1 The critical loss of thrust including any propulsive drag
changes that are rapidly assumed, if applicable;
15.3.3.2 The remaining engine(s) at not more than maximum
continuous power;
15.3.3.3 The landing gear retracted;
15.3.3.4 The wing flaps retracted;
15.3.3.5 A climb speed not less than 1.2 VS1

We will cover more of the “en route” requirements later in this chapter.

Approach Climb
The approach climb configuration is when the airplane is flying under
OEI conditions with the go-around flaps (most often the approach flap con-
figuration). This is meant to define a maximum allowable landing weight
based on the capability to perform a go-around and meet a minimum
climb gradient. The airplane is then limited by the most restrictive climb con-
dition of either the OEI approach flap of §25.121 or the all-engines-operating
requirements of §25.119, discussed in Chapter 12 (author’s underline).
§25.121 Climb: One-engine-inoperative
(d) Approach. In a configuration corresponding to the normal all-engines-
operating procedure in which VSR for this configuration does not exceed
110 percent of the VSR for the related all-engines-operating landing
configuration:
(1) The steady gradient of climb may not be less than 2.1 percent for
two-engine airplanes, 2.4 percent for three-engine airplanes, and 2.7
percent for four-engine airplanes, with—
(i) The critical engine inoperative, the remaining engines at the
go-around power or thrust setting;
(ii) The maximum landing weight;
(iii) A climb speed established in connection with normal landing
procedures, but not exceeding 1.4 VSR; and
(iv) Landing gear retracted.
(2) The requirements of paragraph (d)(1) of this section must be met:
(i) In non-icing conditions; and
(ii) In icing conditions with the most critical of the approach ice
accretion(s) defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g). The climb speed
selected for non-icing conditions may be used if the climb
speed for icing conditions, computed in accordance with
paragraph (d)(1)(iii) of this section, does not exceed that for
non-icing conditions by more than the greater of 3 knots CAS
or 3 percent.

We underlined the 110% requirements to provide additional explanation.

The statement simply says that the stall airspeed of the go-around configur-
ation (approach flap) cannot be more than 10% higher than the stall airspeed
 CHAPTER 13 Climb Capability OEI 589

of the landing flap configuration. The purpose of this limit is to ensure the
airplane does not stall as it starts the go-around maneuver and flaps are
retracted. For airplanes where the difference in stall airspeeds between the
two flaps is more than 110%, the manufacturer can elect to select a higher
VREF (airspeed in the landing configuration) that would result in a similar
110% spread.
The same general requirement is expected of the larger Part 23 airplanes
as shown in §23.67 (pre –Amendment 23-64):
§23.67 Climb: One engine inoperative
(d) For jets over 6,000 pounds maximum weight in the normal, utility
and acrobatic category and commuter category airplanes, the
following apply:
(4) Discontinued approach. The steady gradient of climb at an altitude
of 400 feet above the landing surface must be not less than 2.1
percent for two-engine airplanes, 2.4 percent for three-engine
airplanes, and 2.7 percent for four-engine airplanes, with—
(i) The critical engine inoperative and its propeller in the minimum
drag position;
(ii) The remaining engine(s) at takeoff power;
(iii) Landing gear retracted;
(iv) Wing flaps in the approach position(s) in which for these
position(s) does not exceed 110 percent of the VS1 for the
related all-engines-operating landing position(s); and
(v) A climb speed established in connection with normal landing
procedures but not exceeding 1.5 VS1.

For the new Part 23, the acceptable means of compliance of ASTM F3179
points to similar performance:
ASTM F3179 Section 15. Climb after Partial Loss of Thrust
15.3.4 Discontinued Approach—The steady gradient of climb at an altitude
of 122 m [400 ft] above the landing surface shall be no less than 2.1%
with:
15.3.4.1 The critical loss of thrust including any propulsive drag
changes that are rapidly assumed, if applicable;
15.3.4.2 The remaining engine(s) at takeoff power;
15.3.4.3 Landing gear retracted;
15.3.4.4 Wing flaps in the approach position(s) in which VS1 for
these position(s) does not exceed 110 % of the VS1 for the
related all-engines-operating landing position(s);
15.3.4.5 A climb speed established in connection with normal landing
procedures but not exceeding 1.5 VS1.

Concluding Remarks for Climb Requirements

The overview of the climb requirements following an engine failure for
both Part 23 and Part 25 was meant to provide perspective on what has been
accepted as minimum safe requirements to allow the pilot to safely climb
away from the ground and plan the return to land. They are not exactly the
same for both sets of regulations, but they have achieved a good safety
590 Operational Aircraft Performance and Flight Test Practices

record since their introduction. One must remember that on most flights, the
airplane will have a much greater climb capability than these minimums.

Flight Test and Check Climbs

With the focus of certification requirements being on the applicant being
able to determine the climb capability under OEI conditions, the applicant
must now plan a test campaign that will collect sufficient information to
show compliance. Having OEI conditions will introduce extra risks into the
test campaign that will need to be addressed. The requirements do ask for
out-of-ground effects climb capability with validation of the climb by
continuous takeoff demonstration. We will cover the OEI takeoff in Chapter 16.
The guidance material of FAA AC 25-7D recommends two methods that
can be used to determine the OEI climb performance. The first is the recipro-
cal heading climbs, which is essentially the sawtooth climbs described in
Chapter 12. These climbs are conducted at several thrust-to-weight ratios
in the appropriate configuration with the critical engine shut down; the
aim of the testing is to document and validate the asymmetric thrust
impact on climb performance. The secondary aim of the test is to validate
the impact of one engine shut down on the systems; for example, is there a
loss of flight controls due to one hydraulic system not working? Some man-
ufacturers have shown that they can replicate this effect without actually
shutting down the engine, which, on a two-engine airplane, reduces test risk.
The applicant should try to perform the testing at 90 deg to the wind so
as to minimize the error due to the wind not being necessarily of the same
magnitude for two consecutive climbs. The guidance material also suggests
that reciprocal climbs may not be necessary if inertial corrections are
applied to account for the wind gradients. But as mentioned previously,
this author has yet to find a performance engineer that does not like
data, and two test points per climb configuration does help reduce climb
model data scatter effects.
The testing following this method would proceed in the same way as the
AEO testing was performed (as described in Chapter 12), but the testing
should be expanded to include much lower values of T/W. The applicant
must collect sufficient data to show that the performance model derived
from the testing can be used to validate the lowest expected/limiting climb
condition in service. This can be as low as a zero climb gradient, as discussed
in the first segment climb section!
Another method recommended by FAA AC 25-7D for OEI testing is the
level flight drag polars. For this test, the critical engine is shut down, and OEI
yaw drag testing is performed at various airspeeds (steady state conditions
required to extract yaw drag) with varying thrust inputs. This model may
not cover as much of the T/W range as the reciprocal heading climbs
method, and the guidance recommends that the model be verified by
 CHAPTER 13 Climb Capability OEI 591

performing reciprocal heading check climbs with wings near level. This is
especially true for two-engine airplanes.
This second test method may still be valuable for airplanes with three or
more engines because the thrust on the nontest engine(s) can be adjusted to
maintain the altitude and airspeed of choice while testing a greater range of
T/W. Then the check climb testing can be reduced to validating the critical
WAT climbs condition.
Check climbs are the ultimate climb validation conditions in which the
airplane is in the certification configuration (flap and gear position, most
adverse CG, target weight, propeller in the appropriate configuration, and
engines at the intended maximum thrust/power condition) and are typically
the flight conditions that authorities want to see in flight testing for validation
of the climb model. FAA AC 23-8C provides additional guidance for
propeller-driven airplanes where asymmetric thrust may also have greater
aerodynamic implications than simply reacting to a yawing moment. The
AC states:
Propeller airplanes are susceptible to slipstream drag, and all airplanes are
susceptible to trim drag. This is most noticeable on airplanes with wing-
mounted engines and when one engine is inoperative. Care should be
given so that drag results are not extended from one flight condition to
another. The power and trim conditions must remain very close to those
existing for the actual test conditions. Drag results are only as accurate as
the available power information and propeller efficiency information. The
cooling airflow through the engine is also a factor.

Test Scope
The test scope for OEI, as we said earlier, should be enough to document
the lowest certifiable climb gradient (i.e., 0% for the first segment, 2.4% for the
second segment, etc.) and validate higher thrust conditions that result in
higher yaw drag [Eq. (13.1)]. For the minimum climb capability, the
testing can be done with reduced thrust on the operating engine at
lower altitude, or it can be done at the higher altitude where the thrust will
naturally be lower. The difference between the two is that at the lower alti-
tude, the thrust is achieved at a lower thrust setting (N1 for gas turbines,
RPM for reciprocating engines and electric motors) whereas at higher alti-
tude, the maximum RPM is used and, as we saw in Chapter 8, the character-
istics of the motor will differ. A combination of low- and high-altitude
testing may be useful to show that the applicant has a good understanding
of the engine model. In all cases, one should try to cover the highest
weights (near maximum takeoff weight if possible) to achieve the highest
thrust setting for maximum asymmetric thrust while also having lower
climb gradients.
Figure 13.11 provides an example of a data collection format that can sup-
plement the continuous recording method. On smaller planes with limited
instrumentation, this may be the only data available to the performance
592 Operational Aircraft Performance and Flight Test Practices

Engines
Weight Flap Gear Altitude Airspeed OAT RoC Engine off Fan speed Inlet presure Inlet temp Fuel flow
(lb) (up/dn) (ft) (KCAS) (°C) (fpm) (L/R) (N1) (psi) (°C) (pph)

Fig. 13.11 Example of OEI climb table to be used for testing.

engineer to create a model to show compliance. The exact content of the

table should account for this fact. If more data are continuously recorded,
then the table becomes a guide to find the test data in the data stream; if
minimal data are recorded, the table shown in Fig. 13.11 will contain every-
thing required to create the model.
You will note that rate of climb (ROC) is included in the table instead of
the climb gradient. The ROC value is readily available to the crew and thus
easy to record. The performance engineer can provide the expected ROC
to the test crew for the test conditions of interest so as to get immediate
posttest feedback from the crew. Alternatively, the test procedure could
specify to the crew that thrust shall be set so as to achieve a given ROC to
ensure sufficient combinations of T/W are covered.
For first segment climb, the airplane should be configured in the takeoff
condition to be tested with the gear down and stabilizer at the appropriate
takeoff stabilizer setting for the CG being tested. There is a test requirement
difference between Part 23 and Part 25.
For Part 23, FAA AC 23-8C specifies that the climb performance tests
with landing gear extended in accordance with §23.67(c) should be con-
ducted with the landing gear and gear doors extended in the most unfavor-
able in-transit drag position. This means that the applicant may have to
artificially lock the gear doors open to perform this test. The AC adds, “If
it is evident that a more critical transitional configuration exists, such as
directional rotation of the gear, testing should be conducted in that configur-
ation.” It is also understood that, should the special configuration lead to
unacceptable test risk (risk of gear not coming back to normal landing con-
figuration), the AC allows for correction by analysis. “In all cases where the
critical configuration occurs during a transition phase that cannot be main-
tained except by special or extraordinary procedures, it is permissible to
apply corrections based on other test data or acceptable analysis.”
For Part 25, FAA AC 25-7D recommends that the climb performance
tests with landing gear extended, in accordance with §25.121(a), may be con-
ducted with the landing gear and gear doors in the position they finally
achieve after “gear down” selection. Under Part 25, this is considered accep-
table because the AC specifies that the takeoff path will be determined by
measurement of continuous takeoffs, or checked by continuous takeoffs if
constructed using the segmental method [§25.111(d)], any nonconservatism
arising from the gear doors “closed” climb data will be evident. Also, some
 CHAPTER 13 Climb Capability OEI 593

measure of conservatism is added to the landing gear extended climb per-

formance by the requirement of §25.111(d)(3) for the takeoff path data to
be based on the airplane’s performance without ground effect.

Test Monitoring
Testing for OEI climb performance will result in lower climb angles
than the AEO testing, of course. Additionally, the airplane will not be in sym-
metrical flight anymore. The small angles will amplify the instantaneous rate
of climb variation (see Fig. 13.12), with values that can easily double the mag-
nitude of the ROC as perceived by the crew. Testing should be performed
with a constant thrust setting, target airspeed within approximately 2 to 3
KCAS, bank angle near target (typically near wings-level first segment
climb, no more than 5-deg bank second segment), and heading as steady
as possible (which dictates the amount of rudder input). Then, one can
analyze the output in a way similar to that described in Chapter 12 for the
AEO condition.

Data Reduction and Validation

Once the data have been collected and the points deemed stable enough
to use, one can proceed with the data reduction. The method for data

Altitude (ft) KTAS (ft)

2000 160
1900 155
Altitude (ft)

KTAS (kt)
1800 150
1700 145
1600 140
1500 135
1400 130
FPA (deg) Pitch (deg) RoC (fpm)
12 600
10 500
RoC (fpm)

8
Pitch (deg)
FPA (deg)

6 400
4 300
2 200
0
–2 100
–4 0
Sideslip (deg) Rudder (%) Bank (deg)
100
Sideslip (deg)

4
Rudder (%)
bank (deg)

2 50
0 0
–2 –50
–4
–100
80 90 100 110 120
Time (s)

Fig. 13.12 Example of OEI climb monitoring.

594 Operational Aircraft Performance and Flight Test Practices

1800
FU/GD
1600
TO1/GU
Actual tapeline ROC (ft/min)

1400 TO1/GD
Conservative
1200 TO2/GU
TO2/GD
1000 Ldg/GD
800

600

400

200

0
0 200 400 600 800 1000 1200 1400 1600 1800
Calculated tapeline ROC (ft/min)

Fig. 13.13 Validating the model by flight test.

reduction is the same as described in Chapter 12, but this time, one gets DOEI .
From these data, one can then determine Dyaw [Eq. (13.2)] and create the yaw
drag model [Eq. (13.4)]. If sufficient data are collected, the model as shown in
Fig. 13.5 can be established. It is not necessary to establish the relationship at
all CN ; one just needs enough data to validate the curve for the expected
certification range.
Once the model is established, it must be compared to actual test results
and a best fit (or conservative fit) must be shown. This author likes the format
of Fig. 13.13 because it provides a clear summary of the tests done and fit
obtained. This is used as a summary with authorities followed by detailed dis-
cussion of the analysis performed to obtain the results.

Test Risk and Mitigation

Testing for OEI has the obvious risk that one engine is shut down on
purpose. On a two-engine airplane, that means the airplane is left with
only one engine operating. The main risk for the test comes from losing
the remaining operating engine. As well, the high T/W condition could
result in a high asymmetry condition and loss of control. Some risk mitiga-
tion for OEI climb testing includes:
• Establish the minimum control airspeed prior to the test and validate stall
speeds so as to establish proper operational speeds (V2 , VGA , VREF ) at
which to perform the OEI testing.
• Test at a weight, altitude, and temperature combination that will result in a
minimum climb capability.
 CHAPTER 13 Climb Capability OEI 595

• For twin-engine airplanes, shut down one engine at an altitude that would
allow enough time for the crew to relight the engine should the operating
engine fail during the test. (Bird ingestion is a risk.)
• Perform most of the testing with one engine at idle (to establish yaw
effects) and validate the yaw drag model with a minimum number
of climbs.

Flight Test for Thrust after 8 Seconds

The maximum thrust that can be used for the landing climb (discussed in
Chapter 12) is the thrust available 8 s after initiating a movement of the
power from the minimum flight idle position to the go-around position.
The guidance of FAA AC 25-7D specifies the following procedure to
determine this thrust:
1. At the most adverse test altitude, not to exceed the maximum landing
altitude plus 1500 ft, and with the most adverse bleed configuration,
stabilize in level flight in the landing configuration at 1.23 VSR0 .
2. Retard the throttle to idle and let stabilize to flight idle in a descent at 1.23
VSR0 . Once flight idle has been achieved, rapidly advance the throttle to
the go-around setting. The thrust that is available after 8 s of throttle
movement is the thrust that can be used for §25.119.
The test is meant to capture the impact of slow engine acceleration on an air-
plane’s ability to perform a go-around. The probability of an AEO go-around
at low altitude is greater than that of an OEI (combined engine loss at low
altitude in the approach as well as a go-around, which is in itself a low prob-
ability event), so the authorities have imposed this test to ensure there is a
defined thrust after what is considered an acceptable time (8 s) to ensure
climb and acceleration per procedure. The AEO climb requirements, as
seen in Chapter 12, are of a greater climb angle combined with a possible
lower thrust than maximum possible at the altitude and temperature. This
climb capability is compared to the climb capability at the same weight,
with an OEI and approach flap [as described earlier, §25.121(d)] and lower
climb angle. Whichever gives the lower weight capability becomes the limit-
ing weight for landing based on the airplane’s ability to climb.
This requirement to verify the engine’s ability to accelerate to the higher
RPM to provide thrust was codified with the introduction of Part 25 in 1965
to address the slower acceleration of gas turbine engines (see Fig. 13.14). If a
gas turbine is accelerated too quickly, the engine risks surging and possibly
failing following a surge. Initial gas turbine engines were responding directly
to the pilot input, which added to the variability. Full authority digital engine
control (FADEC) computers are now widely used, and they ensure the engine
does not exceed any limits, but the control laws (CLAWS) used by the
FADEC may now be the new acceleration limit. This author has seen some
engines with slow acceleration response not necessarily due to the physics
596 Operational Aircraft Performance and Flight Test Practices

100
90
Ai = ␲ D2 Stators

Percent maximum go-around thrust

80

ion
4
rmean

it
70

nd
co
60

ion
0%

it
10
. Rotors

nd
50 m RPM

co
Q

%
40

80

$25.119
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12
Time from TLA movement (s)

Turbine temperature limit

pressure ratio
Compressor

TO

e line
␤ Operating line
w ␤ ␤ S urg
w w
Approach it
U U = 2␲ RPM rmean U t lim
Idle ou
60 w-
r blo
Stalled Normal operation Stalled rne
Bu
% Design mass flow rate

Fig. 13.14 Consideration for thrust after 8 s for gas turbine engines.

(large rotational inertia) but simply because of “tricks” in the CLAWS under
certain operating conditions, not necessarily the same behavior that would be
expected at higher altitude and warmer day conditions. A good technical
review of the CLAWS with the engine manufacturer is required before plan-
ning this test. It is meant to capture the slowest acceleration expected so that
the airplane manufacturer can properly plan the landing WAT-limited con-
ditions. Of course, by the time this test (thrust after 8 s) is done for perform-
ance, the applicant will have performed all the engine handling
characteristics testing required to show safe operation and thus reduce the
test risk to normal operating behavior.

Presenting the Takeoff Weight Limited by Climb Requirements to

a Flight Crew
The data for the maximum takeoff weight limited by climb requirements
are typically provided in a simple table (as in Fig. 13.15) or in a chart format
(as in Fig. 13.16) in the airplane flight manual (AFM). The information pro-
vided is weight, altitude, and temperature. This weight represents the most
restrictive weight in the climb to 1500 ft and acceleration to VFTO that
meets the climb requirements of §25.121 for the selected takeoff
 CHAPTER 13 Climb Capability OEI 597

TAKEOFF WEIGHT LIMIT FLAPS 8°

TEMPERATURE
PRESSURE °F 14 32 50 68 86 104
ALTITUDE °C –10 0 10 20 30 40
(FEET) WEIGHT LIMIT ~ POUNDS
10,000 20,500 20,500 20,500 19,760 17,380
9000 20,500 18,200
8000 19,020
7000 19,840
6000 20,500 18,230
5000 19,050
4000 19,870
3000 20,500
2000
1000
S.L.
These data were determined for dry runway, zero wind, zero runway
gradient, anti-ice off, anti-skid on, autospoilers armed, & APR armed.
DO NOT EXCEED MAXIMUM CERTIFIED TAKEOFF WEIGHT

Fig. 13.15 WAT limit in table format.

Maximum allowable takeoff Gross weight

permitted by takeoff climb requirements
anti-icing off, flaps20° 98
60
0 96
50 2
4 94
Airport 5
40 pressure altitude 6 92
–1000 ft 8
30 10 90
Airport ambient temperature – °C

88
Climb limited take-off mass (100 lb)

20 12
14 86
15
10
84

82 0
0
1
80
–10 2
0 f t)

78 3
(100

0
–20
76
tion

5 4
leva

–30 74 Te 10 5
m
rt e

p r 15
72 el. 6
o

–40 to
Airp

ISA 20
70 (C 7
) 25
–50
60 65 70 75 80 85 90 95 68 30 8
–1000 LB
66
20 30 40 50 60 70 80 90 100 110 120 130
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Outside air temperature (F)
–1000 KG
Takeoff Gross weight

Fig. 13.16 WAT limit in chart format.

598 Operational Aircraft Performance and Flight Test Practices

configuration at the pressure altitude and temperature at the time of the

takeoff. The information should be a simple representation of the limit
that is easy to use by the crew. The data presented will typically not indicate
to the crew which of the climb segments is limiting.
The WAT limit becomes the limiting takeoff weight. As we will see later
in this section and in future chapters, the takeoff weight will be limited to the
most restrictive of the following:

• Maximum structural limits (maximum takeoff weight—MTOW)

• Takeoff weight limited by climb (WAT)
• Takeoff weight limited by obstacles
• Takeoff weight limited by runway length
• Takeoff weight limited by brake energy

Obstacle Clearance
Obstacle clearance is a fascinating subject in that it is easy to understand
why it is needed but often hard to clearly put on paper for the crew to use, get
safe values out of the presented information, and at the same time provide
charts that extract the maximum performance out of the airplane. It is the
applicant’s responsibility and obligation to provide a clear and precise (or
conservative) climb chart that will allow an operator to clearly identify the
climb and obstacle clearance capabilities of an aircraft. The initial climb capa-
bility of the airplane, per Part 25, is defined by §25.111 and illustrated by
Fig. 13.17 (author’s underline).

Takeoff Flaps in Flaps

VFTO

AEO OEI
flaps transition up

al
Fin ent En route
Gear Gear gm
VFTO

25.123
V2

Third segment se
down up Min grad
nd t Flap retraction
co ≥1.2%
Min gradient ≥0% Se men acceleration
g 25.121(c)
25.121(a) se
st ≥400 ft ≥1500 ft
Fir ent 25.111(c)(4) 25.111(a)
m
V2

seg
Min gradient ≥1.2%
V=0

Min grad
VLOF
VEF

VR
V1

≥2.4% 25.111(c)(3)
35 ft
25.121(b)

TOFL Takeoff flight path 25.115

25.113

Takeoff path 25.111

Fig. 13.17 Summary of takeoff path—gross climb gradients.

 CHAPTER 13 Climb Capability OEI 599

§25.111 Takeoff path

(a) The takeoff path extends from a standing start to a point in the takeoff at
which the airplane is 1,500 feet above the takeoff surface, or at which the
transition from the takeoff to the en route configuration is completed
and VFTO is reached, whichever point is higher. In addition—
(1) The takeoff path must be based on the procedures prescribed in
§25.101(f);
(2) The airplane must be accelerated on the ground to VEF, at which
point the critical engine must be made inoperative and remain
inoperative for the rest of the takeoff; and
(3) After reaching VEF, the airplane must be accelerated to V2.
(b) During the acceleration to speed V2, the nose gear may be raised off the
ground at a speed not less than VR. However, landing gear retraction
may not be begun until the airplane is airborne.
(c) During the takeoff path determination in accordance with paragraphs (a)
and (b) of this section—
(1) The slope of the airborne part of the takeoff path must be positive at
each point;
(2) The airplane must reach V2 before it is 35 feet above the takeoff
surface and must continue at a speed as close as practical to, but not
less than V2, until it is 400 feet above the takeoff surface;
(3) At each point along the takeoff path, starting at the point at which
the airplane reaches 400 feet above the takeoff surface, the available
gradient of climb may not be less than—
(i) 1.2 percent for two-engine airplanes;
(ii) 1.5 percent for three-engine airplanes; and
(iii) 1.7 percent for four-engine airplanes.
(4) The airplane configuration may not be changed, except for gear
retraction and automatic propeller feathering, and no change in
power or thrust that requires action by the pilot may be made until
the airplane is 400 feet above the takeoff surface; and
(d) The takeoff path must be determined by a continuous demonstrated
takeoff or by synthesis from segments. If the takeoff path is
determined by the segmental method—
(1) The segments must be clearly defined and must be related
to the distinct changes in the configuration, power or thrust, and
speed;
(2) The weight of the airplane, the configuration, and the power or
thrust must be constant throughout each segment and must
correspond to the most critical condition prevailing in the segment;
(3) The flight path must be based on the airplane’s performance without
ground effect; and
(4) The takeoff path data must be checked by continuous demonstrated
takeoffs up to the point at which the airplane is out of ground effect
and its speed is stabilized, to ensure that the path is conservative
relative to the continuous path.
The airplane is considered to be out of the ground effect when it
reaches a height equal to its wingspan.

The synthesis from segments is simply as illustrated in Fig. 13.17 in that

the path is built by the sum of the various segments, determined out of
ground effects, to create a virtual floor for the airplane. That virtual floor is
used to view whether any obstacles penetrate it or if they are safely below.
The takeoff following an engine failure will be a workload-intensive event
600 Operational Aircraft Performance and Flight Test Practices

for the crew. To ensure the airplane does meet this climb floor, clear crew
procedures, as identified by §25.111(a)(1), must be established and validated
by test.
One should expect, based on the requirements of §25.115(a), that the
climb segment begins at the end of the takeoff distance (which we will
discuss in Chapter 15) following an engine failure at the most critical point
of the takeoff run (VEF , Chapter 17). Any other conditions of engine
failure will result in the airplane being higher above ground than the floor
conditions.

§25.115 Takeoff flight path

(a) The takeoff flight path shall be considered to begin 35 feet above the
takeoff surface at the end of the takeoff distance determined in accord-
ance with Sec. 25.113(a) or (b), as appropriate for the runway surface
condition.

The landing gear position has a very large impact on the airplane drag and its
climb capability (Fig. 13.18), so it is imperative that the landing gear be
retracted as soon as possible. So where (altitude) or when would one
expect the landing gear to be retracted? Well, not before liftoff, as stated in
the regulations of §25.111(b): “The landing gear shall not be retracted prior
to lift-off.” The authorities have elected to prescribe that fact in the regulation
following some accidents where the gear was selected up while still on the
ground (more in Chapter 16).
FAA AC 25-7D specifies that the time between liftoff and initiation of
gear retraction during takeoff distance demonstrations should not be less
than that necessary to establish an indicated positive rate of climb plus 1 s.
For the purposes of flight manual expansion, the average demonstrated
time delay between liftoff and initiation of gear retraction may be assumed;
however, this value should not be less than 3 s.
A landing gear can take between 5 and 15 s to retract. (it is a function of
the system’s design, the temperature, and the airspeed of the airplane.) An
initial rate of climb can be anywhere from less than 200 fpm (OEI, near
WAT) to well over 3000 fpm (lightweight AEO), so adding the 3 s after
liftoff before gear is selected up will lead to a gear-retracted height anywhere
from 25 ft AGL to higher than 1000 ft. The AEO lightweight cases typically
are not limiting from a climb point of view, of course, but the most critical
climb condition (OEI) will be. The manufacturer must determine the
longest expected retraction time for the gear and use this in the computation
of the climb floor for obstacle clearance.
Another sizeable impact on climb capability is the availability of
maximum takeoff thrust. When preparing the takeoff path charts, the
effects of thrust variation during the climb must be accounted for.
(Specifically, AC 25-7D mentions, “The effects of changes from takeoff
thrust to maximum continuous thrust should also be included.”) On gas
 CHAPTER 13 Climb Capability OEI 601

Takeoff
AEO OEI
flaps

Gear Gear

V2
down up

nd t
co
Min gradient ≥0% Se men
seg
25.121(a)
≥400 ft
st
Fir ent 25.111(c)(4)
m

V2
seg
V=0

Min grad
VLOF
VEF

VR
V1

≥2.4%
35 ft
25.121(b)

TOFL
25.113

Fig. 13.18 Landing gear retraction height beyond 35 ft AGL.

turbine engines, some airplanes are limited to 5 min of operation at

maximum takeoff thrust, under either AEO or OEI conditions, whereas
others are allowed up to 10 min under OEI (and typically limited to 5 min
for AEO).
When the airplane is near WAT conditions with only 5 min of maximum
takeoff thrust available, it often does not have the ability to transition from
the takeoff configuration to the en route configuration (using maximum con-
tinuous thrust) while maintaining the minimum climb gradient capability
(§25.111 or §23.57) prior to reaching 1500 ft and VFTO. The takeoff pro-
cedure must be such that it ensures the transition is completed prior to
the takeoff thrust time limit.
Figure 13.19 is an example of a climb limited by takeoff thrust avail-
ability for the same takeoff condition (weight, altitude, temperature, flaps,
and OEI). It shows the weight achieved after 10 min of climb from
takeoff, one airplane being limited to 5 min of maximum takeoff thrust
(MTO), the other 10 min. Both airplanes reach VFTO and continue to
climb to VFTO above 1500 ft AGL, and both airplanes were second
segment climb limited (same airplane, only difference was the length of
time MTO was allowed). Note the much steeper average climb gradient
of the second aircraft.
602 Operational Aircraft Performance and Flight Test Practices

3500

Transition to final takeoff

3000 configuration and speed after
reaching 1500 ft AGL. 10 min
2500
Altitude (ft)

2000

1500

1000

500 5 min, transitioned to

MCT and VFTO

0
0 5 10 15 20 25 30
Distance (n miles)

Fig. 13.19 MTO thrust time impact on climb capability.

To account for the variability in the weather, variations in pilot takeoff

technique (did the pilot follow the takeoff climb procedure?), the possibility
that the pilot initiated a turn in the climb, and so forth, the authorities have
prescribed a minimum additional margin of safety over the airplane’s basic
(conservative) climb capability. The airplane’s climb capability, as validated
by flight test, is labeled the gross climb gradient. The requirements of
§25.115 reduce this gross climb gradient by a certain margin to assign a
net climb gradient. The takeoff flight path capability of an airplane is deter-
mined with the net climb gradient (margin incorporated) to ensure safe clear-
ance of obstacles (see Fig. 13.20).

≥400 ft actual
flight path
Net flight path 35 ft
Obstacle flight path

35 ft
Level from
35 point

Fig. 13.20 Net vs gross (actual) climb gradient capability. Source: FAA AC 25-7D.
 CHAPTER 13 Climb Capability OEI 603

§25.115 Takeoff flight path

(b) The net takeoff flight path data must be determined so that they
represent the actual takeoff flight paths (determined in accordance
with Sec. 25.111 and with paragraph (a) of this section) reduced at each
point by a gradient of climb equal to—
(1) 0.8 percent for two-engine airplanes;
(2) 0.9 percent for three-engine airplanes; and
(3) 1.0 percent for four-engine airplanes.
(c) The prescribed reduction in climb gradient may be applied as an
equivalent reduction in acceleration, along that part of the takeoff
flight path at which the airplane is accelerated in level flight.

Now, this may seem like a lot of margins are piled up on top of each other,
and this could be a true statement for an airplane taking off on a clear day
with obstacles clearly in sight. Now imagine one that takes off in the
middle of a dark night or into clouds. You must believe that if you follow
the takeoff procedure you will clear the unseen obstacles. These margins
have so far proven effective.
Similar requirements were specified for Part 23, pre–Amendment 64, for
commuter-category airplanes under §23.61:

§23.61 Takeoff flight path

For each commuter category airplane, the takeoff flight path must be deter-
mined as follows:
(a) The takeoff flight path begins 35 feet above the takeoff surface at
the end of the takeoff distance determined in accordance with §23.59.
(b) The net takeoff flight path data must be determined so that they
represent the actual takeoff flight paths, as determined in accordance
with §23.57 and with paragraph (a) of this section, reduced at each point
by a gradient of climb equal to—
(1) 0.8 percent for two-engine airplanes;
(2) 0.9 percent for three-engine airplanes; and
(3) 1.0 percent for four-engine airplanes.
(c) The prescribed reduction in climb gradient may be applied as an
equivalent reduction in acceleration along that part of the takeoff
flight path at which the airplane is accelerated in level flight.

The new Part 23, Amendment 64, climb requirements are covered by
§23.2125 and an acceptable means of compliance:

§23.2125 Climb information

(a) The applicant must determine climb performance at each weight,
altitude, and ambient temperature within the operating limitations—
(1) For all single-engine airplanes;
(2) For levels 1 and 2 high-speed multiengine airplanes and level 3
multiengine airplanes, following a critical loss of thrust on takeoff in
the initial climb configuration; and
(3) For all multiengine airplanes, during the enroute phase of flight with
all engines operating and after a critical loss of thrust in the
cruise configuration.
604 Operational Aircraft Performance and Flight Test Practices

The acceptable means of compliance include:

ASTM F3179 Section 11 Takeoff Flight Path
11.1 For Levels 1, 2, and 3 high-speed multiengine and all Level 4
multiengine aeroplanes, the takeoff flight path shall be determined as
follows:
11.1.1 The takeoff flight path begins 11 m [35 ft] above the takeoff
surface at the end of the takeoff distance determined in
accordance with Section 10.
11.1.2 The net takeoff flight path data shall be determined so that they
represent the actual takeoff flight paths as determined in
accordance with Section 9 and with 11.1.1 reduced at each point
by a gradient of climb equal to 0.8 %.
11.1.3 The prescribed reduction in climb gradient may be applied as an
equivalent reduction in acceleration along that part of the
takeoff flight path at which the aeroplane is accelerated in
level flight.

On the operational side, one can find additional requirements. As well

as the requirements of §25.115, Part 121 operators have the following
requirement:
§121.189
(d) No person operating a turbine engine powered airplane may take off that
airplane at a weight greater than that listed in the Airplane Flight
Manual—
(2) . . .that allows a net takeoff flight path that clears all obstacles either
by a height of at least 35 feet vertically, or by at least 200 feet horizon-
tally within the airport boundaries and by at least 300 feet horizontally
after passing the boundaries.

Transport Canada Advisory Circular 700-16 [1] provides guidance material

for commercial operators (TCCA Canadian Aviation Regulations—CARs—
Part VII Commercial Air Services) to help them prepare the company oper-
ations manual containing guidance on how to determine OEI climb gradient
and obstacle clearance. The AC rightly notes that the requirements of Part 23
commuter category and Part 25 transport category mandate the performance
of the airplane be limited by OEI climb performance to clear obstacles,
assuming the airplane is at 35 ft AGL at the end of the runway. It also
notes that an airplane taking off on a wet or contaminated runway is
allowed to be runway limited (takeoff distance) with a height at the end of
the runway equal to only 15 ft AGL, 20 ft lower than for takeoff on dry
runways. It offers the image in Fig. 13.21 as a visual aid to operators that
are preparing their manuals to help explain the possible reduced margin in
the initial part of the takeoff (should the airplane be runway length limited
and obstacle limited as well).

En Route Climb Ceiling

The applicant must provide the maximum altitude, in the en route con-
figuration (flaps up, gear up, no more than maximum continuous power or
 CHAPTER 13 Climb Capability OEI 605

Gross flight path

Net flight path
Safety margin is reduced

Start of takeoff flight path

(reference zero)
Terrain
20 ft
15 ft

Fig. 13.21 Takeoff from a wet runway with near obstacles [1].

thrust), for a one-engine-inoperative condition. FAA 14CFR/EASA CS/

TCCA CAR specifies the following for the OEI en route climb:
§25.123 En route flight paths
(b) En route flight paths The one-engine-inoperative net flight path data
must represent the actual climb performance diminished by a gradient of
climb of 1.1 percent for two-engine airplanes, 1.4 percent for
three-engine airplanes, and 1.6 percent for four-engine airplanes.

Simply put, when the airplane reaches the defined climb ceiling, it will still be
capable of climbing under a 1.1% gradient for a twin-engine airplane. The
focus is on ability to climb above an obstacle, not rate of climb. To provide
an example, consider two en route climb speeds, 175 KCAS and 200
KCAS. In a standard atmosphere, an airplane climbing at the 1.1% en
route climb ceiling gradient would have the rate of climb shown in
Fig. 13.22. We note rates of climb in the neighborhood of 200 to 300 fpm,
similar to what would be a normal limit for AEO conditions. But the require-
ment does focus on the ability to clear obstacles (climb gradient). The en
route climb speed is based on OEI and maximum climb gradient; this
tends to be a slower airspeed (most likely below 200 KCAS for most Part
25 airplanes and definitely below 200 KCAS for Part 23 airplanes).
23.67 (c) (3) En-route
The steady gradient of climb at an altitude of 457 m (1500 ft) above the
take-off or landing surface, as appropriate, must be not less than 1.2% with—
(i) The critical engine inoperative and its propeller in the minimum drag
position;
(ii) The remaining engine at not more than MCT;
(iii) The landing gear retracted;
(iv) The wing flaps retracted; and
(v) A climb speed not less than 1.2 VS1.

§23.69 [Enroute climb/descent]

(b) One engine inoperative. The steady gradient and rate of climb/descent
must be determined at each weight, altitude, and ambient temperature
within the operational limits established by the applicant with—
(1) The critical engine inoperative and its propeller in the minimum
drag position;
606 Operational Aircraft Performance and Flight Test Practices

(2) The remaining engine(s) at not more than maximum continuous

power;
(3) The landing gear retracted;
(4) The wing flaps retracted; and
(5) A climb speed not less than 1.2 Vs.

Driftdown
When an engine fails in the cruise phase, the airplane will often be unable to
maintain the altitude. The airplane manufacturer must establish flight pro-
cedures that will allow the crew to continue the flight to a planned alternate des-
tination following the loss of one engine. The procedure will most likely involve
a step to set the remaining engine(s) to maximum continuous thrust/power
(MCT/MCP), and then slow down to a recommended driftdown airspeed
(often the en route climb airspeed) and then allow the airplane to ‘driftdown’
to a lower altitude at a constant airspeed (see Fig. 13.23). The rate of descent
will initially be higher, asymptotically reducing until the airplane nears the en
route climb ceiling for the airplane weight. (The airplane will remain above
because the en route ceiling assumes a 1.1% climb gradient capability.)
If the airplane is in a mountainous area, the procedure may be to maintain
airspeed and engine power, which will naturally make the airplane climb
again as the weight reduces (fuel burn). If no obstacles impact the planned
route, then maintaining the altitude and airspeed may be an ATC require-
ment (see Fig. 13.23).

35,000
En route climb ceiling: 1.1% gradient
ISA conditions
30,000
Venroute
175 KCAS
25,000
Pressure altitude (ft)

20,000
Venroute
200 KCAS
15,000

10,000

5000

0
100 150 200 250 300 350 400
ROC (fpm)

Fig. 13.22 En route climb ceiling.

 CHAPTER 13 Climb Capability OEI 607

Engine(s) at Slow to
MCT/MCP driftdown/en route
airspeed

Maintain airspeed
and power Maintain airspeed
One engine and power
fails
Expected en route
ceiling Maintain airspeed
and altitude

Fig. 13.23 Driftdown procedure.

Another condition where driftdown procedures are important is that

required by the Extended Range Operations (EROPS), as well as ETOPS
for Extended-Range Twin-Engine Operation, FAA §121.162, ETOPS Type
Design Approval, as well as Appendix P to Part 121, and §25.3, Special Pro-
visions for ETOPS Type Design Approvals. The requirements for diversion
under ETOPS will limit the airplane air time following an engine failure.
ETOPS requirements start with a minimum diversion time of 60 min and
typically involve approvals for 75 min, 90 min, 120 min, and 180 min. A
few airplanes have been certified to longer air time than 180 min. To maxi-
mize the range (defined by an airspeed and air time) following an engine
failure, the procedure for the driftdown may now involve the proposed
flight speed (typically much higher than the en route climb speed used for
driftdown over mountains; see Fig. 13.24). A significant portion of the alter-
nate route may be in the driftdown part of the flight. Note that adopting the
higher cruise speed will result in a cruise altitude that is lower, of course, but

Engine(s) at
MCT/MCP Slow to
ETOPS diversion
airspeed

Maintain airspeed
and power
One engine
fails

Maintain airspeed
and altitude
60 minutes 120 minutes 180 minutes

Fig. 13.24 Example of ETOPS procedure over water or nonmountainous terrain.

608 Operational Aircraft Performance and Flight Test Practices

this is often not an issue for ETOPS because most ETOPS routes will occur
over water (though there are a few routes over land).

Guidance on Bank Angle Impact on Climb Gradient

It may not be possible to complete the takeoff and climb in a straight line;
a turn may be required to clear obstacles while in OEI. An airplane in a turn
will see an increase in its load factor, which increases the airplane’s induced
drag; this, in turn, leads to a reduction in the climb gradient. (We will expand
on this in Chapter 14.) The drag equation in a climbing turn would be
2 3
    2
1 6 KW 2 cosðgÞ 7
D¼ rSL s V 2 S CDo þ 4 5 (13:4)
2 1 2 cosðfÞ
r sV S
2 SL
Figure 13.25 depicts a typical impact on the climb gradient capability with
a turn. As can be seen, at the lower airspeeds, an increase in the load factor
due to the bank angle will greatly reduce the climb gradient. If the airplane in
this example had a V2 speed of 125 KIAS, it would have met the minimum
climb gradients of 14 CFR (e.g., 2.4% second segment climb for a twin air-
plane, note: that requirement is for a straight ahead flight) at up to approxi-
mately 35 to 40 deg bank maximum.
For the purpose of takeoff performance, it is often recommended that the
crew use no more than approximately a 15-deg bank in the climb under OEI
conditions until obstacles are cleared. The crew must be provided with the
expected loss of climb capability while executing a turn under the specified
bank angle in the AFM.

15%

10% Bank angle 0 deg

5% 15 deg
30 deg
Gradient

0%
45 deg
–5%

–10%
60 deg

–15%
100 110 120 130 140 150 160 170 180
Airspeed

Fig. 13.25 Impact of bank angle on climb gradient.

 CHAPTER 13 Climb Capability OEI 609

Presenting the Climb Information for Obstacle Clearance to a

Flight Crew
Different manufacturers will use different approaches to show a climb
capability for obstacle avoidance of a given product, and they may even
change the approach from product to product. The applicant can generate
the takeoff flight path information for the AFM by one of two methods:
• Segmented method: Combining the climb gradient capability of the aircraft
as determined by out-of-ground effect during dedicated climb
performance tests.
• Continuous method: Using OEI takeoff flight paths as measured during
dedicated takeoff performance tests. These tests include the favorable
ground effects on the first segment climb performance, thereby increasing
the first segment climb gradient and improving the overall flight path
capability of the aircraft as compared to the segmented method.

For the purpose of obstacle clearance, the segmented method will be used
(validated by continuous takeoff climbs where it is demonstrated that the air-
plane remains above the segmented climb segments at all times). The seg-
mented method is defined by clear procedure and configuration changes.
We first offer some definitions of climb requirements for obstacle clear-
ance to help with our discussion on information to provide to the crew.
• Departure procedures (DPs): Procedures designed to provide obstacle
clearance during instrument departures.
• En route air traffic service (ATS) routes: Routes including VOR and
L/MF-based airways, and VHF omnidirectional range (VOR)/Distance
Measuring Equipment (DME) Area Navigation (RNAV) jet routes that
provide obstacle clearance.
• Standard instrument departure (SID): A preplanned instrument flight rules
(IFR) air traffic control departure procedure published in graphic and
textual form, for the use of pilots and controllers. SIDs provide transition
from runways to the appropriate en route structure.
• Terminal instrument procedures (TERPS): The U.S. Standard for Terminal
Instrument Procedures, FAA Order 8260.3D, is to prescribe the criteria for
the formulation, review, approval, and publishing of procedures for IFR
operations to and from civil and military airports. Transport Canada uses
an equivalent standard, TP 308—Criteria for the Development of
Instrument Procedures, and the ICAO’s equivalent is Procedures for Air
Navigation Services—Aircraft Operations (PANS-OPS). The obstacle
clearance requirements of these standards are based on normal operations
with AEO (see Fig. 13.26). For the purposes of analyzing performance of
procedures developed under TP 308/TERPS/PANS-OPS, it is understood
that any gradient requirement, whether published or not, will be treated as
a plane that must not be penetrated from above until reaching the stated
610 Operational Aircraft Performance and Flight Test Practices

SW-1, 18 JUN 2020 to 16 JUL 2020

(BEVVR1.BEVVR) 19JUL18
BEVVR ONE DEPARTURE (RNAV)

BEVVR ONE DEPARTURE (RNAV)

(BEVVR1.BEVVR) 18312
T ATIS Top altitude: 15,000
DEPARTURE ROUTE DESCRIPTION 135.575
TAKEOFF RUNWAY 25: Begin a climbing left turn as soon as CLNC DEL
124.75
practicable but no later than DER heading 206° to 7100, then DENVER CLNC DEL
direct COPER, then on depicted route to cross APRES at or 124.75
(When lower closed) 14,000
above 13,100, thence. . . . GND CON BELGN
068°
. . . .on (assigned) transition, maintain 15,000. Expect filed 121.8 GGIRO
EAGLE TOWER* (19)
altitude 10 minutes after departure. 119.8
BELGN transition (BEVVR1.BELGN) DENVER CENTER
128.65, 282.2
RUBAY transition (BEVVR1.RUBAY)

(1 7° 0
03 ,00
SPRKT transition (BEVVR1.SPRKT)

14

3)
ZAMRU PARIZ RUBAY
SPRKT 14,000 14,000 14,000
13
,7 077°
29 00

03 1,7 0
(22)
(19 5°

*1 ,10
(1 ° 00
00

AL-6403 (FAA)
13,2 4°

13
)
05 )

7
(12 Note: On departure rapidly rising terrain and trees within

5)
N
GORMN 13 1.5 miles west of the airport and within 0.25 miles
,
13,700 29 100 ZILOP south of the airport.
5 °
(6) ° 054 0 13,200 Note: RNAV 1.
0
13,1 ) Note: GPS required.
EAGLE COUNTY RGNL (EGE)

EAGLE COUNTY RGNL (EGE)

APRES (8 Note: Obstacle protection not ensured for turns delayed
13,100 225K beyond DER.
Note: Do not exceed 210K until established direct COPER.
003°
(7)

EAGLE, COLORADO
6°
EAGLE, COLORADO

BEVVR 20
7100 Takeoff minimums
30 4 )

Rwy 7: NA - Terrain
8°
(

HNDES 263° COPER Rwy 25: 800-2 with a minimum climb of

Note: Chart not to scale (4) 740' per NM to 10,200.
SW-1, 18 JUN 2020 to 16 JUL 2020

Fig. 13.26 Example of DP, Eagle, Colorado. Source: https://flightaware.com/resources/airport/

KEGE/ALL/all/pdf

height, rather than as a gradient that must be exceeded at all points in

the path.
• Engine out procedure (EOP): FAA AC 120-91A, Airport Obstacle Analysis,
provides guidance to develop takeoff and initial climbout airport obstacle
analyses and in-flight procedures to comply with the intent of the
regulatory requirements of Title 14 of the Code of Federal Regulations (14
CFR) Part 121, §§121.177 and 121.189; Part 135, §§135.367, 135.379, and
135.398; and other associated OEI requirements relating to
turbine-engine-powered airplanes operated under Parts 121 and 135.
Although Part 25 provides minimum requirements to determine vertical
obstacle clearance, this AC provides guidance for lateral and horizontal
clearance. EOPs do not need to meet TP 308/TERPS/PANS-OPS. For the
Transport Canada guidance, EOPs are known as engine out departure
procedures (EODPs). Other names for EODPs used by industry include
engine out contingency procedures, engine out escape paths, engine out
SIDS, and emergency escape maneuvers.

An engine failure during takeoff is considered a non-normal condition

and therefore takes precedence over noise abatement, air traffic, SIDs, depar-
ture procedures (DPs), and other normal operating considerations. The pilot
in command has authority to declare an emergency to deviate from
ATC-assigned clearances and instructions. Declaring an emergency may
assist the flight crew in coping with the degraded performance and increased
 CHAPTER 13 Climb Capability OEI 611

workload associated with an engine failure while navigating to remain clear of

obstacles and/or terrain.

Conversions
It is useful to provide a conversion table to the operators to compare the
various climb requirements. The certification requirements are usually given
in percent climb (grad), but departure and obstacle clearance requirements
are often listed in terms of height gained over a given distance. We offer
Table 13.1 as an example of such a table that could be generated.
Another useful conversion chart for the crew is the geometric altitude
gain (height gain) vs pressure altitude gain based on temperature deviation
from standard (see Fig. 13.27). Obstacles are expressed in terms of geo-
metric altitude whereas airplane performance is expressed in terms of
pressure altitude and temperature. The obstacle height must be translated
into pressure altitude gain to find the proper information in the flight
manual.

Table 13.1 Conversion of Various Climb Units

Degrees Grad [%] ft/n miles Degrees Grad [%] ft/n miles
1 1.75% 106 5.2 9.09% 552
1.2 2.09% 127 5.4 9.44% 574
1.4 2.44% 148 5.6 9.79% 595
1.6 2.79% 170 5.8 10.14% 616
1.8 3.14% 191 6 10.49% 637
2 3.49% 212 6.2 10.84% 659
2.2 3.84% 233 6.4 11.19% 680
2.4 4.19% 255 6.6 11.54% 701
2.6 4.54% 276 6.8 11.90% 723
2.8 4.89% 297 7 12.25% 744
3 5.24% 318 7.2 12.60% 766
3.2 5.59% 340 7.4 12.95% 787
3.4 5.94% 361 7.6 13.30% 808
3.6 6.29% 382 7.8 13.66% 830
3.8 6.64% 403 8 14.01% 851
4 6.99% 425 8.2 14.36% 873
4.2 7.34% 446 8.4 14.71% 894
4.4 7.69% 467 8.6 15.07% 915
4.6 8.04% 488 8.8 15.42% 937
4.8 8.39% 510 9 15.77% 958
5 8.74% 531 9.2 16.13% 980
612 Operational Aircraft Performance and Flight Test Practices

Temperature deviation from standard (°C)

10,000
60

40
9000
20

8000 0
–20
7000 –40
Level-off height (LOH) (ft)

–60
6000

5000

4000

3000

2000

1000

0
0 1000 2000 3000 4000 5000 6000 7000 8000
Level-off pressure height (LOHP) (ft)

Fig. 13.27 Height gain vs pressure altitude change.

Net Climb Gradients

The requirements for climb gradient presentation for the AFM are that
the net climb gradient needs to be presented [see §25.1587(b), author’s
underline]. The manufacturer may also elect to present the gross climb gra-
dient as separate information (which may be misleading if not clearly ident-
ified) or simply add a statement to the effect that the gross climb gradient is,
for example, 0.8% greater than the net gradient presented (example for
twin-engine airplanes). So an indicated net value of 1.6% would correspond
to a gross climb capability of 2.4%. The gross gradient would provide the
expected indicated altitude at “level off,” for example, whereas the net gradi-
ent is required to clear all obstacles by at least 35 ft.
 CHAPTER 13 Climb Capability OEI 613

§25.1587 Performance Information

(b) Each Airplane Flight Manual must contain the performance information
computed under the applicable provisions of this part (including
§§25.115, 25.123, and 25.125 for the weights, altitude, temperatures,
wind components, and runway gradients, as applicable) within the
operational limits of the airplane, and must contain the following:
(1) In each case, the conditions of power, configuration, and speeds, and
the procedures for handling the airplane and any system having a
significant effect on the performance information.

The climb gradient information can be presented in chart format (see

Fig. 13.28) or table format (see Fig. 13.29). The chart format is more

Cowl
Wind and wing
component anti-icing
12
11

t
gh
80 wei
10

2,6 oss

)
KG
)
Gr
9
400 foot net gradient – %

0 L eoff KG
8
8 (2 84 G)
4,
k
Ta

B
(2 6K
7 LB 1 G)
7,2
,00

00
0
(2 4K
50

8
55
, B 9, 4 G)
6 0 0 L B (2 5 2K
, 0 L , 7 )
60 000 31 KG
5 5 , L B ( ,020 G)
6 0 34 88 K G)
4 0 ,00 LB ( , 2 6K
7 0 6
(3 ,55 KG)
,00 LB 8
0 Reference line

Off Reference line

3 75 00 B (3 ,824 KG)
, 0 L 0
8 ,00 LB 4 3,092
0 0 (
2 85 ,000 LB (4
90 000
,
1 95
0
–10

10
20
30
40
On

60
Wind – KT Airport 0
PAA 50 pressure 2
minimum altitude 4
1000 ft 6 5
Airport ambient temperature – °C

40
net gradient (1.6 %) 8
30 10
20 12
14
10 15

0
–10
–20
–30
–40
–50

Fig. 13.28 Example of net climb gradient capability chart.

614 Operational Aircraft Performance and Flight Test Practices

First segment net climb gradient (%) and FSH WAI Off
Takeoff weight, LBS (KG)
OAT (°C) HEAD 9000 10,000 11,000 11,500 12,000 12,500 13,000 13,500 13,950
ISA (°C) WIND (KT) (4082) (4536) (4990) (5216) (5443) (5670) (5897) (6123) (6328)
FSH (FT) 128 109 96 90 84 79 74 69 64
–40 °C –10 7.8 6.2 4.8 4.2 3.6 3.0 2.5 2.1 1.7
0 9.0 7.2 5.5 4.8 4.1 3.5 2.9 2.4 1.9
ISA – 55 °C 10 9.5 7.6 5.8 5.0 4.3 3.7 3.0 2.5 2.0
20 10.0 8.0 6.1 5.3 4.5 3.9 3.2 2.6 2.1
30 10.6 8.5 6.5 5.6 4.8 4.1 3.4 2.8 2.2
FSH (FT) 132 112 99 92 87 81 75 70 66
–20 °C –10 7.9 6.3 4.8 4.2 3.6 3.1 2.6 2.1 1.7
0 9.0 7.2 5.5 4.8 4.1 3.5 2.9 2.4 1.9
ISA – 35 °C 10 9.5 7.6 5.8 5.0 4.3 3.7 3.1 2.5 2.0
20 10.0 8.0 6.1 5.3 4.6 3.9 3.2 2.6 2.1
30 10.6 8.5 6.5 5.6 4.8 4.1 3.4 2.8 2.2
FSH (FT) 136 115 101 95 88 83 77 72 67
0 °C –10 7.9 6.3 4.9 4.2 3.6 3.1 2.6 2.1 1.7
0 9.1 7.2 5.5 4.8 4.1 3.5 2.9 2.4 1.9
ISA – 15 °C 10 9.5 7.6 5.8 5.0 4.3 3.7 3.1 2.5 2.0
20 10.0 8.0 6.1 5.3 4.5 3.8 3.2 2.6 2.1
30 10.5 8.4 6.4 5.6 4.8 4.0 3.4 2.7 2.2
FSH (FT) 137 116 102 95 89 83 77 72 67
10 °C –10 7.9 6.3 4.8 4.2 3.6 3.1 2.6 2.1 1.7
0 9.0 7.2 5.5 4.8 4.1 3.5 2.9 2.4 1.9
ISA – 5 °C 10 9.5 7.5 5.8 5.0 4.3 3.6 3.0 2.5 2.0
20 10.0 7.9 6.1 5.3 4.5 3.8 3.2 2.6 2.1
30 10.5 8.4 6.4 5.5 4.7 4.0 3.3 2.7 2.2
FSH (FT) 138 117 103 96 90 84 78 72 67
20 °C –10 7.9 6.3 4.8 4.2 3.6 3.0 2.5 2.1 1.7
0 9.0 7.1 5.5 4.7 4.1 3.4 2.9 2.3 1.9
ISA + 5 °C 10 9.4 7.5 5.7 5.0 4.2 3.6 3.0 2.4 2.0
20 9.9 7.9 6.0 5.2 4.5 3.8 3.1 2.5 2.0
30 10.4 8.3 6.3 5.5 4.7 4.0 3.3 2.7 2.1
FSH (FT) 115 97 84 77 71 65 59 54 49
30 °C –10 6.0 4.5 3.2 2.6 2.1 1.6 1.1 0.7 0.3
0 6.9 5.2 3.6 3.0 2.3 1.8 1.2 0.7 0.3
ISA + 15 °C 10 7.2 5.4 3.8 3.1 2.4 1.8 1.3 0.8 0.3
20 7.6 5.7 4.0 3.2 2.6 1.9 1.3 0.8 0.4
30 8.0 6.0 4.2 3.4 2.7 2.0 1.4 0.8 0.4
FSH (FT) 92 77 64 57 51 46 40
40 °C –10 4.2 2.8 1.6 1.0 0.5 0.1 –0.3
0 4.8 3.2 1.8 1.2 0.6 0.1 –0.4
ISA + 25 °C 10 5.0 3.3 1.9 1.2 0.6 0.1 –0.4
20 5.3 3.5 1.9 1.3 0.7 0.1 –0.4
30 5.6 3.7 2.0 1.3 0.7 0.1 –0.4
Note: Shaded fields indicate conditions exceeding climb limits.

Fig. 13.29 Example of net climb gradient capability table. (FSH represents end of first
segment height.)

concise, presenting most or all of the information on a single page. The vast
amount of data on a single page may lead to some imprecision in the reading
of the actual climb capability of the airplane (but we also know that this
information represents the airplane with its most adverse CG and the
 CHAPTER 13 Climb Capability OEI 615

minimum expected thrust in service, and it is also the net climb capability).
The table format provides a more “precise” value at specifically listed con-
ditions (combination of weight, altitude, temperature, and winds) but still
leaves the pilot the task of interpolating among conditions; this approach
requires a lot more pages in the flight manual. Of course, the advent of
approved computerized flight manuals has greatly simplified the determi-
nation of such information.

Clear Procedure
The flight crew will need clear instructions on how to read the various
charts/tables provided and create their engine out climb path, similar to
that shown in Fig. 13.17 but accounting for the correction to net climb capa-
bility. This can be complicated by allowing the pilot to select the level off
height (LOH) and if the engines provide limited time (5 min) at takeoff
thrust (see Fig. 13.30). Most manufacturers will label the start of the climb
segment for obstacle clearance purposes as “reference zero.” This is the
point that is at 35 ft above ground at the end of the takeoff distance. The
requirements of §25.101 stipulate:
§25.101 General
( f) Unless otherwise prescribed, in determining the accelerate-stop
distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must
be made in accordance with procedures established by the applicant for
operation in service.
(h) The procedures established under paragraphs (f) and (g) of this section
must—
(1) Be able to be consistently executed in service by crews of average
skill;

Maximum 1500 ft
Takeoff thrust (5 minutes max) continuous
thrust
Gear Gear Flaps Up
down up 160 KIAS
Reference LOH 180 KTS
Accelerate
Brake zero
35 ft
release Gear control handle up
V1VR
First Second Third segment Final segment
Liftoff
segment segment acceleration
Takeoff flight path
Takeoff
field length

Fig. 13.30 An obstacle clearance procedure.

616 Operational Aircraft Performance and Flight Test Practices

A Retract flaps at 400 ft. Gross height

2500

2000
ss
1500 Gro
t
1000 Ne

500

0
Height above reference zero – feet

2500
B Retract flaps at maximum level – off height

2000

1500
Gross
1000
Net
500

0
C Extend second segment to 5 minute limit
2500

2000

1500 Gross

1000 Net

500

Typical procedure for obstacle distance

B C A
0 20 40 60 80 100 120 140
Horizontal distance from reference zero – 1000 ft.

Fig. 13.31 Providing LOH options.

(2) Use methods or devices that are safe and reliable; and
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.

One can elect to provide multiple scenarios guidance (see Fig. 13.31) to
help in selecting the LOH height and plan the takeoff or, should the
engine takeoff power be available long enough (e.g., 10 min) to recommend
the operator climb straight to 1500 ft AGL in the second segment climb con-
figuration (gear up, takeoff flaps, V2 ) before leveling off and accelerating.
 CHAPTER 13 Climb Capability OEI 617

Close-In Charts
To help integrate the first and second segments, including the impact
of the gear retraction time, most manufacturers will include close-in
charts (see Fig. 13.32). These charts account for the relationship between
the climb capability of the first and second segments and the conservative
retraction time of the landing gear. These charts include horizontal
distance covered from reference zero while gaining a specific height
(using the net flight path). Corrections for winds are per the requirements
of §25.105(d)(1): “Not more than 50 percent of nominal wind components
along the takeoff path opposite to the direction of takeoff, and not less than
150 percent of nominal wind components along the takeoff path in the
direction of takeoff.”

Gliding Flight
In Chapter 12, we reviewed basic descent equations and the need for the
airplane to have a thrust deficit. The most extreme condition for descent
flight is a gliding flight (no thrust production; see Fig. 13.33). Under these
conditions, the equation for the force equilibrium along the flight path in a

Chart established with a margin of 35 ft.

Landing gear up
10%

200 200
9%
8%
True obstacle height above the air field (ft)

7%

Net height above the air field (ft).

6%

180
uv r

5%
ne gea
er

160
4%
ma g
in ndin

150
140
Close in obstacle clearance
3%
La

120
t
100 2% men 100
nd seg
%
80 1.6 at seco
t
ra dien
60
lim bg
c
40 Net Reference 0
50

20 Example Horizontal distance : 2500 ft

Head wind : 25 kt
0 Net climb gradient : 4 % 0
Head wind True obstacle height : 77 ft
+50
Wind (kt)

Ref.
0
–10
Tail wind

0 1 2 3 4 5 6 7 8
Horizontal distance (×1.000 ft)

Fig. 13.32 Close-In net obstacle clearance chart.

618 Operational Aircraft Performance and Flight Test Practices

␥
V

Fig. 13.33 Glide performance.

constant airspeed descent [Eq. (12.8)] reduces to

   
1 D 1 1
g ¼ sin ¼  sin (13:5)
W E
Therefore, the larger the aerodynamic efficiency, the smaller the glide
angle. The airspeed for minimum glide angle will be achieved while flying
at minimum drag conditions [Eq. (8.25)]. Table 12.1 in Chapter 12 showed
expectations of the maximum lift-to-drag ratio for various classes of air-
planes. These L/D were for the clean configuration (flaps up and gear up).
Gliders will get much larger values, with competition gliders exceeding
maximum L/D of 40.
For small angles, the descent gradient and the descent angle in radian
are about equal [Eq. (12.14)]. Instead of presenting the information in terms
of glide angle, it is often better to present the information in terms of the
distance covered when gliding from a given altitude (see Fig. 13.34).

Certification Requirements for Single Engine Airplanes

Part 23 small airplane requirements allow for the certification of single-
engine airplanes. For these airplanes, losing the only engine available
means they effectively become gliders, but with a much lower maximum
lift-to-drag ratio. Not only is the airplane starting with a lower L/D, but
the failure of the engine will also introduce additional windmilling drag,
further reducing the maximum value.
The FAA introduced a requirement to define the glide capability of
single-engine airplanes in Amendment 23-50 (1996), §23.71.
§23.71 Glide: Single-engine airplanes
The maximum horizontal distance traveled in still air, in nautical miles, per
1,000 feet of altitude lost in a glide, and the speed necessary to achieve this
must be determined with the engine inoperative, its propeller in the
minimum drag position, and landing gear and wing flaps in the most favor-
able available position.
 CHAPTER 13 Climb Capability OEI 619

Post–Amendment 64, this requirement can be found in §23.2125.

§23.2125 Climb information

(b) The applicant must determine the glide performance for single-engine
airplanes after a complete loss of thrust.

–0.05

–0.1

–0.15
Grad

–0.2

–0.25

–0.3

–0.35
0 5 10 15 20 25 30 35 40 45
Maximum lift to drag

10,000
Em = 5 Em = 12 Em = 20 Em = 45
9000

8000

7000

6000
Altitude (ft)

5000

4000

3000

2000

1000

0
0 10 20 30 40 50 60 70 80
Horizontal distance (n miles)

Fig. 13.34 Maximum L/D ratio impact on glide gradient and distance.
620 Operational Aircraft Performance and Flight Test Practices

a) b)
Conditions Example: * Speed 80 MPH (IAS)
Power OFF Altitude 7000 ft. AGL * Propeller windmilling
Propeller Windmilling Airspeed Best glide MAXIMUM GLIDE
Flaps 0% (UP) * Flaps up
Wind Zero Glide distance 12.5 NM
* Zero wind
Best glide speed
3000 lb 96 KIAS 18,000
2500 lb 87 KIAS

Height above terrain (feet)

Maximum glide ratio ~ 10.9 : 1 16,000
14,000
Height above ground - Feet

14,000 12,000
12,000 10,000
10,000
8000
8000
6000
6000
4000
4000
2000
2000
0
0 0 5 10 15 20 25 30 35
0 2 4 6 8 10 12 14 16 18 20
Ground distance - Nautical miles Ground distance - (Statute miles)

Fig. 13.35 Flight manual glide ratio charts. Source: Cirrus Design SR22 POH, 2003 (left);
Cessna Skywagon 185 POH, 1973 (right).

Glide: Presenting the Information to the Flight Crew

For single-engine airplanes, the emergency procedure (power loss) must
focus on memory item. The manufacturer must provide an easy to under-
stand chart—a single speed, if possible, or an expected speed range, and
the most adverse CG (least glide capability, usually the forward CG condition
for traditional, tail aft airplanes). Figure 13.35 provides examples for single-
speed and weight-based speed glide capability. One can easily see that, for
example, one airplane has a glide capability of about 18 n miles from a
height above ground of 10,000 ft whereas the other has a glide capability of
15 n miles from the same height.
From the chart provided, one can easily estimate the airplane maximum
L/D ratio in the glide configuration; with a little manipulation of the data,
once can also estimate the drag coefficients of the airplane from this infor-
mation. For example, the airplane in Figure 13.35a has a glide ratio of 10.9
to 1, which is the maximum L/D ratio of the plane. The minimum drag air-
speeds are also listed for two different weights. If one knows the wing area of
the airplane and the airspeed error to convert KIAS to KCAS, then one can
compute CL and CD .

Exercise
1. The pilot operating handbook (POH) of a small, four-seat, single-engine
airplane provides the engine-out procedure and glide capability as shown
in Fig. 13.36. Estimate the following parameters:
• Drag force and drag coefficients
• Maximum L/D ratio
Assume an airplane weight of 2100 lb and a wing area of 174 ft2 .
 CHAPTER 13 Climb Capability OEI 621

AMPLIFIED PROCEDURES
Engine Failure
If an engine failure occurs during the takeoff run, the most important
thing to do is stop the airplane on the remaining runway. Those extra items
on the checklist will provide added safety after a failure of this type.

Prompt lowering of the nose to maintain airspeed and establish a glide

attitude is the first response to an engine failure after takeoff. In most
cases, the landing should be planned straight ahead with only small
changes in direction to avoid obstructions. Altitude and airspeed are
seldom sufficient to execute a 180° gliding turn necessary to return to the
runway. The checklist procedures assume that adequate time exists to
secure the fuel and ignition systems prior to touchdown.

After an engine failure in flight, the best glide speed as shown in figure
3-1 should be established as quickly as possible. While gliding toward a
suitable landing area, an effort should be made to identify the cause of the
failure. If time permits, an engine restart should be attempted as shown in
the checklist. If the engine cannot be restarted, a forced landing without
power must be completed.

12,000

10,000
Height above terrain - FT

8000

6000

4000 * Speed 65 KIAS

* Propeller windmilling
2000 * Flaps up
* Zero wind
0
0 2 4 6 8 10 12 14 16 18 20
Ground distance – nautical miles
Figure 3-1. Maximum glide

Fig. 13.36 Engine-out procedure and glide capability.

Reference
[1] Transport Canada, “Compliance with Regulations and Standards for Engine-
Inoperative Obstacle Avoidance,” Advisory Circular No. 700-016, 5 Feb. 2010.
Chapter 14 Turn
Performance
and Maneuver
Margin

Chapter Objective
This chapter will describe the steps needed to develop a basic turn perform-
ance model that can provide a good estimate for most practical applications.
We will review the assumptions used to develop the model and some of the
limits caused by these assumptions. We will also review the concept of
energy maneuverability and discuss certification requirements related to
maneuver margin.

Building a Basic Model

T
here are several ways to change an airplane’s heading; the most
common is to bank the wing into the direction of the turn.
Figure 14.1 shows some of the typical parameters one would
monitor in a turn; these data are extracted a flight test airplane data collection
system. This was a climbing left-hand turn with some speed variation. There
were also airplane configuration changes that are not included in the
data shown.
Some of the highlights from Fig. 14.1 include load factor increasing with
bank angle (about 1.1 g when bank angle is steady at about 25 deg), heading
changes proportional to yaw rate, and yaw rate closely following bank angle.
Some noise in the data is normally expected from the flight test instrumenta-
tion, especially for the load factor where atmospheric turbulence, flight
control movement, and the like may induce small variations (0.05–0.2 g)
that are of similar amplitude to what can be expected for some maneuvers.
To build an initial model, we will use a series of assumptions, including
steady state condition. Most turns/changes of heading involve some form
of airplane banking where the airplane lift force vector is pointed toward
the inside of the turn.

623
 624
Operational Aircraft Performance and Flight Test Practices
7000 15 2
Altitude Bank angle
Airspeed 10 Load factor 1.8
6000
210 5 1.6
190 0 1.4
5000

Airspeed (KCAS)
170

Altitude (ft)

Load factor (g)

–5 1.2

Angle (deg)
4000 150 –10 1
130 –15 0.8
3000
110 –20 0.6
90 –25 0.4
2000
70 –30 0.2
1000 50 –35 0
42,700 42,710 42,720 42,730 42,740 42,750 42,760 42,770 42,700 42,710 42,720 42,730 42,740 42,750 42,760 42,770
Reference time (s) Reference time (s)
20 Heading
30 Bank angle
Pitch attitude
10 Yaw rate 1
Bank angle
20 Aileron position
Angle of attack 0 0
10 Angle of sideslip

Yaw rate (deg/s)

Angles (deg)
–10 –1
Angles (deg)

0
–20 –2

–10 –30 –3

–20 –40 –4

–50 –5
–30
42,700 42,710 42,720 42,730 42,740 42,750 42,760 42,770 42,700 42,710 42,720 42,730 42,740 42,750 42,760 42,770
Reference time (s)
Reference time (s)

Fig. 14.1 Climbing turn parameter variations.

 CHAPTER 14 Turn Performance and Maneuver Margin 625

The basic assumptions we will use to build our first model will be:
• Constant altitude
• Constant airspeed
• Constant bank angle
• Constant load factor
• Coordinated turn (no sideslip)
This initial set of assumptions may sound limiting, but it will prove to
provide a good framework for our initial analysis of turn performance.
This will later be expanded, and we will see the impact of some modifications
to these assumptions.

Reference System and Equilibrium of Forces

With the constraints listed in our basic assumptions, we can now build an
initial model for analyzing turn performance.
To change the heading, a side force must be imposed on the flight path.
This side force, within the constraints of our assumptions, is provided by
tilting the lift vector by an angle f towards the inside of the turn.
The first set of axes defined for our analysis is seen in Fig. 14.2. We have
the local vertical and an axis in the horizontal plane pointing towards the
inside of the turn. The bank angle f is defined with respect to the vertical
axis. The airplane weight W acts along the vertical axis pointing down.
From this, and with the assumption of constant altitude, we can see that
the vertical component of lift developed must be sufficient to counter the air-
plane weight. The equilibrium of forces in the vertical plane can be written as
follows:
L cosðfÞ  W ¼ 0 (14:1)
vertical
Local

L 

r Horizon
plane

Fig. 14.2 Banked airplane in coordinated turn.

626 Operational Aircraft Performance and Flight Test Practices

or,
L cosðfÞ ¼ W (14:2)
The load factor n is defined as the ratio of the lift produced to the airplane
weight. By observation of Eq. (14.2), one gets
L 1
n¼ ¼ (14:3)
W cosðfÞ
One can therefore conclude that, for a coordinated (no sideslip), constant
altitude turn, the load factor is only a function of the bank angle of the air-
plane. One caveat must be added to this last statement: The contribution
of the thrust vector was neglected in this analysis.
A more precise statement is provided in the following Rule of thumb: In a
rule of thumb: coordinated, steady state,
Figure 14.3 provides the variation of load factor vs level flight turn, the load
factor is essentially
airplane bank angle for the model we just developed.
proportional to the inverse of
From this, one can observe that as the bank angle the cosine of the bank angle.
approaches 90 deg, the load factor rapidly increases An angle of bank of 60 deg
and tends towards infinity. A steady state, coordi- results in a load factor of 2.0.
nated, level flight turn at a bank angle of 90 deg is
not possible. This is not to say that an airplane cannot fly with a 90-deg
bank angle; it simply means that the vertical force required to sustain level
flight must come from other parts of the airplane such as the engine thrust
or a fuselage flying at a sideslip angle, as shown in Fig. 14.4.

10
9
8

7
Load factor (g)

5
4
3
2
1
0
0 10 20 30 40 50 60 70 80 90
Bank angle (deg)

Fig. 14.3 Load factor variation with bank angle in a coordinated turn.
 CHAPTER 14 Turn Performance and Maneuver Margin 627

Fig. 14.4 Airshow plane flying with 90-deg bank angle.

Part 25 transport category airplanes are designed to a maximum (limit)

load factor anywhere from 2.5 g to 3.8 g. The design requirements are
defined by FAA 14 CFR 25.337(b):
(b) The positive limit maneuvering load factor n for any speed up to Vn
may not be less than 2.1 þ 24,000/(W þ 10,000) except that n may not
be less than 2.5 and need not be greater than 3.8 – where W is the
design maximum takeoff weight.

The resulting design limit load factor would then be as shown in Fig. 14.5.
Some military planes, fighters in particular, are designed to support
higher loads; they are typically capable of achieving between 7 g and 9 g.
These higher values, however, may be too much for the flight crew. As a
side note, approximately 83.6 deg of bank is required to achieve 9 g under
the assumptions used for our model.
Let’s now consider the forces acting along the flight path. To maintain
steady speed per our model assumptions while flying at a constant altitude,
the airplane must generate sufficient thrust to counter drag. As we saw in
Chapter 8, the airplane is flying at an angle of attack, typically small,
and the engines are at an incidence with respect to that angle resulting
in the thrust line being at an angle aT with respect to the airspeed
vector (see Fig. 14.6). The thrust component along the flight path is then
T cos(aT ).
For most planes, in “normal” flight conditions, the angle of attack is
typically small (3–8 deg); therefore, the value of the cosine of this small
angle is anywhere between 0.99 and 1.0. For all practical purposes, we can
assume that

T ¼D (14:4)
628 Operational Aircraft Performance and Flight Test Practices

4.5

4
Design load factor (g) 3.5

2.5

1.5

0.5

0
0

00
,00

,00

,00

,00

,00

,00

,00

,00

,00

0,0
10

20

30

40

50

60

70

80

90

10
Design takeoff weight (lb)

Fig. 14.5 Design limit load factor for Part 25 transport category airplane.

From Fig. 14.6, it can also be noticed that part of the thrust vector will be
pointed towards the inside of the turn. That value is equal to [T sin(aT )
sin(f)]. Considering that we are dealing with a small angle of attack and
bank angles typically smaller than 45 deg, the combination results in less
than 10% of the thrust force being directed towards the inside of the turn.
For most transport category airplanes, the maximum expected thrust force
is less than 25% of the airplane weight (i.e., T/W , 0.25) under normal
flight conditions, leading to a force of less than 2.5% of the airplane weight
pointing to the inside of the turn. So again, for all practical purposes, we

Horizontal
T plane
D
T 
V

Fig. 14.6 Impact of angle of attack on the forces in a turn.

 CHAPTER 14 Turn Performance and Maneuver Margin 629

can neglect the impact of the thrust angle Curved

of attack on the forces affecting turn path
performance. D
With these assumptions, we can now
create a turn performance model to
determine the airplane rate of turn and
radius of turn. Looking at the turn in
the horizontal plane (see Fig. 14.7),
note that the thrust and drag forces act
tangent to the curved flight path; the air- L sin ()
speed is also tangent to the curved flight
path, and the force generating the curve
path is proportional to the lift force
times sin(f). The assumption of con-
stant bank angle will make this force T
constant, which will lead to a constant
arc radius. The equilibrium of forces
along the radius is then
W
L sinðfÞ ¼ V ẋ (14:5) Fig. 14.7 Turning flight.
g
where ẋ is the airplane angular velocity (rad/s). From this, the angular
velocity (or rate of turn) can be written as
L sinðfÞ
ẋ ¼ g (14:6)
W V
Combining Eqs. (14.3) and (14.6), we get
ng sinðfÞ g tanðfÞ
ẋ ¼ ¼ (14:7)
V V
We can make the rate of turn a function of load factor and speed only,
removing the dependency on using the bank angle in the equation by
simple trigonometric analysis.
pffiffiffiffiffiffiffiffiffiffiffiffiffi
g n2  1
ẋ ¼ (14:8)
V

The tangential velocity (the true airspeed) is equal to the radius of the
turn multiplied by the angular velocity. From this, and using Eq. (14.8), we
get an equation for the radius of the turn as a function of true airspeed
and load factor.

V V2
r¼ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi (14:9)
ẋ g n2  1
630 Operational Aircraft Performance and Flight Test Practices

35

0
100
9.0
30
7.0 Ra

00
)
di

r (g

15
25 us
Rate-of-turn (deg/s)

of

to
5.0

00
tu

fac
rn

20
(f t
d
)
20 Loa 00
0
3.0 3

15 00
2.0 50
00
75 00
10 10,0
1.1
1.05
5

0
0 200 400 600 800 1000 1200 1400 1600 1800
True airspeed (ft/s)

Fig. 14.8 Generic grid for turning performance.

The following are preliminary observations for Eqs. (14.8) and (14.9):

• For a given airspeed, the rate of turn is proportional to the load factor,
and the radius of the turn is proportional to the inverse of the
load factor.
• For a given load factor, the rate of turn is proportional to the inverse of the
airspeed, and the radius of the turn is proportional to the airspeed squared.
• These two equations are not a function of aircraft characteristics. Any
airplane that can achieve a given combination of airspeed and load factor
will have the same rate of turn and radius of turn while flying under the
conditions described in the assumptions.

With this last observation, we can create a generic grid for turn perform-
ance that includes the rate of turn, radius of the turn, load factor, and air-
speed (see Fig. 14.8).

Example 14.1
A business jet is in a holding pattern at 200 KCAS and 3000 ft pressure alti-
tude on a standard day. The pilot initiates a 30-deg bank turn. How long
does it take to complete a 360-deg turn?

(Continued)
 CHAPTER 14 Turn Performance and Maneuver Margin 631

Example 14.1 (Continued)

Solution: To compute the time required to complete the 360-deg turn, one
must determine the rate of turn of the airplane. For this, based on Eq. (14.18),
both the true airspeed and load factor are required. The airspeed provided is
knots calibrated airspeed (KCAS), which must first be converted to knots true
airspeed (KTAS). Using the material seen in Chapter 3, one can convert KCAS
to KTAS for the atmospheric conditions specified
200 KCAS ¼ 208:7 KTAS
and
208:7 KTAS ¼ 352:25 ft=s

Next we need the load factor. Knowing the bank angle, and using Eq. (14.3),
we get
1
n¼ ¼ 1:1547
cosð30 degÞ
Finally, the rate of turn is
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi 32:17 ft ð1:1547Þ2 1
2
g n 1
ẋ ¼ ¼ s2 ¼ 0:052728 rad=s ¼ 3 deg=s
V 352:25 ft=s
Therefore, it will take 120 s (2 min) to complete a 360-deg turn under
these conditions (see Fig. 14.9).

18
0
100

16 2.5

14 2.0
Ra
(g)

00

di
15

us
tor
Rate-of-turn (deg/s)

12 of
00
fac

1.5 tu
20

rn
d

00
(f t
Loa

10 1.3 )
30
1.15
8 00
1.1 50
1.05 00
6 75
00
10,0
4

0
0 100 200 300 400 500 600 700 800 900 1000
True airspeed (ft/s)

Fig. 14.9 Results of Example 14.1.

632 Operational Aircraft Performance and Flight Test Practices

Impact of Load Factor on Stall Speed

An increase in the load factor is, by definition, an increase in the lift force
produced by the plane for a given weight. The lift can be written as a function
of the airplane weight and the load factor as follows:

L ¼ nW (14:10)

In terms of lift coefficient, we can write Eq. (14.10) as follows:

nW
CL ¼ (14:11)
qS

As was seen in Chapters 6 and 7, the airplane can produce only so much
lift before it stalls. We saw that the maximum usable lift coefficient (CLmax )
was dependent on several factors, including maximum allowable for certifi-
cation purpose. For this defined maximum lift coefficient, the stall speed is
calculated using
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u nW pffiffiffiu W
Vstalln u
¼t ¼ nu
1 t1
r s S CLmax r s S CLmax
2 SL 2 SL
pffiffiffi
¼ nVstall1:0g (14:12)

So it may be possible to stall an airplane that was

flying in level flight away from the 1.0 g stall speed by Note: Under Part 25
increasing the load factor. transport category
certification, Vstall is referred
While flying at a given airspeed V, one can esti- to as VSR under a 1.0-g,
mate the load factor increase required to reach stall most-adverse-CG,
conditions. Another way of saying this is, how zero-thrust condition, as
much margin is there in terms of load factor before seen in Chapters 6 and 7.
the airplane would stall?
 
V pffiffiffi V 2
¼ n ! n¼ (14:13)
Vstall Vstall

Example 14.2
Part 25 certification requirements state under paragraph 25.125(b)(2) that a
stabilized approach must have a reference airspeed (VREF) of not less than
1.23 VSR. When flying at that speed, how much load factor margin does
the airplane have before reaching stalling conditions?

(Continued)
 CHAPTER 14 Turn Performance and Maneuver Margin 633

Example 14.2 (Continued)

Solution: The minimum ratio of approach speed to reference stall speed is
VREF pffiffiffi
¼ 1:23 ¼ n
VSR
n ¼ ð1:23Þ2 ¼ 1:5129

So the airplane, while flying at VREF , has a 1.5129-g margin to stalling con-
ditions. If we compare this to Eq. (14.3), we can see that the airplane can
execute a maximum bank angle of
 
1
f ¼ cos1 ¼ 48:6 deg
n
before getting into stalling conditions.

Example 14.3
Consider an airplane with a wing loading (W/S) of 75 lb/ft2 and a maximum
lift coefficient of 1.5. This airplane has a maximum structural load factor capa-
bility of 2.5 g.
1. Chart the change in stall speed as a function of load factor from level flight
(1 g) to the maximum structural limit for both sea level and 15,000 ft
pressure altitude conditions.
2. If the 1-g stall condition corresponds to the VSR condition of Example 14.2,
what would be the rate of turn at stall and at sea level while flying at the
VREF of Example 14.2?
3. What is the impact of increasing altitude on the stall speed at all load
factors?

Solution:
1. See Fig. 14.10.
2. VSR at sea level for this example is 205.1 ft/s true airspeed. This gives a
VREF (1.23VSR ) of 252.3 ft/s, as shown in Fig. 14.10. The load factor at stall,
while flying at VREF , is 1.5129 per Example 14.2. The rate of turn of this
airplane while flying at VREF and under a load factor for stall conditions
would be
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi 32:17 ft ð1:5129Þ2 1
g n2  1 s 2
ẋ ¼ ¼ ¼ 0:1448 rad=s ¼ 8:3 deg=s
V 252:3 ft=s

(Continued)
634 Operational Aircraft Performance and Flight Test Practices

Example 14.3 (Continued)

18

0
16 2.5

100
14 2.0
Ra

00
g)

di

15
(

us
Rate-of-turn (deg/s)

tor

12 of

00
1.5
fac

tu

20
rn
d

(f t
00
Loa

10 1.3 )
30
l
leve

1.15
8 00
Sea

t
50
0f

1.1
,00

1.05 00
6 75
15

00
10,0
4

0
0 100 200 VREF300 400 500 600 700 800 900 1000
True airspeed (ft/s)

Fig. 14.10 Rate of turn chart for Example 14.3.

3. Increasing the altitude will increase the stall speed, in terms of true
airspeed, at all load factors.

A search of the Internet has uncovered an interesting chart from the Min-
istry of Aircraft Production, England dating back to 1940 (see Fig. 14.11). The
bottom scale is true airspeed (and equivalent airspeed) for the chart altitude
of 12,000 ft pressure altitude; the right scale is time to complete a 360-deg
turn (equivalent to 360 divided by rate of turn). Other information added
to the chart includes:

• CLmax variation with load factor; this is specific to this one airplane.
• Airplane weight and engine maximum brake horsepower.
• Wings level climb capability of the airplane (angle of straight climb). This
line represents the airplane available excess power.
• The left scale is the difference in climb angle between a wings level climb
and a climb with a given load factor.

This chart is a good introduction to the next section. Does the airplane
have enough thrust to sustain the turning condition?
 CHAPTER 14 Turn Performance and Maneuver Margin 635

Fig. 14.11 Spitfire turn performance capability, 1940 report.

Impact of Load Factor on Thrust Required

To maintain steady airspeed in level flight per our assumptions, the
airplane needs to be able to generate enough thrust to counter the drag.
An increase in load factor is an increase in the generation of lift to maintain
altitude; therefore, there is an increase in the lift-induced drag. Expanding
Eq. (14.4),
0 1
2
B1 K ðnW Þ C
T  @ rSL s V 2 S CD0 þ A¼0 (14:14)
2 1 2
r sV S
2 SL

One can observe that the induced drag component is proportional to the
square of the load factor. This means that an airplane performing a 2-g turn
will encounter 4 times more induced drag than flying in level flight (1 g) at
the same speed and altitude. An airplane flying under 9-g conditions will
have 81 times more induced drag. Induced drag can rapidly become the
dominant drag component of a maneuvering airplane.
636 Operational Aircraft Performance and Flight Test Practices

To better understand the impact of the load factor on the airplane

drag, consider this first example for an airplane flying at 15,000 ft pressure
altitude. The airplane has a weight of 75,000 lb and a wing area of 750 ft2 .
The maximum lift coefficient for the plane in the configuration for this
example is 1.5, and the airplane has a limit load factor (structural limit) of
2.5 g. The drag polar, neglecting compressibility effects for simplicity,
has a minimum drag coefficient CDo of 0.022, and the wing has an
aspect ratio (AR) of 7.5 and an aerodynamic lift efficiency e of 0.8
[K ¼ 1/(p AR e) ¼ 0.05305]. Finally, consider the thrust available to be
equal to 14,000 lb, independent of airspeed. Such an airplane is representa-
tive of existing regional jets.
Figure 14.12 represents the results of the analysis for load factors varying
from 1.0 g to 2.5 g.
We can note the following from Fig. 14.12:
• The airplane stall speed increases with increasing load factor.
• The drag at any speed increases with increasing load factor, its effect being
more pronounced at low speed where induced drag is dominant.
• The minimum drag value increases with increasing load factor, and the
airspeed at which the minimum drag occurs increases as well.

Overall, the behavior is similar to an actual increase in airplane weight, as

seen in previous chapters.
We can also plot the same example on a rate-of-turn chart, as in Fig. 14.13.

20,000
18,000
16,000 2.5 g
Available thrust
14,000
Drag, thrust (lb)

2.0 g
12,000
n
talll,

10,000
Vs

8000 1.3 g
6000
1.0 g
4000
2000
0
0 200 400 600 800 1000 1200
True airspeed (ft/s)

Fig. 14.12 Thrust variation with airspeed and load factor.

 CHAPTER 14 Turn Performance and Maneuver Margin 637

12
0
2.5 200

10

(g)
2.0
00
30

tor
Ra

ac
diu

df
so
f tu
a
Rate of turn (deg/s)

8 Lo 1.5 1
0 rn
500 (f t)
1.3
0
6 2 750

tall,n
1.15
00
1.1 10,0
Vs 12,5
00
00
1.05 15,0
4

T=D
2

0
100 200 300 400 500 600 700 800 900 1000 1100
True airspeed (ft/s)

Fig. 14.13 Rate-of-turn chart.

From this chart, we can observe the stall speed line (Vstall,n), below
which speed the stall conditions are encountered. We also have the line
where maximum available thrust equals the drag (T ¼ D); the airplane has
sufficient thrust available below that line to maintain the turn condition.
Two regions of interest stick out of this chart:
• Zone 1: This zone is bounded on the left by the airplane stall speed, on
the right by the airplane limit structural load factor (2.5 g), and on the
bottom by the available thrust (T ¼ D). In this region, it is possible for an
airplane to reach a given rate of turn without stalling or exceeding
limit loads, but there is not enough thrust to sustain the airspeed and/or
altitude. This is the region of instantaneous turn capability where, in
order to maintain airspeed, the airplane will need to lose some altitude,
or if one wants to maintain altitude, airspeed will decrease. The load
factor required to achieve this T ¼ D condition can be derived from
Eq. (14.14)
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 ffi
q 1 ðT =SÞ
n¼  CD0 (14:15)
ðW =S Þ K q

• Zone 2: This zone is bounded to the left by the limit load factor of the
airplane and to the right by the available thrust. The airplane should
not venture into this region for risk of damage to the structure.
638 Operational Aircraft Performance and Flight Test Practices

There is also a section at high speed, not highlighted in Fig. 14.13, where
the T – D ¼ 0 curve goes back below the maximum structural limit.
Not shown in this example are:
• Compressibility effects that could increase the drag at high speed,
thereby decreasing the high-speed capability more than shown here
• The thrust lapse rate that was set to zero but most of the time appears
as a reduction of thrust available with speed
• The maximum operating speed of the plane (VMO ) that is typically
around 320 KCAS (395.2 KTAS or 667 ft/s for this example under
standard atmospheric conditions) for this type of airplane

Special Cases for Turn Performance

From observation of Fig. 14.13, four speeds can be identified:
• Maneuvering speed
• Fastest turn speed
• Stall turn speed
• Minimum radius of turn speed

Maneuvering Speed
The condition where the stall speed curve meets the maximum design
load factor curve is called the maneuver speed or corner speed condition. If
the airplane flies at a speed less than the corner speed, it will stall before
reaching limit load factor. Above corner speed, the airplane will reach limit
load factor before stalling.
The speed for the corner condition can be derived by using the load factor
driven stall speed from Eq. (14.12) and setting the load factor equal to the
limit load factor nlimit
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
un
limit ðW =S Þ
Vcorner ¼ u
t1 (14:16)
rSL s CLmax
2

Under these conditions, the rate of turn and radius of turn are derived
from Eqs. (14.8) and (14.9), respectively. Writing these equations specifically
for the corner speed turn condition yields
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
g n2limit  1
ẋcorner ¼ (14:17)
Vcorner
2
Vcorner
rcorner ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (14:18)
g n2limit  1
 CHAPTER 14 Turn Performance and Maneuver Margin 639

The FAA/EASA/TCCA define the design maneuvering speed VA in

§25.335(c):
25.335(c) Design maneuvering speed VA. For VA, the following apply:
p
(1) VA may not be less than VS1 n where—
(i) n is the limit positive maneuvering load factor at VC; and
(ii) VS1 is the stalling speed with flaps retracted.
(2) VA and VS must be evaluated at the design weight and altitude
under consideration.
(3) VA need not be more than VC or the speed at which the positive CN max
curve intersects the positive maneuver load factor line, whichever is less.

So VA , as defined in (c)(1), is the same as the corner speed per Eq. (14.18) with
additional limitations as stated under §25.335(c)(2) and (3).
Flying at the corner speed allows the airplane to reach the maximum rate
of turn possible without stalling or exceeding structural limits; however, most
of the time this is not sustainable (i.e., the airplane will not have enough
thrust to counter the drag generated in the turn at the combination of
nlimit and CLmax ).

Example 14.4
For the airplane as defined in Fig. 14.12, what would be the impact on the air-
speed if the airplane were to try to fly at the corner conditions while maintain-
ing constant altitude? How much thrust deficit does it have?

Solution: For this airplane, nlimit ¼ 2.5, CLmax ¼ 1.5, and the wing loading
(W/S) ¼ 100 lb/ft2 . Pressure altitude is 15,000 ft under standard conditions,
leading to an air density ratio s ¼ 0.6292. The resulting corner speed is then
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi v
u
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 ffi
un ð W =S Þ u 2:5 g 100 lb=ft 2
ft
Vcorner ¼ ut1
limit
¼u
t1 ¼ 472:1 ¼ 279:7 KTAS
s
r s CLmax r ð0:6292Þ 1:5
2 SL 2 SL

Under these conditions, the airplane drag coefficient is

2
CD ¼ CD0 þ KCLmax ¼ 0:022 þ 0:05305 ð1:5Þ2 ¼ 0:1414
and the drag is
1 2
D¼ r s Vcorner S CD ¼ 17,670 lb
2 SL
This is a thrust deficit of 3670 lb with respect to the example maximum
available thrust of 14,000 lb. Although the airplane could generate a turn
rate of up to almost 9 deg/s at the corner condition, the thrust deficit
under constant-altitude turn would result in a deceleration of the airplane

(Continued)
640 Operational Aircraft Performance and Flight Test Practices

Example 14.4 (Continued)

of the order of
W dV
T D¼
g dt
   
dV T D ft 3670 lb ft
¼g ¼ 32:17 2 ¼ 0:04894 g ¼ 1:574 2
dt W s 75; 000 lb s
¼ 0:93 KTAS=s

So the airplane would lose about 1 kt true airspeed per second, and
because it is flying at stall condition, it would enter the stall regime.

Fastest Turn
The fastest turn (subscript FT) is defined as the condition that results in
the highest rate of turn under sustained turning conditions (i.e., there is
enough thrust available to counter the drag produced). Mathematically,
this can be determined by differentiating the rate of turn [Eq. (14.8)] and
setting the result equal to zero.
   
@ ẋ 2 @n
¼ 0 ! n  1  2nV ¼0 (14:19)
@V FT @V
The load factor required to maintain steady level flight is defined per
Eq. (14.15), where the thrust is typically a function of speed (and altitude)
and the aerodynamic coefficients (CDo , K) may be impacted by the Mach
number. Equation (14.19) can be solved to find the conditions for fastest
turn (see [1], Appendix D), assuming zero thrust lapse rate with speed
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 
Tmax
nFT ¼ 2 Em  1 (14:20)
W

and the airspeed is

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffi
2 ðW =SÞ 4 K
VFT ¼ (14:21)
rSL s CD0

Let’s define the maximum aero load factor as

nmaxaero ¼ Tmax=W Em (14:22)

By observation, the fastest turn speed for zero thrust lapse rate is equal to
the level flight minimum drag airspeed. This is expected because it is the
 CHAPTER 14 Turn Performance and Maneuver Margin 641

flight condition where the excess thrust is maximum. In practice, that con-
dition would be slightly away from the minimum drag condition, typically
at a slower speed based on thrust often decreasing with increasing airspeed.
With the load factor and airspeed known, one can compute the resulting rate
of turn for the fastest turn condition.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
g n2FT  1
ẋ ¼ (14:23)
VFT

In order to maximize the sustained rate of turn, an airplane should

be designed to have a large thrust-to-weight (T/W ), a low wing loading
(W/S), a small induced drag coefficient (K, which is equivalent to a large
aspect ratio and Oswald coefficient near 1.0), and a low parasitic drag coeffi-
cient (CDo ). The influence of these major parameters on the turn perform-
ance curve is shown in Fig. 14.14.
To show some impact of these critical parameters on turn performance,
we will use an example with three different airplanes designed to perform a
similar mission but of different generations. For this example, we will con-
sider the North American F-86F, the McDonnell F-4, and the General
Dynamics F-16 (see Table 14.1 and Fig. 14.15). This represents an evolution
of front-line fighters in the U.S. Air Force from the late 1940s to the
mid-1970s (before stealth became a driving factor). The data were retrieved
from the Internet and are not vetted; they are simply used to show/discuss
the impact of the critical parameters on turn performance. Note that all air-
planes had about the same wing span.
For each plane, we included the typical three limit curves (i.e., Vstall;n ,
T ¼ D, and max structural load factor), and all airplanes are flying at approxi-
mately 5000 ft pressure altitude. Note that for the F-4, the max load factor
value changes from 7.5 g at low speed to near 5 g at higher speed. We can
identify the sustained envelope as limited by these three curves as well as
the instantaneous turn performance above the T ¼ D curve. Note that for
the F-4 and F-16, the airplanes have enough thrust to reach and maintain
max structural load factor over a certain speed regime. Comparing the
three airplanes:
• Stall speed curve: Both the F-86 and the F-4 have a similar limitation in
maximum lift; the higher wing loading of the F-4 pushed the stall speed
higher, meaning the F-86 could operate at speeds below those of the F-4
and outmaneuver the F-4. The F-16 had a much higher angle of attack
range and higher CLmax , so even if the wing loading is higher than for the
F-86, the higher angle of attack (AoA) and larger T/W are used favorably
to get the stall speed slower, and the F-16 can out-turn the other two
planes (the F-86 has a small advantage above 3.5 g near stall).
• Maximum load factor curve: There is a small increase from the F-86 to the
F-4. The main benefit is above Mach 0.70 for the F-4. The F-16 has the
642 Operational Aircraft Performance and Flight Test Practices

12
0
2.5 200

10 2.0 Ra

)
00

(g
diu
30 so

tor
Th f tu
ru

fac
st i rn
nc (f t)
8 ad 1.5 rea
Rate of turn (deg/s)

se
Lo
5000
1.3
0
6 1.15 750

n,n
00
stali
1.1 10,0
00
V
1.05 12,5 00
15,0
4

T=D
2

0
100 200 300 400 500 600 700 800 900 1000 1100
True airspeed (ft/s)
12
0
2.5 200

10 2.0 Ra
(g)

00 diu
30 so
tor

f tu
ac

rn
df

(f t)
8
Rate of turn (deg/s)

1.5
a
Lo

0
500
1.3
0
6 1.15 750
,n

00
stall

1.1 K 10,0
V

00
1.05 increase 12,5 00
4 15,0

T=D
2
CDo
increase
0
100 200 300 400 500 600 700 800 900 1000 1100
True airspeed (ft/s)

Fig. 14.14 Influence of thrust and drag coefficients on T ¼ D curve.

highest maximum load factor at 9 g and can again outmaneuver both

planes. The higher load factor did, however, introduce other issues related
to the physical limits of the pilot.
• The T ¼ D curve: Even if the F-4 had a larger T/W than the F-86, its lower
aspect ratio wing led to higher induced drag at the lower speed and a lower
T ¼ D curve (lower sustained rate of turn) below Mach 0.70. The F-4 had a
 CHAPTER 14 Turn Performance and Maneuver Margin 643

Table 14.1 Main Characteristics of Three U.S. Air Force Fighters

F-86F F-4 F-16A

Wing area S 313 ft2 530 ft2 300 ft2
Wing span b 37 ft 38 ft 32.7 ft
Aspect ratio AR 4.4 2.7 3.5
Weight 15,000 lb 41,550 lb 20,875 lb
Wing loading W/S 47.9 lb/ft2 78 lb/ft2 70 lb/ft2
Static thrust sea level 5910 lb 35,640 lb 23,744 lb
T/W static sea level 0.394 0.86 1.14
Max structural load factor 7g 7.5 g (low speed) 9g
Year of first flight 1947 (XP-86) 1958 (XF-4H-1) 1974 (YF-16)

marked advantage above that Mach number and at speeds above Mach
0.90 where the F-86 could not reach in level flight. The F-16, with its T/W
higher than unity, exceeds the sustained rate-of-turn performance of the
other two airplanes.

Stall Turn Speed

Equation (14.12) showed the impact of the load factor on the stall speed.
The condition under which an airplane achieves both stall and maximum

35 F-4 F-16A
F-86F
9g

30
7g
tor

25
Rate of turn (deg/s)

5g
fac
d

4g
Loa

20
3g

15
2g
g
1.5

10

0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Mach

Fig. 14.15 Turning performance of three U.S. Air Force fighters, 5000 ft pressure altitude.
644 Operational Aircraft Performance and Flight Test Practices

available thrust condition in a steady state is called the stall turn condition
(subscript ST). From Fig. 14.13, that speed is at the intersection of the stall
speed curve and the thrust equals drag curve. Under these conditions, the
load factor is
   
T T CLmax
nST ¼ EST ¼ (14:24)
W max W max CD0 þ K CL2max

Combining this with Eq. (14.12) leads to a stall turn speed of

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u
u 2 Tmax=S
VST ¼ t (14:25)
rSL s CD0 þ K CL2max

This is typically the condition of minimum radius of turn under sustained

flight conditions that meets our basic assumptions; however, it is strongly
dependent on the drag variation with airspeed as the stall is approached,
on the variation of thrust vs airspeed, and on how the load factor affects
CLmax .

Minimum Radius of Turn

Observing the T ¼ D curve on Fig. 14.13, one can see that, theoretically, it
is possible to have a smaller radius of turn by flying even slower than the stall
speed (also shown on Fig. 14.10 with the wings level climb capability curve).
This is not a practical flight condition for most planes, however. Figure 14.13
can be redone in terms of radius of turn vs airspeed, as shown in Fig. 14.16.
Note that the radius of turn keeps decreasing while following the T ¼ D
curve to slower than the stall speed (based on the models used).
Mathematically, the minimum radius of turn (subscript TT for tightest
turn) that an airplane can sustain per our turning model assumptions can
be found by differentiating Eq. (14.9) with respect to airspeed and setting
the results to zero.
   
@r @n
¼ r2  1  q n ¼0 (14:26)
@V TT @q TT

Solving this equation (see [1], Appendix D), yields a speed for
tightest turn
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðW =SÞ K
VTT ¼2 (14:27)
rSL s ðT =W Þ
 CHAPTER 14 Turn Performance and Maneuver Margin 645

Load factor
16,000 n = 1.05 n = 1.1 n = 1.15 n = 1.5 n = 2.0

14,000

12,000

s)
eg/
Radius of turn (ft)

10,000

rn (d
2
of tu
8000
T=D
Rate

6000 Vstall

4000 3
4
2000
6
8
10
0
0 100 200 300 400 500 600 700 800 900 1000
True airspeed (ft/s)

Fig. 14.16 Radius of turn chart derived from Fig. 14.13 data.

a load factor for tightest turn of

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 1 1
nTT ¼u
t2  h i2 ¼ 2  (14:28)
Tmax=W Em ½nmaxaero 2

and finally a radius of turn for the tightest turn conditions

VTT 2
rTT ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi (14:29)
g n2TT  1

Example 14.5
An airplane flying at 15,000 ft pressure altitude in a standard atmosphere has
the following characteristics:
CDo ¼ 0.025, AR ¼ 5, Oswald coefficient (e) ¼ 0.8, W ¼ 15,000 lb, S ¼ 300 ft2 ,
CLmax ¼ 1.5, T/W ¼ 0.22
1. Assuming the T/W and the aerodynamic coefficients are not a function of
airspeed, what would be the minimum sustainable radius of turn and what
would be the highest sustainable rate of turn?

(Continued)
646 Operational Aircraft Performance and Flight Test Practices

Example 14.5 (Continued)

2. What would happen to the minimum radius and highest rate of turn
condition if the wing aspect ratio was increased to 8, and no other changes
were made to the baseline coefficients?
3. For the aspect ratio of 5, what would happen if the wing area was reduced
to 200 ft2 ?
Solution: Based on the coefficients provided, the induced drag coeffi-
cient K is
1
K¼ ¼ 0:0796
p AR e

The maximum lift-to-drag ratio Em would be 11.21.

1. With this, the resulting turn conditions of interest are shown in Table 14.2.
The condition for minimum sustainable radius of turn for this example is
the stall turn (ST) condition at 1762 ft, because the tightest turn condition
(TT) would require more lift (CL ) than is available. The fastest sustainable
rate of turn would be the FT condition at 9.1 deg/s with a lift coefficient less
than maximum.
2. If the aspect ratio was increased to 8, it would reduce the induced drag
coefficient K to
1
K¼ ¼ 0:0497
p AR e
and increase the maximum lift-to-drag ratio to 14.18. The impact on the
stall conditions would be as shown in Table 14.3.
The condition for minimum sustainable radius of turn for this example
is still the ST condition at 1522 ft, because the TT condition would require
more lift (CL ) than available. The fastest sustainable rate of turn would now
be the ST condition as well at 12.3 deg/s, because the FT condition would
also require more lift than available. Reducing the induced drag coefficient
(through aspect ratio increase) has pushed the peak conditions for

Table 14.2 Conditions of Interest for Example 14.5

FT ST TT
Load factor (g) 1.98 1.62 1.35
True airspeed (ft/s) 345.4 268.6 219.8
Rate of turn (deg/s) 9.1 8.7 7.7
Radius of turn (ft) 2165 1762 1643
Required CL 1.11 1.5 1.88

(Continued)
 CHAPTER 14 Turn Performance and Maneuver Margin 647

Example 14.5 (Continued)

Table 14.3 Impact on Stall Conditions

FT ST TT
Load factor (g) 2.29 2.41 1.38
True airspeed (ft/s) 307.1 328 173.8
Rate of turn (deg/s) 12.4 12.3 10
Radius of turn (ft) 1424 1522 991
Required CL 1.62 1.5 3.05

minimum radius and FT to an airspeed below the maximum lift capability

of the wing; therefore, the ST condition now becomes the best turn
performance condition for radius of turn and rate of turn.
3. The third condition is to reduce the wing area (increase the wing loading)
with the other coefficients remaining unchanged. The new turn conditions
are as shown in Table 14.4.
The higher wing loading has essentially pushed all airspeeds to higher
values. The radius of turn for all conditions is higher than for condition 1,
and the rate of turn for all is lower. The minimum radius of turn condition
is still the ST condition, now with a radius of 2643 ft rather than 1762 ft for
condition 1. The rate-of-turn condition is the FT condition, but the rate is
reduced to 7.5 deg/s instead of the previous 9.1 deg/s.

Table 14.4 New Turn Conditions

FT ST TT
Load factor (g) 1.98 1.62 1.35
True airspeed (ft/s) 423 329 269.2
Rate of turn (deg/s) 7.5 7.1 6.3
Radius of turn (ft) 3248 2643 2464
Required CL 1.11 1.5 1.88

Altitude Effects on Sustainable Turn Performance

Flight Envelope
Changing altitude will impact the generation of thrust, the airplane drag,
and the stall speed. In a previous section we saw that the stall speed will
increase with altitude for a given weight and load factor; this was shown in
Example 14.3. We saw in Chapter 8 that an increase in altitude has a tendency
to push minimum drag to a higher true airspeed, but the minimum drag value
648 Operational Aircraft Performance and Flight Test Practices

does not change for a given airplane weight. From a thrust generation point
of view, Fig. 14.12 shows that increasing the load factor affects both the high-
and low-speed capability of the airplane. We also know that there is generally
a decrease in thrust with increasing altitude. Equation (14.14) can be rewrit-
ten as follows:

S2 CD0 q2  T S q þ Kn2 W 2 ¼ 0 (14:30)

Solving for the dynamic pressure,

" sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#
ðT =S Þ 4K CD0 n2
q¼ 1+ 1 (14:31)
2CD0 ðT =W Þ2

Leading to a true airspeed equation

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
u sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi3
u
u ðT=SÞ 4 n2 5
V ¼t 1+ 1 (14:32)
rSL s CD0 ½Em ðT=W Þ

Equation (14.32) has two solutions, as was seen in Chapter 8; the plus sign
(þ) represents the thrust-limited Vmax , and the minus sign (–) represents the
thrust-limited Vmin . The altitude effects can be clearly seen in the air density
ratio. Not as clear is the impact on the thrust (absorbed through the use of T’
only in the equation). As well, the thrust lapse rate (change with speed) is also
not clearly seen, but it is accounted for by the use of T’ again.

Example 14.6
Consider a small general aviation single-engine prop airplane that has the fol-
lowing characteristics: available power (shaft horsepower times propeller effi-
ciency) of 160 HP at sea level, weight of 2000 lb, wing area of 174 ft2 , parasitic
drag coefficient of 0.033, induced drag coefficient K of 0.0530, and a maximum
lift coefficient of 1.5. Assuming the available power decreases proportionally
to the air density ratio and is independent of airspeed, build the level flight
envelope (Vstall and Vmax ) under a 1.0-g condition. Repeat under a 2.0-g
(60-deg bank) condition. Additional information is that the airplane is
limited to less than 19,000 ft pressure altitude by certification and has a
maximum normal operating speed (VNO ) of 126 KCAS.
Solution: This problem can be best summarized by Fig. 14.17. For this
example, we assumed that the power available varies proportional to the air
density ratio and is constant with airspeed; therefore, the available thrust is
equal to P/V.

(Continued)
 CHAPTER 14 Turn Performance and Maneuver Margin 649

Example 14.6 (Continued)

20,000
Max altitude
18,000

Vmax,1 g
16,000
14,000 VNO
Pressure altitude (ft)

V max
l,1 g
12,000

Vstal

,2 g
l,2 g
10,000

Vstal
8000

6000

4000

2000

0
0 50 100 150 200 250 300
True airspeed (ft/s)

Fig. 14.17 Impact of the load factor on the sustainable flight envelope.

The figure shows that at the lower altitudes, the speed envelope where the air-
plane can maintain both airspeed and altitude gets narrower at the larger load
factor; in this case, however, the airplane still has enough thrust to maintain
both. As the altitude increases, the envelope becomes narrower until an alti-
tude of about 14,700 ft, where both the stall speed curve and the Vmax curve
meet. While flying above that altitude, should the pilot execute a 2-g maneu-
ver, there will be no flight conditions where the altitude and airspeed can be
maintained. This leads us to our next subject: Is there a better way to show the
airplane capability including excess and deficit in thrust?

Energy Maneuverability
The equations developed in the previous sections were based on a set of
assumptions (constant speed, constant altitude, constant bank angle) used
to develop a relative measure of sustained (fastest turn, stall turn) or instan-
taneous (corner speed) maneuverability. These equations can be used to get
a good approximation of an airplane’s capability for a specific set of con-
ditions, but when it comes time to compare one airplane against another
to see whether one has a turn advantage over the other, these equations
650 Operational Aircraft Performance and Flight Test Practices

may not tell the full story, as seen in Fig. 14.15. They also fall short on pre-
dicting turn performance if the assumptions used to develop them are not
respected. Let’s consider the following windup turn example. The flight test
maneuver shown in Fig. 14.18 starts with the airplane wings leveled and the
thrust set to maintain altitude and speed. Then the bank angle is progress-
ively increased while the pilot tries to maintain airspeed, letting the
altitude change.
Figure 14.18a shows the variation of speed (true ground speed) and alti-
tude with time. Figure 14.18b shows the variation of the bank angle, the
recorded load factor, and the load factor computed using the initial turn per-
formance assumption that says the load factor is proportional to one over the
cosine of the bank angle per Eq. (14.3). This is a very dynamic maneuver. It
does show that our modeling of the load factor, as compared to the flight test
recorded data, holds up very well until the nose of the plane starts dropping
(Fig. 14.18d). Beyond reference time 52,427 s (approximately), too many of
the assumptions used to develop the load factor model are not respected
anymore (speed decreasing, altitude loss), and the load factor model from
Eq. (14.3) starts overestimating the load factor value of the maneuver as
recorded during the test.
Figure 14.18 shows that the airplane loses both airspeed and altitude
during the maneuver with increasing load factor. This is expected because
the airplane was trimmed for level flight (just enough thrust to counter
drag), so when the load factor was increased, the induced drag increased,
leading to a thrust deficit for this constant thrust maneuver. This loss of
speed/altitude combination could be different if the pilot had tried to main-
tain airspeed; that would have resulted in a greater altitude loss. If instead of
looking at the airplane instantaneous turn performance we were to consider
its energy state, what would it look like?
The concepts of energy height and specific excess power were explained
in Chapter 12. For convenience, we will rewrite these equations here. The air-
plane specific energy (also called energy height) is the airplane’s total energy
(potential plus kinetic) divided by the airplane’s weight

1
mgh þ mV 2 V2
He ¼ 2 ¼hþ (14:33)
W 2g

The specific excess power is the variation of that energy vs time. It can be
written in the following way:
 
T q CD0 K n2 ðW =SÞ dh V dV
Ps ¼ V   ¼ þ (14:34)
W W =S q dt g dt

Plotting the energy height and specific excess (deficit) power of the
airplane from Fig. 14.18, we would obtain the data presented in Fig. 14.19.
a) 37,200 500 b) 0 3
Bank angle n = 1/cos ()
Altitude –10
37,000 480 2.5

True ground speed (KTGS)

–20

Bank angle (deg)

36,800 460 –30 2

Load factor (g)

Altitude (ft) –40
36,600 440 1.5
KTGS –50
36,400 420 –60 Load factor 1

CHAPTER 14
–70
36,200 400 0.5
–80
36,000 380 –90 0
52,380 52,390 52,400 52,410 52,420 52,430 52,440 52,380 52,390 52,400 52,410 52,420 52,430 52,440
Reference time (s) Reference time (s)
c) d)

Turn Performance and Maneuver Margin

2.5 300 2.5 2.5

Pitch angle (deg), rate of climb (deg/s)

0 2
Rate of climb Pitch angle
2 –300 2 1.5
–600 1
Load factor (g)

Load factor (g)

Rate of climb (ft/min)
1.5 –900 1.5 0.5
–1200 0
1 –1500 1 0.5
Load factor Load factor
–1800 –1
0.5 –2100 0.5 –1.5
Rate of turn
–2400 –2
0 –2700 0 –2.5
52,380 52,390 52,400 52,410 52,420 52,430 52,440 52,380 52,390 52,400 52,410 52,420 52,430 52,440
Reference time (s) Reference time (s)

Fig. 14.18 Example of windup turn maneuver.

651
652 Operational Aircraft Performance and Flight Test Practices

46,000 2.6
Energy height 2.4
45,500 2.2
2
45,000
Energy height (ft)

1.8

Load factor (g)

1.6
44,500 1.4
1.2
44,000
1
Load factor
0.8
43,500 0.6
0.4
43,000 0.2
52,380 52,390 52,400 52,410 52,420 52,430 52,440
Reference time (s)

200 2.5

150 Load factor 2.25

Specific excess power (ft/s)

100 2

50 1.75
Load factor (g)
0 1.5

–50 1.25

–100 1

–150 Specific excess power 0.75

–200 0.5
52,380 52,390 52,400 52,410 52,420 52,430 52,440
Reference time (s)

Fig. 14.19 Energy height variation during windup turn maneuver of Fig. 14.18.

The load factor was added to the chart as a reference of where the airplane is
in the maneuver.
From Fig. 14.19, we can see the airplane was fairly well trimmed on wings
level condition with essentially a constant energy height (Ps near zero). Then
we see that as the maneuver begins and the load factor increases, the airplane
 CHAPTER 14 Turn Performance and Maneuver Margin 653

energy height starts to decrease rapidly (Ps becoming negative). The specific
excess power in this maneuver reaches a peak of –110 ft/s (–6600 ft/min
sink rate equivalent).
One can rapidly get a feel for an airplane’s maneuverability (both sus-
tained and instantaneous) by glancing at specific excess power curves
overlaid on a rate of turn vs airspeed chart for a given altitude, airplane
weight, thrust setting, and configuration. Figure 14.20, believed to be for

T.O. 1F-16C-1-1

Turn performance–Sea level

Data basis flight test
Engine F110-GE-100/Big inlet
Configuration:
Conditions:
• Drag index = 0
• Standard day
• GW =20,000 pounds
• Max AB
Note: Refer to section V for airspeed limitation.
GW effect Temperature effect Radius temp effect
E 1.1
0.9M –3
(degrees/second)

(degrees/second)

0.95M 0.6M J
–4

Turn radius factor

–2 1.2M
 Turn rate

 Turn rate

1.0M –1
10 20
–2 F F
–10 
0.6 1.0 –20 –10 0
–1 M 10 20
0.9

1.2M
D 1.2M
0 –2
5M

A
20 22 24 26 –3 G
GW–1000 Pounds Temp dev from std day–°C
Temp dev from std day–°C 0.9
–400 Quickest turn
8 (25.2 degrees/second)
7

0
6 –20
24
5
C
PS B
= 0F
Ps
–g

4
200
or
act

20
df

400
a

3
Lo

9
Turn rate-degrees/second

00
15

2 600
15
00
20

H I
Tightest turn
300 –400
(radius =
0
877 feet)
3 00
1
12
–20
0

0
200

00
40
00
400

10
8 00
60
Maximum airspeed

0
800

0 0
10,0

4
CAT III limiter

0
CAT I limiter

20,00
Turn radius
– feet

0
0 0.2 0.4 A 0.6 0.8 1.0 1.2 1.4
Mach number

Fig. 14.20 Specific excess power chart for energy maneuverability.

654 Operational Aircraft Performance and Flight Test Practices

an F-16C, was found on the Internet. No attempts were made to validate the
data; it is simply used to illustrate our point.
Lines of equivalent specific excess power are drawn (using a spacing of
200 ft/s typical for high-performance airplanes) to show the airplane’s
ability to maintain a given flight condition (Ps ¼ 0), to accelerate/climb
(Ps . 0), or to decelerate/descend (Ps , 0). For the chart in Fig. 14.20,
one can see that this particular plane, under the conditions listed at the
top of the chart (weight, altitude, temperature,
etc.), has 1000 ft/s specific excess power under a Note: A +200 ft/s specific
5-g, Mach 0.87 flight condition. excess power is the equivalent
As a side note, during air combat maneuvering of a 12,000 ft/min rate of
climb at constant airspeed or
(ACM), a fighter airplane will have an energy maneu-
as much as (2258/V) KTAS/s
verability advantage over another one if: speed gain for a constant
• It enters the engagement at a higher energy level altitude turn, where V is in
KTAS. This would translate
and maintains more energy than the other airplane to 7.53 KTAS/s speed gain
during the fight while going through 300
• It enters the engagement at a lower energy level KTAS, for example.
but can gain energy faster than the other airplane
To determine whether one fighter aircraft has the advantage, one air-
craft’s specific excess power curves can be superimposed on a similar plot
for its adversary.

Inner Wing Stall in a Turn

Aircraft with long wings and slow stall speed face an additional constraint
when performing a turn at a slow speed. The wing tip on the inside of the
turn may be subjected to much lower speeds than the wing tip on the
outside of the turn (see Fig. 14.21).

Airspeed (V)
Sideslip () = 0
r Vi = ri · 
b
2 cos()
ri

Fig. 14.21 Inner wing tip airspeed vs airplane CG airspeed.

 CHAPTER 14 Turn Performance and Maneuver Margin 655

The speed differential between inner and outer wing tips is

proportional to

Vo  Vi ¼ ẋ ðr þ b=2 cosðfÞÞ  ẋ ðr  b=2 cosðfÞÞ

pffiffiffiffiffiffiffiffiffiffiffiffiffi
g n2  1 b
¼ (14:35)
V n

From this equation, one can also determine that the inner wing tip is
slower than the airplane center of gravity by half the value of Eq. (14.35)
and will therefore be more subject to stalling first, inducing a roll
towards the inside of the turn. Even without stalling the inner wing tip,
there may be substantial lift difference between the inner and outer wing
that would tend to generate a rolling moment to increase the bank angle
more. This is counterbalanced by the use of ailerons to maintain the
bank angle.
As an example, consider tree airplanes with identical wing spans of 50 ft.
One airplane can maneuver at 100 KTAS, one at 40 KTAS, and one at 20
KTAS, and all three can handle up to 2 g (60-deg bank). Figure 14.22 was gen-
erated to show how much slower the inner wing tip was as compared to the
airplane center of gravity.
Although there are no regulations on stall speed in a turn, there are
requirements on the stall characteristics of the plane in a turn that must

0
Airplane speed
–2 100 KTAS
Inside turn wing tip speed minus
airplane CG speed (KTAS)

–4
40 KTAS
–6
Reference wing
span = 50 ft
–8

–10
20 KTAS

–12

–14
0 10 20 30 40 50 60
Bank angle (deg)

Fig. 14.22 Inner wing speed minus airplane CG speed.

656 Operational Aircraft Performance and Flight Test Practices

be proven by flight testing. For FAA Part 25 airplanes, typically larger with
higher stall and operating speeds, these are (underline added by author):
25.201 Stall Demonstration

(a) Stalls must be shown in straight flight and in 30 degree banked turns...

25.203 Stall Characteristics

(a) It must be possible to produce and to correct roll and yaw by unreversed
use of the aileron and rudder controls, up to the time the airplane is
stalled. No abnormal nose-up pitching may occur. The longitudinal
control force must be positive up to and throughout the stall. In addition,
it must be possible to promptly prevent stalling and to recover from a
stall by normal use of the controls.
(b) For turning flight stalls, the action of the airplane after the stall may
not be so violent or extreme as to make it difficult, with normal
piloting skill, to effect a prompt recovery and to regain control of the
airplane. The maximum bank angle that occurs during the recovery may
not exceed—
1. Approximately 60 degrees in the original direction of the turn, or 30
degrees in the opposite direction, for deceleration rates up to 1 knot
per second; and
2. Approximately 90 degrees in the original direction of the turn, or 60
degrees in the opposite direction, for deceleration rates in excess of 1
knot per second.

Under Part 23 (small airplanes), for airplanes that typically operate at slower
speeds and therefore are more susceptible to inner wing tip stalling, FAA
regulations state:
23.203 Turning flight and accelerated turning stalls
Turning flight and accelerated turning stalls must be demonstrated in tests as
follows:
(a) Establish and maintain a coordinated turn in a 30 degree bank. Reduce
speed by steadily and progressively tightening the turn with the
elevator until the airplane is stalled, as defined in §23.201(b). The rate of
speed reduction must be constant, and—
1. For a turning flight stall, may not exceed one knot per second; and
2. For an accelerated turning stall, be 3 to 5 knots per second with
steadily increasing normal acceleration.
(b) After the airplane has stalled, as defined in §23.201(b), it must be possible
to regain wings level flight by normal use of the flight controls, but
without increasing power and without—
1. Excessive loss of altitude;
2. Undue pitchup;
3. Uncontrollable tendency to spin;
4. Exceeding a bank angle of 60 degrees in the original direction of the
turn or 30 degrees in the opposite direction in the case of turning
flight stalls;
5. Exceeding a bank angle of 90 degrees in the original direction of the
turn or 60 degrees in the opposite direction in the case of accelerated
turning stalls; and
6. Exceeding the maximum permissible speed or allowable limit
load factor.
 CHAPTER 14 Turn Performance and Maneuver Margin 657

Maneuver Margin and Certification Regulations

From a certification point of view, it must be shown that the airplane has
sufficient load factor margin, as expressed by an equivalent bank angle, while
flying at approved operational speeds such as takeoff, en route climb,
and approach. This margin is not between the operational speed and the
stall speed, but rather between the operational speed and the stall warning
speed.
Under FAA 14 CFR Part 25, regulation 25.143(h) [Amendment 25-129]
states:
(h) The maneuvering capabilities in a constant speed coordinated turn at
forward center of gravity, as specified in the following table, must be free
of stall warning or other characteristics that might interfere with normal
maneuvering:

Manoeuvring
Bank Angle in A Thrust/Power
Configuration Speed Coordinated Turn Setting
TAKE-OFF V2 308 ASYMMETRIC
WAT-LIMITEDð1Þ
TAKE-OFF V2 þ xxð2Þ 408 ALL ENGINES OPERATING
CLIMBð3Þ
EN-ROUTE VFTO 408 ASYMMETRIC
WAT-LIMITEDð1Þ
LANDING VREF 408 SYMMETRIC FOR 238
FLIGHT PATH ANGLE
ð1Þ A combination of weight, altitude, and temperature (WAT) such that the thrust or power
setting produces the minimum climb gradient specified in §25.121 for the flight condition.
ð2Þ Airspeed approved for all-engines-operating initial climb.
ð3Þ That thrust or power setting which, in the event of failure of
the critical engine and without
any crew action to adjust the thrust or power of the remaining engines, would result in the
thrust or power specified for the takeoff condition at V2, or any lesser thrust or power setting
that is used for all-engines-operating initial climb procedures.

The European Aviation Safety Agency (EASA), under its equivalent Part 25
Certification Specification (CS), para 25.143(h) [Amendment 14], has
similar requirements to the FAA 14 CFR at Amendment 25-129. The same
is true for Transport Canada under Canadian Aviation Regulations (CAR)
Part V Subpart 525, para 525.143(h).

Takeoff Maneuver Margin

The minimum safety takeoff speed V2min is defined under FAA 14 CFR
25.107(b)(1) to be no lower than 1.13VSR . This theoretically provides a
margin of 1.2769 g (38.4-deg bank) to stall conditions. Certification require-
ments are that there be at least 30-deg bank margin (1.154 g under
658 Operational Aircraft Performance and Flight Test Practices

coordinated turn) available to stall warning in the takeoff configuration,

under a one-engine-inoperative condition, with takeoff flaps down and
gear up while flying in the published takeoff safety speed V2 under WAT
limits (minimum certification climb gradient as seen in previous classes).
This translates into a required minimum speed margin of
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2  nmargin VSW ! V2  1:0746 VSW (14:36)

The minimum acceptable margin between stall warning speed VSW and
reference stall speed VSR , per FAA 14 CFR 25.207(d), is 3 kt or 3%. This
margin must be validated by flight testing, a subject we will address a little
later in this chapter. Suffice to say at this time that some scatter in the
data will exist. Figure 14.23 shows typical test data collected with part of
the data being impacted by the 3-kt margin (below 100 KEAS) and some
data being impacted by the 3% margin (above 100 KEAS). It can be seen
that there is a very narrow band of allowable scatter in the test results if
one wants to both meet the stall margin requirements and be able to use
the minimum takeoff safety speed V2min . If the scatter is larger, the
minimum V2 that can be used may need to increase, with the resulting
impact on takeoff performance. (Higher V2 means longer takeoff distance.)
Because V2 is typically defined for an OEI condition and a single speed for
takeoff must be used for both OEI and AEO cases, the extra thrust of the
AEO takeoff case results in a larger acceleration for the plane, while the
crew follows essentially the same takeoff procedure for both cases. Therefore,

1.16
1.15
1.14
V2min, 25.107(b)(1)
1.13
1.12
1.11 30 deg bank
1.10
1.09 Vsw = V2min/1.0746
V/Vsr

1.08 SW data
1.07
1.06
1.05
1.04
25.207(d)
1.03
1.02 3 knots 3%
1.01
1.00
80 85 90 95 100 105 110 115 120
KEAS

Fig. 14.23 Stack-up of margins and example of flight test results scatter.
 CHAPTER 14 Turn Performance and Maneuver Margin 659

resulting takeoff speed would be larger than the V2 . (We will elaborate
more on the subject in Chapter 15.) The additional margin is typically
defined as an additive to V2 , often 10 KCAS higher. Let’s define that speed
as V35 . [That speed must be reached by the time the aircraft reaches 35 ft
above ground, per FAA 14 CFR 25.107(e)(1)(iii).] Under an AEO condition,
the maneuver margin requirements are that the airplane be able to reach
40 deg of bank (1.305 g under coordinated turn) under the same WAT con-
dition as the limiting OEI condition. This translates into a required minimum
speed margin of
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V35  nmargin VSW ! V35  1:1424 VSW (14:37)

Being a higher thrust condition and a faster one as well, the AEO V35 con-
dition is rarely limiting. The margin must nevertheless be demonstrated by
testing at the critical condition. If it is limiting, the speed margin to V2 can
be increased slightly with little to no impact on the takeoff distance (most
likely limited by the V2 OEI condition).

Landing Maneuver Margin

The requirement for maneuver margin in the landing phase requires that
the airplane be in the landing configuration (landing flaps and gear down,
plus any other device that must be used in that configuration), thrust set sym-
metrically for a flight path angle of –3 deg (sometimes shown as TF-3, which
stands for thrust for –3 deg). Then, while flying at VREF [1.23VSR minimum
per 25.125(b)(2)], the airplane must be able to bank 40 deg without

1.25
VREF min 25.125(b)(2)

1.2

40 deg bank
VSW ≤ VREF/1.1424
1.15
V/VSR

1.1

Stall warning data

1.05
25.207(d)

3 knots 3%
1
90 95 100 105 110 115
KCAS

Fig. 14.24 Landing configuration stack-up of margins and flight test data scatter..
660 Operational Aircraft Performance and Flight Test Practices

60 deg 45 deg
30 deg

15 deg

Runway Extended
runway
centerline

Fig. 14.25 Overshooting runway centerline and overcorrecting in bank.

encountering stall warning. The margin between stall warning and VREF
must therefore be (see also Fig. 14.24)
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi V
VREF  nmargin VSW ! VREF  1:1424 VSW ! SW  0:8753
VREF
(14:38)
Understanding what the maneuver margin really provides in terms of
protection during an approach is essential. Many accidents occur while the
pilot’s attention is diverted from the primary mission of flying the airplane
to address an event while flying low and slow. A typical problem is an airplane
turning on final, low altitude and slow speed, where the pilot overshoots the
extended runway centerline and then elects to bank more to come back on
centerline. The overbank generates a larger load factor and leads to the
wing stalling (see Fig. 14.25).

NTSB Identification: ANC11FA065, 17 July 2011, Piper PA-12. Witnesses

reported that the airplane was departing a private airstrip. It reached an alti-
tude of about 200–300 ft above the ground when the engine lost all power, and
the airplane began a left turn. During the turn, the airplane’s nose pitched
down abruptly, and the airplane collided with the ground in a steep, nose-
down attitude.... The accident site was adjacent to a road and about 300 yd
from a clear field. Despite these potential landing areas, the pilot decided to
attempt to return to the airstrip by initiating a turn at low altitude and low air-
speed following the loss of engine power, which resulted in an aerodynamic
stall and a loss of control with insufficient altitude to recover.
 CHAPTER 14 Turn Performance and Maneuver Margin 661

En Route Maneuver Margin

The en route configuration is OEI with flaps up and landing gear up, and
the remaining engine(s) at no more than maximum continuous thrust
(MCT). The speed provided by the manufacturer for this configuration typi-
cally provides the best climb gradient under the OEI condition. To ensure
this speed is not too low, it must be demonstrated that at the selected
speed, the airplane has at least 40-deg bank capability without encountering
stall warning.
NTSB provides an interesting link for further reading: https://www.ntsb.
gov/safety/safety-alerts/documents/SA_019.pdf.

Flight Testing for Maneuver Margin

It can be shown by analysis of the data that an airplane has sufficient man-
euver margin if there is little scatter in the stall warning data and the test
points fall near the minimum margins. That being said, the traditional way
of validating the margin has been to demonstrate it in flight.
The flight test demonstration must demonstrate that the airplane does
have the required minimum margin. To do so, the test crew first sets up
the test configuration (flaps, landing gear, airspeed, and thrust setting) with
wings level. (Note that some conditions will result in the airplane climbing
whereas for others the airplane may be descending.) Once conditions are
met, the pilot initiates a slow windup turn where the bank angle is increased
progressively until the value to be demonstrated is reached. Then some test
facilities demonstrate margin (i.e., how much more bank angle is available to
stall warning), whereas some just return to wings level.
Figure 14.26 provides an example of flight test results to verify the 40-deg
maneuver margin of an airplane. During this test, the pilot first stabilized at
the appropriate speed, aircraft configuration, thrust level, and wings level.
Then the pilot progressively increased the bank angle until the desired
target, 40 deg, was reached without adjusting thrust. The pilot stabilized
just above the 40-deg bank to show the condition was not marginal. After
a few seconds, when the test condition was deemed successful, the pilot pro-
ceeded to increase the bank angle further until the stall warning was reached.
When the stall warning was reached, the pilot backed out of the bank slightly
and then progressively increased it again to validate the warning and bank
angle margin over the 40-deg minimum requirement, about 8–9 deg in
this case. For this airplane, the stall warning was provided by artificial
means (stick shaker).
As can be seen from Fig. 14.26, there is a lot of small amplitude noise in
the load factor trace (elevator movement to increase load, ailerons movement
to track bank angle, atmospheric turbulence, etc.), so performing the test in
calm air will greatly increase the chances of having good results. The artificial
stall warning is usually triggered when a certain angle of attack is reached,
662 Operational Aircraft Performance and Flight Test Practices

16 0

14 –10
Stall warning angle
12 –20
Angle of attack (deg)

Bank angle (deg)

10 –30
Test
requirements
8 –40
Bank angle
6 –50
Bank angle
Angle of attack
4 –60

warning
Initial
2 –70

0 –80
0 20 40 60 80 100 120
Reference time (s)

2 0
Load factor (flight test data)
1.8
–10
n = 1/cos ()
1.6
–20
1.4

Bank angle (deg)

Load factor (g)

1.2 –30

1 –40
0.8
Bank angle –50
0.6
–60
warning
Initial

0.4
–70
0.2

0 –80
0 20 40 60 80 100 120
Reference time (s)

Fig. 14.26 Maneuver margin test.

so turbulence can affect local AoA readings. AoA vanes installed on a very
flexible fuselage may trigger the stall warning by the simple oscillation of
the fuselage. A stall warning system equipped with angle of attack phase
advance (lowering of the stall warning AoA due to AoA rate) may see an
earlier firing of the system during rapid entry into the stall. Flight testing
 CHAPTER 14 Turn Performance and Maneuver Margin 663

≥5%
§25.207(c) Maneuver margin, 40 deg bank
§25.143(h)
≥2%
if pusher Maneuver margin, 40 deg bank
≥3% §25.143(h)
installed §25.207(d)
§25.103(d) Maneuver margin, 30 deg bank Applicant
§25.143(h) selected

Vstall ID VSR VSW V2min V2+XX VREF

§25.201 §25.103(a) §25.207
≥13%
§25.107(b)(1)

≥23%
§25.125(b)(2)

Fig. 14.27 Low-speed margins for transport category airplane certification.

is done to verify overall stall protection system behavior and confirm

margins.

Summary of Certification Requirements: Low Speed

Figure 14.27 shows the impact of minimum margin on the certification
speeds for takeoff and landing defined by FAA 14 CFR Part 25 for transport
category airplanes. It clearly shows that the aim of requirements defined by
25.143(h) is to ensure the crew has sufficient capability to perform normal
maneuvering for this type of plane without encountering an attention-
grabbing stall warning.

Buffet Envelope
On the high-speed side of the flight envelope, as the airplane approaches
sonic speeds, conditions will exist where local airspeed on part of the airplane
will reach Mach 1.0 even while the airplane is still flying at subsonic speeds.
Beyond that point, shock waves will develop, most of the time on the top of
the wing first. The flow behind these shock waves is very turbulent and may
even separate from the wing. The flow separation will lead to loss of lift and
increase drag. For the loss of lift condition, Fig. 14.28 best illustrates the
general trend of what could be expected.
Some observations can be extracted from Fig. 14.28. As the airplane
accelerates from a low Mach number to a higher one, while at constant alti-
tude, the required lift coefficient decreases proportional to the inverse of the
airspeed squares (or Mach number squares). Initially, per the chart in
Fig. 14.7, the airfoil will produce more lift for a given angle of attack as the
 664
Operational Aircraft Performance and Flight Test Practices
.8 o Aerodynamic phenomena at various speeds
(deg) Subsonic Transonic Supersonic
.7 5
.6
4
.5 3 Subsonic
Lift coefficient, Ci

type
.4 2 airfoil
.3 1
.2 0

.1 –1 No shock waves Lambda-type shock waves Bow- and trailing-edge-

type shock waves
0
Steady flow pattern Fluctuating flow pattern Steady flow pattern
–.1 Small boundary layer Large boundary layer Moderate boundary layer
–.2 Small wake Large, violent wake Small wake
Readily susceptible to Susceptible by special Susceptible to experimental
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0
theoretical and experi- techniques to experi- analysis, and, the first order
Mach number, M mental analysis mental analysis effects, to theoretical analysis

Fig. 14.28 Extracts from NACA WR L-143 (1945) and NACA LAL 57359.
 CHAPTER 14 Turn Performance and Maneuver Margin 665

Mach number increases, thereby leading to reducing the angle of attack

needed. Beyond a critical Mach number (changing depending on the required
lift), the airfoil will start losing lift capability at a given angle of attack (flow
separation due to the presence of shock waves), requiring the angle of
attack to increase to maintain lift. To maintain level flight (given the lift coef-
ficient), the angle of attack must be increased, leading to stronger shock
waves, more flow separation, and possibly buffeting. Buffeting is a high-
frequency vibration of the airframe, caused by either the flow separation or
shock wave oscillation over the airframe. Buffeting is undesirable in flight.
From a certification point of view, FAA 14 CFR 25.251, Vibration and
Buffeting, requires the following:
(d) There may be no perceptible buffeting condition in the cruise configur-
ation in straight flight at any speed up to VMO/MMO, except that stall
warning buffeting is allowable.
(e) For an airplane with MD greater than .6 or with a maximum operating
altitude greater than 25,000 feet, the positive maneuvering load factors
at which the onset of perceptible buffeting occurs must be determined
with the airplane in the cruise configuration for the ranges of airspeed
or Mach number, weight, and altitude for which the airplane is to be
certificated. The envelopes of load factor, speed, altitude, and weight
must provide a sufficient range of speeds and load factors for normal
operations. Probable inadvertent excursions beyond the boundaries of
the buffet onset envelopes may not result in unsafe conditions.

To help understand the practical impact of this aerodynamic characteristic

on the airplane flight envelope, consider this example: An airplane has a
weight W of 230,000 lb and a reference wing area S of 1951 ft2 . To fly level
flight, this airplane would require a given lift coefficient depending on the
altitude and airspeed conditions. This airplane has a lift coefficient at
buffet onset as defined in Fig. 14.29. At low Mach numbers, the buffet
onset is typically triggered by stall conditions, whereas at higher Mach
numbers, the buffet is induced by shock waves.
Lift coefficient is defined as
nðW =SÞ nðW =dÞ
CL ¼ ¼ (14:39)
1 2 1
g pSL d M g pSL S M2
2 2
It can be noted from Fig. 14.29 that:
• For a given Mach number, the lift coefficient would increase under the
1.0-g condition as altitude is increased, as seen in previous chapters.
• For a given altitude, the lift coefficient would decrease as the Mach number
increases while flying under a 1.0-g condition.
• For a given altitude, the lift coefficient would increase if the load factor
were to increase.
The combination of weight, altitude, load factor, and Mach numbers
where the buffeting occurs is typically well behaved if the buffeting is a
666 Operational Aircraft Performance and Flight Test Practices

a) 2
Altitude: SL 10k 20k 30k 40k feet pressure altitude
1.8

1.6 Buffet envelope

Required lift coefficient (CL)

1.4

1.2

0.8

0.6

0.4

0.2

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Mach

b)
2
Load factor: 1.0 1.3 2.0 g
1.15 1.5
1.8
Buffet envelope
1.6
Required lift coefficient (CL)

1.4

1.2

0.8
30,000 ft pressure altitude
0.6

0.4

0.2

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Mach

Fig. 14.29 Required CL example. a) Left altitude effects; b) right load factor effects,
fixed altitude.

wing shock wave induced with a small variation with center of gravity effects;
for a conventional wing-tail layout, the most forward CG is the most unfavor-
able location (lowest buffet limits). If the data in Fig. 14.29 are replotted in
 CHAPTER 14 Turn Performance and Maneuver Margin 667

1,600,000

1,400,000

1,200,000

1,000,000
n (W/ ) (lb)

800,000

600,000

400,000

200,000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Fig. 14.30 [n(W/d)] capability vs Mach example.

terms of [n(W/d)] at buffet onset vs Mach number, one gets the chart shown
in Fig. 14.30.
From this curve, the transition from low-speed stall buffet to shock wave–
induced buffeting can clearly be seen, starting near Mach 0.70 and being very
strong past Mach 0.80.
This buffet envelope must be defined by flight testing. The FAA guidance
material on acceptable means of compliance to determine the buffet bound-
ary is found in Advisory Circular AC25-7D.

A determination of a buffet onset envelope is to be established for the ranges

of airspeed and/or Mach number, weight, altitude, and load factor for which
the airplane is to be certificated.
This boundary should be established by pilot qualitative evaluation or by
correlation with pilot qualitative evaluation, as there is no predetermined cri-
terion for buffet level at the pilot station. A normal acceleration of +0.05 g
has been used in some cases; however, the appropriate acceleration level
will vary from airplane to airplane and may also be affected by the
dynamic response of the accelerometer. If a measured normal acceleration
is to be used, the acceleration level and specific accelerometer should first
be correlated against a pilot’s assessment of the onset of buffet.

Per the guidance material, the buffet boundary is based on a pilot qualitative
assessment. It is often useful to record these data to help correlate the start of
vibration with an airplane load factor. For that purpose, an accelerometer
installed in the cockpit, often on the pilot seat rail, is used to help with
data reduction.
668 Operational Aircraft Performance and Flight Test Practices

Buffeting is a high frequency random vibration of, typically, low ampli-

tude. One must select the right accelerometer to collect the data so as to dis-
tinguish the information from the much lower frequency, motion-based load
factor. Figure 14.31 shows flight test data from two different accelerometers,
one selected to pick up the low-frequency airplane load factor and one
selected to pick up the high-frequency vibration. The first accelerometer is
typically installed close to the airplane center of gravity, whereas the
second one is installed in the cockpit. We can clearly see the difference in
the signal output and how selecting the right accelerometer can greatly
enhance collection of the test data for the purpose of identifying buffeting
over the lower amplitude noise or higher amplitude (and lower frequency)
load factor.
If one looks at the vibration signal, it can be seen that vibration is always
present on the airplane. At a given load factor (about 1.7 g for this example),
buffeting is triggered. The amplitude of the vibration signals shows a marked
change. We compare this signal to the FAA guidance of +0.05 g, and
although this is not the first time the vibration signals exceed that guidance
value, the clear change to a larger amplitude does indicate the presence of
buffet. For some airplanes, that onset is slower and may lead to more
scatter in the test data, as compared to a rapid onset of buffet, when one
tries to document test results and create a buffet onset boundary.

2 0.25

1.8 0.20

1.6 Load factor 0.15

1.4 0.10
Airplane load factor (g)

Cockpit vibration (g)

1.2 0.05

1 0

0.8 –0.05
AC25-7C
0.6 guidance –0.10

0.4 Cockpit vibration –0.15

0.2 –0.20

0 –0.25
0 10 20 30 40 50 60
Reference time (s)

Fig. 14.31 Buffet onset test.

 CHAPTER 14 Turn Performance and Maneuver Margin 669

Buffet margin testing typically involves a windup turn maneuver where

the airplane is first trimmed in level flight (constant thrust, altitude, and
speed), and then the bank angle is increased gradually to increase load
factor until buffeting is encountered.

Presenting Buffet Information to the Flight Crew

The information collected during flight testing must be provided to the
flight crew in an easy-to-use format. FAA 14 CFR part 25.1585(d), Operating
Procedures, requires, under subpara (d), that:
(d) The buffet onset envelopes, determined under §25.251 must be furn-
ished. The buffet onset envelopes presented may reflect the center of
gravity at which the airplane is normally loaded during cruise if correc-
tions for the effect of different center of gravity locations are furnished.

A chart that would reflect the impact of the CG location on the maneuvering
capability of the airplane could take the form shown in Fig. 14.32.
A chart like the one in Fig. 14.32 is typically read from left to right, enter-
ing the chart at the current indicated Mach number and going up until the
current cruising altitude is reached. It is then read towards the right, at the
CG reference line, following a parallel line up until the airplane current
CG is reached. Then it is read across until one finds the current airplane
weight. From there, read down; this is the maximum available load factor
(and bank angle) for the current flight condition.

Pressure altitude (1000 FT)

25
Note:
Reference line

27 1.0 lb = 0.4536 kg
Mmo limitation

29

31
b
33 0l
,00
42 0 lb
35 , 0 0
40 lb
0
37 ,00
36
b
39 0 0l
,0
32
41 2 3 4 5 lb
0
,00
43 28
0 0 lb
45 25,0 00 lb
23,0 ht
eig
ss w
Gro
8 7 6
9 11

10 1 12
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 20 25 30 35 40 45 1.00 1.25 1.50 1.75 2.00 2.25 2.50
Indicated mach number – MI CG location (% Mac) Load factor (g)
10 30 40 50 55 60 65
Bank angle (°)

Fig. 14.32 Typical buffet onset chart for an AFM with CG location impact.
670 Operational Aircraft Performance and Flight Test Practices

FT
- lb 600,000
ht

e-
eig

ud
W 550,000

l tit

el

T
T
FT
FT
T

T
re a

0F
0F

0F
lev
0F
500,000

000
000

,00

,00
,00
ss u

500
Sea

25

30
15,

20
10,
Pre
450,000
FT
400,000 ,000
35
350,000

T
0F
,00 FT
40 000
45,
Bank angle – deg
0 15 30 45 50 55 60

1.0 1.5 2.0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Load factor – g Buffet onset mach number

Fig. 14.33 AFM buffet onset chart for most adverse CG location.

Figure 14.32 was created with information that is readily available to the
crew in the cockpit [weight, altitude, Mach, bank angle, and CG, with the CG
typically being estimated from fuel burn curve defined in the aircraft flight
manual (AFM) weight and balance section]. A more aft CG increases load
factor capability (less download from the horizontal tail) for a given
weight, altitude, and Mach combination.
At the discretion of the manufacturer, the effects of the CG location can be
omitted from the chart, provided that the most adverse CG is selected to gen-
erate the chart. Then the chart could take on the form shown in Fig. 14.33.
Figure 14.33 is formatted slightly differently than Fig. 14.32, in that one
enters the chart at the desired bank angle (load factor), going up until the
current airplane weight is reached. Then, reading across to the current
cruise altitude, one may find two intercepts that represent the low and
high speed at which buffet onset is reached for the given load factor,
weight, and altitude [n(W/d)] combination.

EASA Additional Requirement

EASA adds additional requirements in the Acceptable Means of
Compliance (AMC), Book 2, of CS25:
AMC25.251(e) 2:
2.1 CS 25.251(e) requires that the envelopes of load factor, speed, altitude
and weight must provide a sufficient range of speeds and load factors for
normal operations.
2.2 An acceptable means of compliance with the requirement is to establish
the maximum altitude at which it is possible to achieve a positive normal
acceleration increment of 0.3 g without exceeding the buffet
onset boundary.
 CHAPTER 14 Turn Performance and Maneuver Margin 671

1,600,000

1,400,000
0.3 g
1,200,000 margin

1,000,000
n (W/ ) (lb)

800,000

600,000

400,000

200,000

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Mach

Fig. 14.34 New [n(W/d)] curve for 0.3-g maneuvering capability.

The impact of the 0.3-g normal acceleration (load factor) margin is to push
the available buffet curve down by a factor of 1.3. Using the model in
Fig. 14.30, the curve is pushed down as shown in Fig. 14.34.
One can see from Fig. 14.34 that adding a 1.3-g margin lowers the avail-
able [n(W/d)] at all speeds, and the available speed range is also reduced for a
given [n(W/d)].
This information now needs to be included in AFM for EASA-certified
airplanes. It is presented in the form of maximum weight at altitude for a
given Mach number, the Mach number representing a typical cruise speed.
Such a chart could take on the form shown in Fig. 14.35.

350,000

Mach 0.80
300,000
Max altitude for 1.3-g capability
Weight (lb)

250,000

200,000

150,000

100,000
30,000 32,000 34,000 36,000 38,000 40,000 42,000
Pressure altitude (ft)

Fig. 14.35 Airplane limit altitude for a given weight range and Mach 0.80.
672 Operational Aircraft Performance and Flight Test Practices

Note that, although Fig. 14.35 provides a maximum weight at altitude for
a given speed, it does not account for the available airplane thrust and there-
fore does not guarantee that the airplane can reach the combination of
weight, altitude, and speed that provides the 1.3-g margin to buffet.

Unrestricted Turn Performance

In the previous sections, we built models using assumptions of coordi-
nated unaccelerated turn with no loss of altitude and constant bank angle.
We then treated the problem from an energy point of view where we
allowed a certain excess or deficit in thrust and looked at the impact on
the turn performance. One assumption we made previously was to neglect
the effect of the angle of attack of the thrust vector on the turn performance.
Let’s now quantify the effect of pointing the thrust vector towards the
inside of the turn (inducing sideslip, uncoordinating the turn) as well as
taking into account the effect of angle of attack on the thrust vector.
Figure 14.36 shows the forces and angles during a yawed turn in level flight.
From Fig. 14.36, the forces along the flight path are
T cosðaT Þ cosðbÞ ¼ D (14:40)
where b is the sideslip angle; for this exercise, the sideslip angle lies in the
horizontal plane instead of the more traditional airplane reference axis water-
plane. The vertical forces are
L cosðfÞ þ T sinðaT Þ cosðfÞ ¼ W (14:41)
Finally, the forces along the radius of the turn are
W
L sinðfÞ þ T sinðaT Þ sinðfÞ þ T cosðaT Þ sinðbÞ ¼ V ẋ
g
W V2
¼ (14:42)
g r

T Horizontal
D plane

T

V

Fig. 14.36 Forces and angles in a yawed level turn.

 CHAPTER 14 Turn Performance and Maneuver Margin 673

If one solves for the radius of turn

V2
r¼ (14:43)
g ½n sinðfÞ þ ðT =W ÞfsinðaT Þ sinðfÞ þ cosðaT Þ sinðbÞg

And of course, the rate of turn can be computed using Eq. (14.9).
The angle of attack is tied to a combination of airspeed, airplane weight,
load factor, altitude, and other factors. As well, the maximum expected angle
of sideslip in service is typically limited by either structural limits or a
maximum normally expected in service. To best illustrate the impact of
adding angle of attack and angle of sideslip onto the radius of turn model,
we will define a model airplane that is typical of a jetliner configuration.
Our model airplane will have a wing loading (W/S) of 100 lb/ft2 , an
available thrust-to-weight (T/W ) up to 0.3, and a drag coefficient (CD) of
0.022 þ 0.0468 CL2 for flaps up, gear up configuration. For this airplane,
we get a lift curve slope for low Mach number per Fig. 14.37. For this type
of airplane, the maximum sideslip expected would be defined by the airplane
encountering a 25-kt crosswind. The resulting curve is shown in Fig. 14.37.
To simplify the discussion, it will be assumed that the thrust line angle of
attack is equal to the airplane angle of attack for the purpose of computing
the impact of thrust on the turn. The first condition we will consider is a low-
speed, 1.23VSR . Under this condition, the airplane should be capable of at
least 40 deg of bank per the maneuver margin section. The load factor at
40-deg bank is n ¼ 1.3054.
For the flaps up position, with a maximum lift coefficient of 1.5, the
1.23VSR condition results in an airspeed of 172.6 KTAS under sea level stan-
dard conditions. With the required lift coefficient under 1.23VSR and a
40-deg bank, and a lift coefficient of 1.29, from the chart, we can see this is
about a 9.9-deg angle of attack. The radius of turn with the simplified
model [Eq. (14.9)] is 3144 ft. To account for the angle of attack, one would
compute a required thrust T/W of about 0.14 and a resulting radius of

16 2.5
Flaps DOWN
14
Based on 25 kt 2
12
Sideslip angle (deg)

Lift coefficient

crosswind
10 1.5
Flaps UP
8
1
6
4
0.5
2
0 0
100 150 200 250 300 350 0 5 10 15 20
Airspeed (kt) Angle of attack (deg)

Fig. 14.37 Model airplane aero characteristics.

674 Operational Aircraft Performance and Flight Test Practices

T2 T3
T1

Failed
T4

yeng
DW

Fig. 14.38 Asymmetric thrust on a multiengine airplane.

turn of 3087 ft. This represents a little less than a 2% reduction. Is 2% signifi-
cant? Using the simplified model, we can compute an equivalent load factor
required to achieve a 3087-ft turn radius; the value would be 1.315 or the
equivalent of 40.5-deg bank, clearly within the capability of a flight crew to
track a given bank angle in a turn. Therefore, using the simplified model pro-
vides an adequate means of estimating the airplane’s turn performance.
If we add the maximum expected sideslip in service (about 8.2 deg at
172.6 KTAS), the radius of turn reduces to 3018 ft—a 4% reduction com-
pared to the simplified model.
On the high-speed side, for the same 40-deg bank angle, the required lift
coefficient would be 0.35 with a corresponding angle of attack of 0.54 deg.
The simplified model radius of turn is 11,492 ft. If we are to take into
account the angle of attack, the radius of turn would become 11,480 ft—
the equivalent of 0.03 deg of bank using the simplified model. Finally, if we
add the maximum expected sideslip (about 4.3 deg at 330 KTAS), the
radius of turn would be 11,337 ft.

Turn Performance under OEI Conditions

For multiengine airplanes, the loss of thrust from one engine (OEI con-
dition) results in the generation of a yawing moment due to the thrust asym-
metry. That yawing moment can often be compensated for by the application
of rudder deflection when the airplane is flying above its minimum control
 CHAPTER 14 Turn Performance and Maneuver Margin 675

speed. Depending on the thrust asymmetry, the rudder used, the airplane
bank angle, and many other parameters, a sideslip angle may result. This
will typically be small (within +5 deg), but this does mean the turn will
not be a coordinated turn.
The yawing moment generated by the thrust asymmetry is, for a four-
engine airplane per Fig. 14.38,
1
N¼ r V 2 S b CnT ¼ ðT4  T1 Þ youtboard þ ðT3  T2 Þ yinboard (14:44)
2
Note that if an engine fails, it will produce windmilling drag Dw ; this is
negative thrust.

Exercises
1. The airplane in Example 14.1 took 2 min to perform a 360-deg turn while
flying at 200 KCAS and 3000 ft pressure altitude in a standard atmosphere.
What would be the radius of turn, in miles, and the total distance covered
during the same time?
2. It is said that the SR-71 (see Fig. 14.39) could easily cruise at Mach 3.2 and
78,000 ft pressure altitude in a standard atmosphere. What would be the
turn radius of the plane, in miles, in these conditions for a 30-deg banked
turn? How much time would it take to do a 360-deg turn? How much
distance would the plane cover during this turn? Here is an interesting link
on the SR-71: http://www.916-starfighter.de/SR-71_Waever.htm.
3. From Example 14.5, change the parasitic drag coefficient (CDo) from
0.025 to 0.020; all other characteristics remain unchanged. Produce the

Fig. 14.39 NASA artwork of the SR-71.

676 Operational Aircraft Performance and Flight Test Practices

same table and identify the conditions for minimum sustainable radius of
turn and maximum sustainable rate of turn.
4. Again for Example 14.5, the airplane now flies at sea level standard
condition with a T/W of 0.35; all other coefficients remain unchanged.
Produce the same table and identify the conditions for minimum
sustainable radius of turn and maximum sustainable rate of turn.

Reference
[1] Asselin, M., An Introduction to Aircraft Performance, AIAA Education Series, AIAA,
Reston, VA, 1997.
Chapter 15 Takeoff
Performance

Chapter Objective
This chapter will define what can be used as a takeoff distance for transport
category airplanes and will show the building blocks required to create a
model to compute such distance. Takeoff performance is heavily regulated,
so the various speeds influencing the takeoff distance will be defined
and explained.
Our discussion and analysis will be restricted to tricycle-type landing gear
layout where the pitch attitude is essentially constant during the initial phase
of the takeoff and is increased to lift off the ground. We will cover the
all-engines-operating (AEO) “normal” takeoff and the case of a continued
takeoff following the failure of one engine [one engine inoperative (OEI)].

Takeoff Distance Definition

T
here are several ways to define a takeoff distance, but all involve
at a minimum an acceleration from a low speed to the point where
the airplane becomes airborne. This distance must not be longer
than the runway available for takeoff, of course; because many parameters
impact the required takeoff distance, some safety margins are usually
applied. For transport category airplanes (FAA Part 25), this is defined by
§§25.113 and 25.111:
§25.113 Takeoff distance and takeoff run.
(a) Takeoff distance on a dry runway is the greater of—
(1) The horizontal distance along the takeoff path from the start of
the takeoff to the point at which the airplane is 35 feet above the
takeoff surface, determined under §25.111 for a dry runway; or
(2) 115 percent of the horizontal distance along the takeoff path, with
all engines operating, from the start of the takeoff to the point at
which the airplane is 35 feet above the takeoff surface, as determined
by a procedure consistent with §25.111.
(b) Takeoff distance on a wet runway is the greater of—
(1) The takeoff distance on a dry runway determined in accordance with
paragraph (a) of this section; or

677
678 Operational Aircraft Performance and Flight Test Practices

(2) The horizontal distance along the takeoff path from the start of the
takeoff to the point at which the airplane is 15 feet above the takeoff
surface, achieved in a manner consistent with the achievement of V2
before reaching 35 feet above the takeoff surface, determined under
§25.111 for a wet runway.

§25.111 Takeoff path.

(a) The takeoff path extends from a standing start to a point in the takeoff at
which the airplane is 1,500 feet above the takeoff surface, or at which the
transition from the takeoff to the en route configuration is completed
and VFTO is reached, whichever point is higher. In addition—
(1) The takeoff path must be based on the procedures prescribed in
§25.101(f);
(2) The airplane must be accelerated on the ground to VEF , at which
point the critical engine must be made inoperative and remain
inoperative for the rest of the takeoff; and
(3) After reaching VEF , the airplane must be accelerated to V2 .
(b) During the acceleration to speed V2 , the nose gear may be raised off the
ground at a speed not less than VR . However, landing gear retraction may
not be begun until the airplane is airborne.
(c) During the takeoff path determination in accordance with paragraphs
(a) and (b) of this section—
(1) The slope of the airborne part of the takeoff path must be positive at
each point;
(2) The airplane must reach V2 before it is 35 feet above the takeoff
surface and must continue at a speed as close as practical to, but not
less than V2 , until it is 400 feet above the takeoff surface;

From these certification requirements, four takeoff distances are defined, two
for dry runways and two for wet runways. What can be extracted from these
is the following:
• All takeoff distances start from a standing start (zero ground speed).
• The takeoff distance is defined from the standing start to the point where
the airplane reaches 35 ft above ground level (AGL) for a dry runway and
15 ft AGL for a wet runway.
• For an AEO condition, a safety margin of 15% of distance is added (i.e., the
runway must be at least 15% longer than the distance from a standing
start to the point where the airplane reaches 35 ft above the runway).
(See Fig. 15.1.)

+15%

AEO
OEI 35 ft

Takeoff distance

Fig. 15.1 Transport category takeoff distance on dry runway.

 CHAPTER 15 Takeoff Performance 679

• Regulations also require that the takeoff distance with one engine
inoperative (OEI) be computed. In this case, the “critical” engine is
assumed to fail at the most critical speed VEF , and the airplane is allowed to
continue the takeoff. No distance margin is applied for this condition.
• The airplane must reach a takeoff safety speed V2 before reaching 35 ft
AGL (dry or wet runway).
• The crew must use an approved takeoff procedure.
• The available runway must be at least as long as the defined AEO takeoff
distance or OEI takeoff distance, whichever is the longest.
FAA Part 23 small airplanes and military requirements define takeoff dis-
tance differently; we will discuss the differences later in this chapter.

Basic Takeoff Performance Model: Takeoff Segments

The creation of a mathematical model for the takeoff distance is easiest
when the takeoff is divided into segments of similar airplane behavior. For
the AEO case, we break down the takeoff as follows:
1. Ground run, all gears on ground, from brake release (ground speed ¼ 0)
to the start of rotation (airspeed reaching VR )
a. The thrust is set prior to brake release to ensure maximum thrust is
available for the start of the acceleration. This is seen in Fig. 15.2 by
an increasing engine fan speed (N1) prior to brake release.
b. When thrust is set (stable N1), the brakes are released. Sometimes a
small pitch wobble is seen as the airplane starts accelerating, as seen
by the small pitch rate fluctuation at brake release in Fig. 15.2.
c. The airplane accelerates, all wheels are on the ground, and there is
essentially a constant pitch attitude.
d. This segment ends at the start of rotation, seen in Fig. 15.2 as a sharp
elevator input.
2. Acceleration from VR to the liftoff of the last main landing gear (VLOF )
a. During this segment, the nose of the plane is rotated to increase the
airplane’s angle of attack and generate the required lift to lift off.
b. This is typically achieved by the pilot providing a longitudinal control
input (elevator deflection).
i. Elevator input is shown in Fig. 15.2.
ii. Enough elevator is set to reach a target pitch rate (typically around
3 deg/s) until a target pitch attitude is reached.
c. The nose landing gear comes off the ground soon after the longitudinal
control input.
i. In Fig. 15.2, note the small time difference between elevator input
and pitch attitude change.
d. The segment ends at the last main gear off the ground.
i. The airplane lifts off when sufficient lift is produced.
 680
Operational Aircraft Performance and Flight Test Practices
KTAS (kt) Wheel spd (kt) AEO takeoff
Wheel spd (kt) 200
KTAS (kt)
150
100
50
0

Elevator (deg)
10
Elevator (deg)

0
–10
–20

Pitch angle (deg) Pitch rate (deg/s)

Pitch angle (deg)

Pitch rate (deg/s)

30 10
20 5
10
0 0
–10 –5

RADALT (ft)
100
RADALT (ft)

50
0

N1 left (%)
N1 left (%)

200
100
0
150 200 250 300 350 400 450
Thrust Brake Time (s)
Rotate 35 ft
set release LOF

Fig. 15.2 AEO takeoff.

 CHAPTER 15 Takeoff Performance 681

ii. This is best seen by looking at wheel speed on the test airplane.
When the gear lifts off the ground, the wheel speed starts to decay.
3. Acceleration from VLOF to V2 and simultaneous climb to 35 ft AGL
a. During this segment, the pitch may continue to increase.
b. The landing gear may be selected up and be in transition.

Modeling the Ground Run: AEO

The forces acting on an airplane during the ground run are similar to the
ones seen in flight in that:
• The airplane weight W acts along the local vertical.
• Thrust T is required to accelerate.
• There are aerodynamic forces of lift L and drag D that exist when the
airspeed is nonzero.
In addition to these forces, ground forces exist when the wheels are on the
ground. These are:
• A normal ground reaction force N opposing the weight of the plane before
sufficient lift is generated to sustain flight. This force is distributed over all
wheels on the ground. For the purpose of performance modeling, it is
represented by a single force vector.
• There is a rolling friction force f between the tires and the ground acting in
the opposite direction to the movement of the airplane.
Figure 15.3 illustrates all of these forces.
The ground reaction force N acts on the landing gears to counteract the
airplane weight on wheel. It acts perpendicular to the ground. For a runway
with zero gradient, it takes of the following form:

N ¼W L (15:1)

D T

f
N
W

Fig. 15.3 Ground run forces.

682 Operational Aircraft Performance and Flight Test Practices

The sum of the moments will vary during the takeoff run and will
impact the ground reaction force on each individual landing gear, but the
variation of the lift force during the ground run will have the most
impact on the total weight on wheel. We are limiting our analysis to tricycle
landing gear configurations, so we can expect that the pitch attitude during
the ground run will be essentially constant, as can be seen in Fig. 15.2. The
constant pitch, and therefore angle of attack on the ground, will result in a
constant lift coefficient available CLa that depends on the takeoff configur-
ation (flaps) selected; we will define that lift coefficient for the ground run as
CL;TO . The lift produced is then simply proportional to the lift coefficient
times the dynamic pressure and the reference wing area. Figure 15.4
shows a typical variation in lift coefficient from the ground run to after
takeoff.
To get an idea of the evolution of the normal force during an AEO
takeoff, we produced Fig. 15.5, which shows main landing gear (MLG)
and nose landing gear (NLG) shock strut displacement vs ground speed.
These displacements are for a high thrust line engine with respect to the
center of gravity, as can be seen by the initial compression of the NLG
strut during thrust application (before brake release). These represent
the same takeoff conditions as in Fig. 15.2. Note how the load on the
NLG is reduced once the brakes are released. Then note the general
reduction in NLG and MLG loads (decreased shock strut compression)
as the speed increases. Once the rotation starts, the NLG load decreases
rapidly, followed by the MLG loads reduction. Note also the very small
increase in MLG load at rotation (around time 307 s), which is typical of

2
Lift coefficient required (CL1g)
1.8 for 1-g flight
n = CLa /CL1g
1.6
1.4
Lift coefficient

1.2
1
Typical lift coefficient (CLa)
0.8 during a takeoff run
0.6
0.4
Constant AoA on ground Rotation
0.2
0
100 110 120 130 VR 140 VLOF 150 160 170
Airspeed

Fig. 15.4 Typical lift coefficient variation during a takeoff.

 Wheel spd (kt) Elevator (deg)
160 5
140
0
120
Wheel spd (kt)

Elevator (deg)
100 –5
80
60 –10

40
–15
20
0 –20
MLG shock strut (in) NLG shock strut (in) Pitch angle (deg)
25

15 20

CHAPTER 15
Pitch angle (deg)
MLG shock strut (in)
NLG shock strut (in)

15
10
10
NLG
LOF 5
5

Takeoff Performance
0

0 –5
200 250 300 350 400
Thrust Brake Time (s) Rotate
set release MLG
LOF

Fig. 15.5 MLG and NLG compression during takeoff.

683
684 Operational Aircraft Performance and Flight Test Practices

an aft tail airplane where a download is applied (increased apparent

weight) to start lifting the nose; this download on the gear reduces
very rapidly as the wing lift quickly increases with increasing angle of
attack.
The rolling friction force f acting on the landing gear is opposed to the
airplane’s movement. It is proportional to the ground reaction force (to the
weight on wheel). It can be modeled in the following way:

f ¼ mr N (15:2)

where mr is the rolling friction force coefficient. The rolling friction force
coefficient for airplane tires is mostly due to deformation hysteresis where
the energy of deformation of the tire (see Fig. 15.6) is greater than the
energy of recovery. This deformation energy loss results in the tire pressure
increasing during taxi and takeoff. Underinflation of the tires may lead to
failure. Some other energy loss comes in the friction of the wheel system
and from small slippage of the tire.
Typical values of rolling friction force coefficients on hard surfaces are:
• Dry asphalt and concrete: 0.015 to 0.05
• Wet asphalt and concrete: 0.05
The larger values on dry surfaces are typical of lower pressure tires.
On wet runways, some additional work is done to displace the small
amount of water on the runway surface. A coefficient of 0.02 is typical for
dry runway and will be used for most of the discussions. Note that per Eq.
(15.2), a coefficient of 0.02 means that the actual rolling friction force is
2% of the weight on wheels.
Just as the lift coefficient was determined to be essentially constant on
the ground, so is the drag coefficient. For the takeoff ground run, the drag

A


C E

R
B
D F 
B

Fig. 15.6 Tire deformation under load. Source: NASA CR3629, 1982.
 CHAPTER 15 Takeoff Performance 685

NTSB DCA08MA098: On 19 Sept. 2008, a Learjet 60, tail number N999LJ,

overran runway 11 during a rejected takeoff at Columbia Metropolitan
Airport, South Carolina. The National Transportation Safety Board deter-
mined the probable cause(s) of this accident to be the operator’s inadequate
maintenance of the airplane’s tires, which resulted in multiple tire failures
during takeoff roll due to severe underinflation, and the captain’s execution
of a rejected takeoff (RTO) after V1, which was inconsistent with her training
and standard operating procedures.

Photo source: Associated Press.

coefficient can be expressed in a similar way as the flight drag coefficient,

specifically
2
CD,TO ¼ CDo,TO þ F KTO CL,TO (15:3)
The parasitic drag coefficient CDo;TO is for the takeoff flaps deployed and
gear down. The induced drag parameter KTO represents the inflight (far from
ground) parameter with takeoff flaps and gear down.
The ground effect factor F is discussed in details in Chapter 19 and
takes the form specified by Eq. (19.6). Suffice to say here that while all
gears are on the ground, the induced drag of the airplane will be much
lower than that in flight away from the ground. During the ground run,
this factor is constant.

Summing It Up!
If we sum up the forces along the runway, we get the following equation:
W dV W dV
T Df ¼ ¼ V (15:4)
g dt g dX
686 Operational Aircraft Performance and Flight Test Practices

Because we expect an acceleration, and the aim of the exercise is to

predict takeoff distance, we rewrote the acceleration term (dV/dt) as a func-
tion of change in distance (V dV/dX ). If we further expand the equation into
its various components, we get
 
1 2 1 2
T  rSL s VTAS S CDo,TO  mr W  rSL s VTAS S CL,TO
2 2
 
W dV
¼ V (15:5)
g dX TGS
The drag force and lift force are proportional to the square of the true air-
speed around the airplane, so we added the subscript TAS to the airspeed
term for clarity. Although the sign of the drag contribution to the airplane
acceleration is negative (retarding forces), under certain tailwind conditions
(typically less than 10 kt for takeoff), the drag may actually be contributing
to the acceleration for a very short time. The last part of Eq. (15.5) uses sub-
script TGS for true ground speed because the acceleration is an inertial force.
The last part of the equation can be written in terms of TAS and wind speed
as follows:
 
dV dV
V ¼ ðVTAS  VW Þ (15:6)
dX TGS dX
where a headwind has a positive value for this equation; therefore, the ground
speed is the true airspeed minus the headwind component. The dV term also
represents the change in inertial forces, not the change in true airspeed. The
wind is never steady in direction or magnitude, so any testing for takeoff dis-
tance is typically best done under no wind conditions; we will discuss this
further in the next chapter.
To get a feel for the relative impact of each force component during
the ground run acceleration, we will review the forces of Fig. 15.7. The
figure shows typical forces from brake release to liftoff during a takeoff on
a level runway under no wind conditions for a Part 25 transport category
airplane.
The first force of interest is the thrust T. It is usually maximum at brake
release and reduces at a given lapse rate (change of thrust vs change of
airspeed) during the acceleration. For Part 25 airplanes, this force is of the
order of 25–35% of the airplane maximum takeoff weight under AEO con-
ditions; it is the dominant force of the takeoff.
The next major force is the rolling friction force f. On a hard and
dry runway, this force is of the order of 2% of the airplane weight, as
specified earlier in this chapter. This force, which is proportional to the
weight on wheels, decreases as the airspeed increases and the lift increases.
That force typically does not decrease much for the ground run (brake
release to start of rotation), and then goes down rapidly from rotation to
liftoff.
 40,000
Wheel speed (kt) KTAS (kt)
200

Wheel speed (kt)

35,000
T 150

KTAS (kt)
100
30,000
50
0
25,000
Nx (g)

Rotation speed
Forces (lb)

0.6

Liftoff
20,000
Excess 0.4

Nx (g)
thrust 0.2
15,000
0.0
–0.2
10,000

CHAPTER 15
RADALT (ft)
(f + D)

RADALT (ft)
5000 50
D
f

Takeoff Performance
0 0
0 20 40 60 80 100 120 140 160 20 25 30 35 40 45 50
Airspeed (kt) Time (s)
35 ft
Rotation
Liftoff

Fig. 15.7 Forces during a takeoff ground run, level runway.

687
688 Operational Aircraft Performance and Flight Test Practices

Finally, the drag of the plane will increase proportional to the square of
the airspeed during the ground run (pitch attitude essentially constant,
therefore constant drag coefficient) and much more rapidly during the
rotation to liftoff as the lift coefficient rapidly increases with increasing
angle of attack.
From the previous information, including Fig. 15.7, we can observe that
the excess thrust (what provides the acceleration for the takeoff) will be
largest at brake release and will decrease as the airplane accelerates. One
means of visualizing this excess thrust is to look at the acceleration along
the runway. The acceleration recorded in Fig. 15.7 (parameter Nx) shows
the largest excess thrust at brake release (time 21 s) slowly reducing as the
airplane accelerates.
To compute the ground run distance (from brake release to rotation), we
can rewrite Eq. (15.5) in the following form:
W
ð XR ð VR (VTAS  VW ) dV
g
XR ¼ dX ¼
0 VW 1 2
T  mr W  rSL s VTAS S (CD,TO  mr CL,TO )
2
(15:7)
Note how the integration does not start from zero airspeed, but rather
from the wind component along the runway to the rotation airspeed. To
solve this equation, we will need to make some assumptions. They are:
• Distance is measured from a standing start (zero ground speed).
• Thrust is set to maximum prior to brake release.
• The winds are constant, both in magnitude and direction, during the
acceleration.
• If the wind variation at the airplane vs time is known, then the actual
winds can be included in the equation to be solved.
• The weight of the plane is considered constant during the ground run.
• This is an acceptable assumption because the weight will typically reduce
less than 1% over the course of the acceleration.
• The rolling friction force coefficient is considered constant.
With this, Eq. (15.7) can now be solved numerically.
Another way to write Eq. (15.7) is
ð VR
(VTAS  VW ) dV
XR ¼ (15:8)
VW a

where a is the airplane acceleration.

 
1 2
g T  mr W  rSL s VTAS S (CD,TO  mr CL,TO )
2
a¼ (15:9)
W
 CHAPTER 15 Takeoff Performance 689

0.45
Measured
0.4
Smoothed
0.35

0.3
Average
0.25
Nx (g)

0.2

0.15

0.1

0.05
V 2R
0
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000
TAS2 [(ft/s)2]

Fig. 15.8 Determination of average acceleration during ground run.

The parameters in Eq. (15.9) have known behaviors vs airspeed and can
allow us to determine an average acceleration ā during the ground run. With
this average acceleration now being a constant value, Eq. (15.8) can be solved
as follows:
ð VR
1 1
XR ¼ ðVTAS  VW Þ dV ¼ ðVR  VW Þ2 (15:10)
ā VW 2 ā

The ground distance is proportional to the square of the rotation speed.

We will soon address VR to see what is allowable for FAA Part 25 transport
category airplanes. One simple way to determine the average acceleration is
to determine the thrust, drag, and lift at an airspeed equal to 12(VR – VW )2 [or
about 0.707(VR – VW )]. This is shown in Fig. 15.8. The resulting value has an
accuracy of approximately 1% compared to the more precise numerical
integration.
We can make one further simplification to the takeoff distance and accel-
eration equations. From Fig. 15.7, we can observe that the retarding forces do
not change much during the ground run and are of the order of the ground
rolling friction force. Therefore, if we neglect the aerodynamic forces of drag
and lift, Eq. (15.9) simplifies to
 
T
ag  mr (15:11)
W
690 Operational Aircraft Performance and Flight Test Practices

Combining with Eq. (15.10) and zero winds leads to a simplified ground run
distance of

VR2
XR    (15:12)
T
2g  mr
W

And, with basic kinematic equations, we find the time required to accelerate
to rotation speed from brake release to be

2 XR VR
tR     (15:13)
VR T
g  mr
W

Example 15.1
You are the pilot of a commercial jetliner about to get on a runway for takeoff.
You notice that the winds are calm (essentially zero) and you think, why not
try to estimate the airplane rotation speed and average T/W for the takeoff.
You align the airplane on the runway end where the distance remaining
sign shows 4000 ft, and then apply thrust. When set, you release the brakes
and start the timer. As the airplane reaches the 2000-ft runway marker, you
feel the nose coming up, so you stop the timer. It reads 20.7 s. What is the
approximate rotation speed and the average T/W for this takeoff?

Solution: We don’t have much in terms of airplane description, but we have

time and distance. We also know the airplane started from a standing start at
the beginning of the runway. Let’s assume a hard and dry runway with a rolling
friction force coefficient of 0.02 (typical value). We will use this to compute
the average T/W for the takeoff. From basic kinematic equations we first

(Continued)
 CHAPTER 15 Takeoff Performance 691

Example 15.1 (Continued)

find the average acceleration for the takeoff run described
2 XR 2  2000 ft ft
a¼ ¼ ¼ 9:33 2 ¼ 0:29 g
tR2 ð20:7 sÞ2 s

We also know that VR will be

2 XR ft
VR ¼ ¼ 193:2 ¼ 114:5 kt
tR s
Combining this with Eq. (15.11), we can extract the average T/W during the
acceleration.
a
T =W ¼ þ mr ¼ 0:29 þ 0:02 ¼ 0:31
g

Rotation Speed (VR )

Before going any further, it might be worth looking at the limits imposed
on VR by FAA Part 25. The certification authorities do allow the applicant
(the airplane manufacturer) to select VR , and typically a lower VR results
in a shorter takeoff distance. Experience has shown, however, that certain
limits on that selection must be met for safety reason; these are listed in
§25.107(e):
§25.107 Takeoff speeds

(e) VR, in terms of calibrated airspeed, must be selected in accordance with

the conditions of paragraphs (e)(1) through (4) of this section:
(1) VR may not be less than—
(i) V1;
(ii) 105 percent of VMC;
(iii) The speed (determined in accordance with §25.111(c)(2)) that
allows reaching V2 before reaching a height of 35 feet above
the takeoff surface; or
(iv) A speed that, if the airplane is rotated at its maximum
practicable rate, will result in a VLOF of not less than—
(A) 110 percent of VMU in the all-engines-operating condition,
and 105 percent of VMU determined at the
thrust-to-weight ratio corresponding to the
one-engine-inoperative condition; or
(B) If the VMU attitude is limited by the geometry of the
airplane (i.e., tail contact with the runway), 108 percent of
VMU in the all-engines-operating condition, and 104
percent of VMU determined at the thrust-to-weight ratio
corresponding to the one-engine-inoperative condition.
(2) For any given set of conditions (such as weight, configuration, and
temperature), a single value of VR, obtained in accordance with this
paragraph, must be used to show compliance with both the
one-engine-inoperative and the all-engines-operating takeoff
provisions.
692 Operational Aircraft Performance and Flight Test Practices

(3) It must be shown that the one-engine-inoperative takeoff distance,

using a rotation speed of 5 knots less than VR established in
accordance with paragraphs (e)(1) and (2) of this section, does not
exceed the corresponding one-engine-inoperative takeoff distance
using the established VR. The takeoff distances must be determined
in accordance with §25.113(a)(1).
(4) Reasonably expected variations in service from the established
takeoff procedures for the operation of the airplane (such as
over-rotation of the airplane and out-of-trim conditions) may not
result in unsafe flight characteristics or in marked increases in the
scheduled takeoff distances established in accordance with
§25.113(a).

We will provide a quick explanation of this regulation, because some material

impacting the takeoff run has yet to be covered (like the failure of one engine
with a continued takeoff).
§25.107(e)(1)(i) states that VR may not be less than V1 , where V1 is the
maximum brake on speed to reject a takeoff. We will cover this speed later
in this chapter and in Chapter 17.
§25.107(e)(1)(ii) states that VR may not be less than 5% faster than the
minimum control speed VMC of the airplane. Part 25 airplanes are always
multiengine, and should one engine fail at low speed, the manufacturer
must determine the speed at which the airplane can maintain a minimum
level of control. We reviewed the definition of minimum control speed
both in the air VMCA and on the ground VMCG in Chapter 9 and will
review testing for those speeds in the next chapter.
§25.107(e)(1)(iii) states that VR must be selected to ensure that the air-
plane, with the available acceleration due to its excess thrust and the rotation
procedure to be followed by the crew, will reach a minimum of V2 before
reaching 35 ft above the takeoff surface. This modeling will be part of our
takeoff performance estimates. We will describe V2 later in this chapter.
§25.107(e)(1)(iv) requires that the VR selected will result in a safe liftoff
speed (VLOF , more on VMU later). Paragraph (e)(2) requires the selection
of a single VR speed that is independent of whether an engine has failed or
all engines are operating because the crew does not have time to select a
rotation speed during this critical phase of the takeoff. Paragraphs (e)(3)
and (e)(4) state that the takeoff should be safe with reasonably expected devi-
ation in VR (crew rotates early).
As we can see, the “simple” VR speed is not so simple when looked at in
the perspective of certification requirements. Let’s continue the review of the
ground run modeling with this in mind.
Runway Gradient
The previous approach assumes a level runway (zero gradient).
Of course, runways generally are not at zero gradient or even of constant
gradient from one end to the other. A runway gradient off zero will
introduce a new force in the takeoff equation—a component of the
 CHAPTER 15 Takeoff Performance 693

22
Elev Elev Elev Elev

0%
.5
N

12
6581 ft 6565 ft 6421 ft 6368 ft

m
8%

4
13
.6
18
05

213 ft
6%
.6
18

FUEL
23
H 0% 9% 18.5% 12.5%
04
TWR 105 m 55 m
HA
NG
A R
238 m 137 m
535 m
1755 ft

Fig. 15.9 Courchevel runway gradient.

weight [2W sin(u)] will now act to increase or decrease the airplane’s accel-
eration.
1  
T  mr W  rSL s V 2 S CD,TO  mr CL,TO  W sinðuÞ
2
W dV
¼ ðV  VW Þ (15:14)
g dX
Note that we did not include cos(u) in the expression (mr W) because the
runway slope would rarely be more than 1–2%; therefore, cos(u)  1. We will
discuss why these are limiting in the OEI modeling section. There are,
however, some exceptions to the runway gradient maximum value. Courch-
evel altiport is one such example (see Fig. 15.9).
Using the simplified acceleration Eq. (15.11), we can see the relative
impact of the runway gradient.
 
T
ag  mr  sinðuÞ (15:15)
W

Minimum Unstick Speed (VMU )

The minimum unstick speed VMU of an airplane is defined by Part 25,
§25.107(d):
§25.107 (d) VMU is the calibrated airspeed at and above which the airplane
can safely lift off the ground, and continue the takeoff. VMU speeds must be
selected by the applicant throughout the range of thrust-to-weight ratios to
be certificated. These speeds may be established from free air data if these
data are verified by ground takeoff tests.

The manufacturer essentially needs to demonstrate that, should the airplane

lift off at its earliest capable speed, it will be capable of safely flying away from
the ground. This includes moving out of ground effects while flying at high
angle of attack near the ground.
The concept of VMU was first introduced into regulations with the
Special Civil Air Regulation No. SR-422A in 1958, to allow an aircraft to
lift off before reaching V2 , but at a speed at least “10 percent greater than
694 Operational Aircraft Performance and Flight Test Practices

a speed at which no hazardous characteristics are displayed by the airplane,

such as a relatively high drag condition or a ground stall.”
Two de Havilland Comet I crashes predating this regulation exemplify
the problem with early jetliners. In both accidents (G-ALYZ, Rome, 26
Oct. 1952, and CF-CUN, Karachi, 3 March 1953), the aircraft captain
adopted a nose-high pitch attitude too early in the takeoff run; the aircraft
failed to lift off the ground, and acceleration was reduced or stopped.
The manufacturer indicated that, for this early Comet design, the ideal
takeoff incidence was approximately 6 deg. The manufacturer also indicated
that, for the takeoff configuration selected, an increase in the takeoff inci-
dence past 9 deg would result in a partially stalled wing with corresponding
large drag increase. This condition could be noted by the pilot as a
low-frequency buffet.
We will cover VMU in the next chapter as part of takeoff testing assess-
ments. Suffice to say at this stage that VR must be selected by the applicant
to ensure that the airplane will lift off the ground at a safe margin above
VMU , as defined by §25.107(e)(1)(iv).

ICAO Circular 38-AN/33, No.37: B.O.A.C. de Havilland Comet I, regis-

tration G-ALYZ, accident at Ciampino Airport, Rome, Italy, 26 Oct. 1952.
The captain lined up the airplane on Runway 16 and applied thrust. When the
thrust was set, the brakes were released, and the airplane made a normal accel-
eration. At an IAS of 75–80 kt, the nose wheel lifted off from the runway and a
slight tendency to swing to starboard was corrected. At an IAS of 112 kt, the
captain lifted the aircraft from the ground by a positive backward movement
of the control column, and when he considered that the aircraft had reached a
safe height he called for “undercarriage up.” At about the same instant, the
port wing dropped rather violently and the aircraft swung to port; the controls
gave normal response, and lateral level was regained. At this point, the captain
realized that the aircraft’s speed was not building up, although he made no
reference to the ASI. A pronounced buffeting was felt, which he associated
with the onset of stall. In spite of two corrective movements of the control
column, the buffeting continued. Before the first officer had time to select
undercarriage up, the aircraft came down on its main landing wheels and
bounced. The airplane was approaching the end of the runway, so the
captain made the decision to abandon the takeoff. The undercarriage struck
a mound of earth as the captain was closing the throttle, the undercarriages
were wrenched off and the airplane slid 270 yd over rough ground.

6°
 CHAPTER 15 Takeoff Performance 695

Acceleration from VR to VLOF

Upon reaching VR , the pilot will apply a nose-up flight control input,
typically via a downward force on the aft horizontal tail, so the airplane
can adopt the flight attitude for a safe takeoff (see Fig. 15.10). The airplane
will lift off when the right combination of lift coefficient (angle of attack)
and airspeed is reached, per Fig. 15.4; while on the ground, the airplane
pitch attitude and angle of attack are essentially the same because the
flight path is nothing less than the runway slope.
To better visualize the multiple factors affecting this part of the takeoff,
we prepared Fig. 15.11. To lift off, the airplane must be brought to an
angle of attack that will generate enough lift to sustain flight. That lift will
be achieved at a combination of airspeed and wing loading for a given con-
figuration. The performance engineer must pick the correct rotation speed,
within the allowable range provided by the certification requirements, that
will optimize the takeoff distance while being easy enough for a crew of
average skill to follow so as to be predictable and repeatable.
The pitch rate is a driving factor for the time from rotation speed to
liftoff Dt. On traditional configurations (tail in the back), a download from
the tail is required to rotate the nose up, the airplane rotating about the
MLG contact point. In traditional tricycle landing gear configurations, the
center of gravity (CG) location is a big driver of the horizontal tail’s ability to
generate the required pitching moment, with more forward CG requiring
more tail download. The thrust line also has an impact. A high thrust line gen-
erates a nose-down pitching moment whereas a low thrust line generates a
nose-up pitching moment (see Fig. 15.12). It also means that the nose-up/
down contribution from the thrust line will be impacted when the airplane
goes from an AEO condition to an OEI condition. Finally, the tail proximity
to the ground during the rotation will also impact pitch rate capability. An air-
plane with a high tail will generally see less ground effects and be capable of a
more predictable pitch rate. An airplane with a low tail will see the tail come
close to the ground near liftoff (Fig. 15.13) where the downwash is reduced,
and it could also be exposed to the exhaust flow from the engine, further
impacting its aero capability and ultimately the rotation time from VR to VLOF .

D Rotation force
T required

f
Download to N
W
rotate

Fig. 15.10 Initiation of rotation.

 696
Operational Aircraft Performance and Flight Test Practices
Pitch attitude on ground 2
Lift coefficient required (CL1g)
1.8 for 1-g flight
n = CLa/CL1g
1.6
1.4

Lift coefficient
1.2
Flap shape and deflection 1
Basic section
Typical lift coefficient (CLa)
Slotted 0.8
coefficient, cl

3.0 Fowler during a takeoff run • Airspeed

Section lift

Section lift coefficient, cl

2.5 Plain flap

Split 0.6 • Pitch (, lift IGE)
2.0
Plain • Flaps
1.5 Slotted flap 0.4
1.0 Constant AoA on ground Rotation • Tail clearance
Basic 0.2
0.5 section Split flap •  stall IGE
0 0
–10 –5 0 5 10 15 20
Section angle of attack
Fowler flap
100 110 120 130 VR 140 VLOF 150 160 170
0, degrees
Airspeed
t, V
• Excess thrust
• Pitch rate

Fig. 15.11 Factors affecting speed spread between VR and VLOF .

 CHAPTER 15 Takeoff Performance 697

moment

moment
Pitching

Pitching
Thrust line below CG
Thrust line above CG Thrust line near CG

Fig. 15.12 Impact of thrust line on thrust pitching moment.

Horizontal tail away from Horizontal tail impacted by

engine exhaust flow and engine exhaust flow and
relatively far from ground close to the ground

Fig. 15.13 Ground effects and engine exhaust flow impact on horizontal tail.

From VR to liftoff, drag starts to increase proportional to the square of the

lift coefficient as the airplane initiates a rotation towards a target pitch angle,
but this increase typically has only a small impact on the excess thrust for
AEO conditions and the short duration of this part of the takeoff (see
Fig. 15.7 earlier in the chapter). Experience has shown that the airplane’s
rotation time from VR to VLOF can be represented as a function of excess
thrust of the airplane (see Fig. 15.14).

12

10

8 V
V (KTAS), t (s)

t
2

0
0 0.05 0.1 0.15 0.2 0.25
[(T – D)/W ] @ VR

Fig. 15.14 Speed spread model for the rotation phase from VR to VLOF .
698 Operational Aircraft Performance and Flight Test Practices

When the rotation time is plotted in this fashion, given the typical short
duration of this event, one can estimate the speed gained as a function of
excess thrust at VR multiplied by the rotation time, Eq. (15.16). This
becomes the speed spread between VR and VLOF .
 
(T  D)
DV  g Dt (15:16)
W VR

The distance covered by the airplane during that time is equal to the
average speed multiplied by the delta time.
1
DXRLOF  (VR þ VLOF ) Dt (15:17)
2
The generic speed spread of Fig. 15.14 is validated by flight testing.

Acceleration from VLOF to V35 , 35 Feet above Ground

The next phase of the takeoff involves a complex maneuver with the air-
plane both accelerating and climbing while the load factor increases to
change the flight path from one that is parallel to the runway to one that
will equal the excess thrust capability of the plane (see Fig. 15.15). The accel-
eration capability of the airplane is a representation of its excess thrust.
The combination of acceleration from VLOF to the airspeed at 35 ft (V35 )
plus the climb to 35 ft AGL can be represented as
(T  D) dh V dV
V ¼ þ (15:18)
W dt g dt
Airspeed is the change of distance over time, so Eq. (15.18) is rewritten as
(T  D) dh V dV
¼ þ (15:19)
W dX g dX
which leads to a distance equation equal to
 
W V
dX ¼ dh þ dV (15:20)
(T  D) g
The change of altitude dh is fixed and equal to 35 ft, per §25.113(a). This
leaves the speed increase to be optimized by the performance engineer. The
task is to select the speed increase dV in combination with the airplane’s
capability to rotate to the proper liftoff attitude and follow-on pitch to
climb away at V35 while minimizing the takeoff distance.
Figure 15.16 shows an example of the evolution of acceleration during an
AEO takeoff; note that for this takeoff, the thrust was not set to max prior to
brake release (not until about 500 ft). After thrust is set, observe how the
excess thrust decreases while accelerating on the ground (longitudinal accel-
eration) with all wheels on the ground. This acceleration decreases faster as
the airplane starts the rotation (at VR ); this reduction is further accentuated
 KTAS (kt) Wheel spd (kt)
160
Wheel spd (kt)

150
KTAS (kt)

158.9369
140 147.1808 123
130
134
120
Elevator (deg)
Elevator (deg)

10
–5.904212 –6.352623
0
–10
–20

Pitch angle (deg) Pitch rate (deg/s)

Pitch angle (deg)

30 6
3.625543 16.64738

Pitch rate
(deg/s)
20 4
6.874457
10 2
1.5
0 0

RADALT (ft)

CHAPTER 15
60
RADALT (ft)

34.9423
40
–0.5191575 Curved path
20
0

Nz (g)

Takeoff Performance
0.5
Nz (g)

–0.0545326
0.0 0.2550574

off 365 370 375 380 ft 385

Lift 425 Time (s)
35 264
2.9 1.5
36 3 8

Fig. 15.15 Complex flight path and acceleration from VLOF to V35 .

699
700 Operational Aircraft Performance and Flight Test Practices

160 5000 160 0.4

V35 V35
140 4500 140 0.35
VLOF VLOF

Longitudinal acceleration (g)

4000 VR
120 VR 120 0.3
3500
Airspeed (KTAS)

Airspeed (KTAS)
100 100 Ac 0.25

Distance (ft)
3000 ce
ler
ati
80 2500 80 on 0.2

ed
ed

pe
pe

rs
2000
rs

Ai
60 Ai 60 0.15
ce
stan 1500
40 Di 40 0.1

Thrust set
1000
20 500 20 0.05

0 0 0 0
0 5 10 15 20 25 30 35 0 500 1000 1500 2000 2500 3000 3500 4000 4500
Time (s) Distance (ft)

Fig. 15.16 Example of time, speed, and distance for an AEO takeoff.

after the airplane lifts off. In this example, the airplane reached 35 ft above
ground in a little more than 4100 ft of takeoff distance. The air distance
(VLOF to V35 ) could have been reduced had the pilot pitched up more to
have zero acceleration by the time the airplane got to V35 .
From observation of how fast the takeoff distance increases with increas-
ing airspeed in Fig. 15.16, one wants to minimize the V35 in order to mini-
mize takeoff distance. We have used V35 and V2 on several occasions in
the previous text to represent the speed at 35 ft. V2 is meant to be the
minimum speed that must be reached at 35 ft under OEI; under AEO with
the extra thrust, that speed will be higher.

Takeoff Safety Speed V2

As is implied by its name, the takeoff safety speed V2 is expected to provide
a minimum safety level when the airplane does reach the speed under what
is expected to be an OEI condition. The airspeed is thus well regulated! As it
is for any of the speeds of interest, the regulations state that the speed may
be selected by the applicant (airplane manufacturer), but must meet a
certain set of minimum requirements defined under §25.107(c).
§25.107(c) V2 , in terms of calibrated airspeed, must be selected by the appli-
cant to provide at least the gradient of climb required by §25.121(b) but may
not be less than—
(1) V2MIN ;
(2) VR plus the speed increment attained (in accordance with §25.111(c)(2))
before reaching a height of 35 feet above the takeoff surface; and
(3) A speed that provides the maneuvering capability specified in §25.143(h).

This statement has four minimum requirements to meet, and then 25.107(b)
adds the following:
§25.107 (b) V2MIN , in terms of calibrated airspeed, may not be less than—
(1) 1.13 VSR for—
(i) Two-engine and three-engine turbopropeller and reciprocating
engine powered airplanes; and
 CHAPTER 15 Takeoff Performance 701

(ii) Turbojet powered airplanes without provisions for obtaining a

significant reduction in the one-engine-inoperative power-on stall
speed;
(2) 1.08 VSR for—
(i) Turbopropeller and reciprocating engine powered airplanes with
more than three engines; and
(ii) Turbojet powered airplanes with provisions for obtaining a
significant reduction in the one-engine-inoperative power-on stall
speed; and
(3) 1.10 times VMC established under §25.149.

Based on the single VR , OEI or AEO, and §25.107(c)(2), one can expect that
the AEO case with the much greater thrust would result in a higher airspeed
by 35 ft than the OEI case. Similarly to the previous phase, one can generate a
speed spread model that will mimic the airplane’s capability during this
complex phase of the takeoff (see Fig. 15.17).
Note the highly nonlinear behavior of the time curve in Fig. 15.17. This
simply comes from the climb capability and is best visualized by rewriting
Eq. (15.18) as follows:
V
dh þ dV
g
dt ¼ (15:21)
V (T  D)=W
The minimum height to climb is 35 ft, so even while expecting reduced
acceleration (dV) during the maneuver, one can see that as the specific

14

12

dV
10
V (KTAS), t (s)

4
dt

0
0 0.05 0.1 0.15 0.2 0.25
(T – D)/W @ VLOF

Fig. 15.17 Speed spread model from VLOF to V35 .

702 Operational Aircraft Performance and Flight Test Practices

excess thrust [(T – D)/W ] reduces, the time to climb will become asympto-
tic. The exact shape of the curve will depend on the procedure defined by the
OEM for the airplane and used by the crew, §25.101(f):
§25.101 General
(f) Unless otherwise prescribed, in determining the accelerate-stop
distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must be
made in accordance with procedures established by the applicant for
operation in service.

The OEM must then specify the V-speeds (VR , V2 , and the AEO adder to V2 )
and the takeoff procedure that will ensure repeatable takeoff.

Target Pitch Rate

For a traditional Part 25 airplane (tail in the back, tricycle landing gear), a
target average pitch rate of 3 deg/s, from VR to a target pitch attitude, is
typical and it is a comfortable rate for passengers. Most jetliners are
capable of higher rates, however, so the manufacturer must clearly define
what is expected of the pilot, and proper training must be provided to
ensure repeatability of the maneuver. In the end, the pitch rate must be suffi-
cient that the airplane will achieve the specified V-speeds in the expected
time so that the takeoff distance becomes predictable and the airplane flies
safely away from the ground in AEO and OEI conditions.
The following are extracts of procedures from various aircraft flight
manuals (AFMs) of Part 25 jets found online. Some OEMs provide more
details than others.

Airplane 1: At VR, promptly and smoothly apply and hold

approximately 2/3 aft sidestick to achieve a rotation rate of
approximately 2 to 3 deg/sec, assessed primarily by outside visual
reference. The rotation rate may take time to establish but for a
given stick input, once it has developed it remains relatively constant.
The rotation rate is important as too low a rate would compromise
take-off performance, whereas too high a rate would increase the risk
of tailstrike. An indication of the correct rotation rate is achieving
the target pitch attitude approximately 5 seconds after rotation
commences (not 5 seconds after sidestick input).
Airplane 2: With all engines operating, or if an engine failure occurs at or
after V1 speed, rotation of the airplane is initiated at VR . Recommended
nose-up attitude is set. Gear retraction is initiated less than 3 seconds after
lift-off with all engines operating, and less then 4.5 seconds after lift-off
with one engine inoperative.
Airplane 3: Takeoff speeds are established based on minimum control
speed, stall speed, and tail clearance margins. Shorter-bodied airplanes are
normally governed by stall speed margin while longer-bodied airplanes
are normally limited by tail clearance margin. When a smooth continuous
rotation is initiated at VR, tail clearance margin is assured because
computed takeoff speeds are developed to provide adequate tail clearance.
Above 80 knots, relax the forward control column pressure to the neutral
 CHAPTER 15 Takeoff Performance 703

position. For optimum takeoff and initial climb performance, initiate a

smooth continuous rotation at VR toward 158 of pitch attitude. The use of
stabilizer trim during rotation is not recommended. After liftoff, use the
attitude indicator as the primary pitch reference. Using the technique
above, liftoff attitude is achieved in approximately 4 seconds. Resultant
rotation rates vary from 2 to 2.5 degrees/second with rates being lowest
on longer airplanes.
Airplane 4: At VR rotate to the required pitch attitude to maintain an
airspeed at or above V2 . When a positive vertical speed is confirmed, select
landing gear up.
Airplane 5: At VR , rotate the airplane to 148 (for flaps 98). With positive rate
of climb, landing gear up and minimum airspeed V2 . If maneuvering is
required, maintain a minimum airspeed of V2 þ 10 KIAS with a
maximum bank of 258.
Airplane 6: At VR , the Pilot Flying should rotate smoothly toward the
target pitch attitude in one continuous motion. Use a rotation rate not to
exceed 3 degrees per second. High weights and temperatures or engine
failure will require slightly lower rotation rates. Correct rotation
technique is important to ensure that adequate performance is obtained.
Lift-off will occur prior to reaching initial climb target attitude.

The only requirement for the takeoff procedure published in the AFM is
that it can be executed repeatably and precisely by a crew of average skill
and that it results in the correct speeds being achieved by 35 ft above
ground within the prescribed takeoff distance, with or without an engine
failure. The OEM is free to prescribe the procedure it believes achieves
this goal.

Artificial Runway Lengths Avert Long A340 Takeoff Rolls

Air France and Lufthansa introduced artificial reductions of runway length at
Bogota as a precautionary measure after incidents involving prolonged
take-off runs by Airbus A340s.
French investigators have disclosed the carriers’ actions in an analysis of a
serious departure incident from the Colombian capital in March 2017. The
A340-300 started rotating at 142kt some 2,760m from the threshold of
runway 13R, which is 3,800 m long. But it did not lift off for another 11s
at which point it was only 140m from the threshold of the opposite-direction
runway 31L, crossing it at a height of just 6ft. The take-off distance was
nearly 990m longer than that defined by performance models.
Lufthansa had previously looked into rotation issues after a take-off inci-
dent in Johannesburg in 2004 and long take-off runs in 2007, 2011 and 2012.
Its studies showed an average continuous rotation rate of 1.98/s rather than
the 3.18/s required by the A340-300 performance model.
But BEA says that Lufthansa has resisted specific training to adapt take-off
techniques, in order to avoid a possible increase in tail-strike risk. Air France’s
flight-crew techniques manual had specifically cautioned against low rotation
rate. It pointed out that rates of 2-38/s would have “minimal impact” on the
704 Operational Aircraft Performance and Flight Test Practices

take-off run, but rates “significantly” below 28/s “should be avoided”. Airbus
last year [2018] revised its manual for the type, quantifying the effect of a
2-38/s variability in rotation rate as translating into a 300m extension of the
take-off run.

Source: https://www.flightglobal.com/news/articles/artificial-runway-lengths-
avert-long-a340-take-off-r-459634/

Target Pitch Attitude

To achieve the correct airspeed at 35 ft and maintain it during the first
part of the climb, the aircraft must maintain the proper flight path angle.
But the flight path angle is not what is provided to the crew as a reference
for the takeoff procedure because it depends on too many variables (includ-
ing the airspeed, rotation rate, and load factor) for a crew to reliably achieve
in a dynamic maneuver like a takeoff (see Fig. 15.18).
From the crew’s perspective, a pitch attitude is more readily understand-
able and is predictably away based on a pitch rate capability, so the flight path
angle must be translated into a pitch angle that the pilot can aim for during
the rotation. As we have seen in previous chapters, the pitch angle is the sum
of the airplane angle of attack and flight path angle (proportional to excess
thrust), and the angle of attack depends on the airspeed, airplane configur-
ation, and load factor. The expected load factor during the rotation to the
target pitch attitude will be of the order of 1.02 to 1.2 g, the larger value
going with a higher pitch rate. Figure 15.18 shows how two takeoffs with
similar liftoff speeds, one AEO and one OEI, can have different rotation
speeds and V35 .
From Fig. 15.18, one can see that for the OEI case, the airplane stabilized
on a target pitch angle of about 8 deg, which was achieved prior to reaching
35 ft; in contrast, for the AEO case, the pitch attitude continued to increase
beyond the 35-ft height. This is expected because the procedure followed by
the crew in terms of pitch rate should not be dependent on engine condition,
but one can adjust the target pitch attitude to be followed based on such a
condition. In selecting a target pitch attitude, the OEM must remember
that a target pitch attitude that is too large will lead to a deficit in specific
excess thrust and a decrease in airspeed during the maneuver, and the aircraft
will not reach the target airspeed by the time it reaches 35 ft altitude.
The contrary is true as well—too low of a target pitch attitude, and the
aircraft will climb more slowly and exceed the target airspeed (see
Fig. 15.19). The takeoff distance is a function of when the aircraft reaches
35 ft altitude, so the distance will be longer than the minimum required by
regulation.
On several aircraft, a target pitch for AEO and OEI can be a single value
for each condition that would cater to all possible T/W variations expected in
 AEO takeoff OEI takeoff
Wheel speed (kt) RADALT (ft) KCAS (kt) Wheel speed (kt) RADALT (ft) KCAS (kt)
200 800
137.6939 147.9997 150.28 140 142.5 120
700
120 146.2281 135.125 100
150 600
Wheel speed (kt)

Wheel speed (kt)

100 133.1143

RADALT (ft)
500 80
KCAS (kt)

RADALT (ft)
130.977

KCAS (kt)
80 35.42173
100 133.8385 127.5799 400 60
122.6384 300 60
35.52257 40 40
50 200 1.039502 1.414239 1.125
1.23036 0.301041
100 20 20

0 0 0 0
Pitch (deg) FPA (deg) Elevator (deg) Pitch (deg) FPA (deg) Elevator (deg)
–2.037226 10.75375
12 0
6.968781

Flight path angle (deg)

15 0 10 –2.65301 1.684665
–2
–5.175261 –2 8

Elevator (deg)
Elevator (deg)
10 3.146365 –6.598316 –4
Pitch (deg)

Pitch (deg)
FPA (deg)

–4 6 0.2204686
–0.1460393

CHAPTER 15
5 –6 4 –0.07975803 –6
4.917605 –8 2 –8
0
–3.170477 –10 0
–5 –2 –10
–0.9671298 –12 1.424955
0.610302 –0.7794436 –9.567954
–4 –12
35 40 45 50 55 55 60 65 70 75 80
Time (s) Time (s)

03 4
.1 off

67 e
6

85
62 Vr

.0 lin
87 5

31
64 Lift

Takeoff Performance
.2 off

.7 ne

82

70 nt
29

14
Vr

.4
45 t li
42 Lift

e
84

40

Ev
en
.4

Ev
40

Fig. 15.18 Dynamics between VR and V35 .

705
706 Operational Aircraft Performance and Flight Test Practices

L T V OEI TO
 V2  too low
V  Vlof Correct 

VR  too high
D
W
Time

Fig. 15.19 Proper pitch capture helps with achieving target speed by 35 ft.

service to ensure the aircraft can reach at least V2 (or V2 þ XX for AEO) by
35 ft. Single values mean that in most cases, the aircraft will accelerate to
higher than the target speed (preventing a deceleration), resulting in a
larger takeoff distance than the minimum required. The target pitch values
can be adjusted by crew procedure. Here are a few examples extracted
from AFMs found online.
Airplane 1: As the rotation progresses and the runway environment
disappears from view, transfer attention to the PFD [Primary Flight
Display] to establish the initial pitch attitude of 158. With all engines
operating, the airplane has considerable excess thrust. It is normal for a
correctly flown rotation to result in a stabilized speed in excess of
V2 þ10 kt. With all engines operating, the performance of the airplane is
not compromised by the additional speed.
Airplane 2: At VR , rotates smoothly towards the target pitch attitude of the
Flight Director (FD) in one continuous motion. Adjust pitch to achieve an
airspeed of V2 þ10-15 KIAS for all engines operating.
Airplane 3: The Flight Director pitch bar indicates a pitch of 88 up. The
guidance becomes active after liftoff and the command bar moves to a
pitch attitude to maintain a target speed of V2 for an engine failure case
and a speed of V2 þ15 knots for all engines operating.
Airplane 4: When VR is attained the airplane is rotated to 128 pitch attitude
until 35 ft or until reaching V2 . Pitch angle adjusted as required to
maintain not less than V2 . Landing gear retracted when a positive rate
is achieved.
Airplane 5: OEI Procedure: Following recognition of engine failure at or
after V1 , accelerate to VR . Rotate at 38 to 58 per second to the target
takeoff pitch attitude while accelerating to achieve V2 at a height of 35 feet
above the runway. Maintain V2 once achieved, Small pitch attitude
corrections may be required to maintain V2 .
Airplane 6: If an engine failure occurs after V1 but prior to VR, continue
acceleration to a scheduled VR, at which rotate the airplane with an
average 3-4 degrees per second to an initial pitch attitude of 9-11
degrees. Adjust the airplane’s pitch attitude to maintain V2. During a
normal takeoff (all engines operating), rotate the airplane at VR with an
average of 3-4 degrees per second to an initial pitch attitude of 13-15
degrees. Adjust pitch attitude to obtain v2þ10 knots but limit to
258 maximum

Some OEM may instead elect to provide more guidance in the AFM in the
form of a table or chart that contains target pitch attitudes (AEO and OEI)
 CHAPTER 15 Takeoff Performance 707

as a function of weight to better optimize the takeoff distance (see Fig. 15.20).
This table must provide sufficient information to the crew to account for the
excess thrust condition related to the combination of flap settings, airplane
weight, altitude, and temperature at the time of takeoff, and crew procedure
if it calls for reduced thrust setting.
Other aircraft are equipped with a smart pitch director that accounts in
some form for the aircraft T/W and provides an optimum pitch attitude
for takeoff that will just bring the aircraft to the target airspeed by 35 ft.
This kind of director greatly simplifies the pilot’s speed capture task. These
pitch directors also recognize engine failure (usually through a split in N1)
and automatically reduce the target pitch to the OEI value. In the previous
takeoff procedure examples, it can be seen that some airplanes use such
an approach.

Trim Setting
Another item that greatly helps reduce the crew workload and make a
takeoff more repeatable from a distance point of view is the proper pitch
trim setting. The ideal pitch trim setting will naturally bring the airplane to
the proper pitch attitude and airspeed combination during the takeoff.
Based on this author’s experience, a good pitch trim setting would bring
the airplane to a trimmed condition (zero pilot pitch force), landing gear
up at V35 AEO. This means that for most airplanes, a slight pull force
would still be required for landing gear down and V2 . The actual value of
the ideal pitch trim setting depends on the airplane thrust line (high vs low
thrust) and thrust, CG location, and flap configuration. Most manufacturers
will offer a pitch trim setting for CG and flap. Traditional Part 25 transport
category airplanes will need more nose-up pitch trim with more forward
CG and larger flap deflection (Fig. 15.21).

Modeling the Ground Run: OEI

FAA Part 25 transport category airplanes must be able to safely continue
a takeoff following the failure of the most critical engine if the airplane is
beyond a certain airspeed (V1 ) during the ground run. That speed can be

Temperature (deg C)
Altitude Weight
Target OEI Pitch table
(ft) (lb)
Note: For AEO add 4 deg to the OEI target pitch.

Fig. 15.20 Example of table to optimize pitch target based on available T/W.
708 Operational Aircraft Performance and Flight Test Practices

–10

–9

–8
Airplane nose up (units)

–7 Increasing
Tak flap
–6 e off
3 deflection
Tak
–5 e off
2
–4 Tak
eo ff 1
–3 Flaps
up
–2

–1

0
0 10 20 30 40 50
Center of gravity (% MAC)

Fig. 15.21 Recommended takeoff trim settings.

selected by the applicant (airplane manufacturer) but will have limitations

imposed on it that we will cover in this section.
The main impact on the distance will now be the rapid loss of thrust vs
time after the failure that now introduces a new dynamic to the computation
of the takeoff distance (see Fig. 15.22). It can be observed that at about the
same time the engine fails, as seen in this example by a very rapid reduction
in fan speed (N1) at time 61.5 s, the pilot must address the rapid heading
deviation with the application of rudder. The airplane soon reaches rotation
speed (time 62 s), at which point the pilot must now increase the pitch angle
to continue the takeoff as soon as the MLG leaves the ground (time 65 s). A
bank angle develops due to the rolling moment generated by the deflected
rudder (force above CG) and the loss of MLG ground reaction force. The
pilot must now also address this by deflecting ailerons to maintain wings
as level as possible while close to the ground. The pilot flies the airplane
away from the ground in this configuration, only retracting the landing
gear when climb is assured.
Transport category airplanes are designed for continued safe takeoff fol-
lowing an engine failure, so the airplane’s manufacturer must choose con-
ditions that will ensure such a safe takeoff and demonstrate that the
airplane will have acceptable handling characteristics. The aviation industry,
manufacturers, and authorities have defined additional airspeeds over the
years to come up with a standardized airspeed to be used by aircrew to maxi-
mize the chance of a successful continued takeoff.
 KCAS (kt) RADALT (ft) KCAS (kt) RADALT (ft)
150 150
300 300

RADALT (ft)

RADALT (ft)
KCAS KCAS
KCAS (kt)

KCAS (kt)
100 100

failure
RADALT 200 RADALT 200

failure

Eng
50

Eng
50 100 100
0 0 0 0

Pitch (deg) Flight path angle (deg) Heading (deg)

Flight path angle (deg)

39

Heading (deg)
10 Pitch
Pitch (deg)

38
5 37
Flight path
36
0
35

N1-L (%) % max FFL (%) Rudder (% max)

1.2

Rudder (% max)
% max FFL (%)

1.0 0.6
N1-L (%)

0.8
0.4

CHAPTER 15
0.6 Rudder
Rudderforce
force
0.4 0.2
0.2 N1
0.0 0.0

Elevator (% max) Roll (deg)

0.0 2

Roll (deg)
Elevator
(% max)

–0.2 0

Takeoff Performance
–0.4
–0.6 –2
–0.8 –4
40 50 60 70 80 90 100 110 40 50 60 70 80 90 100 110
Time (s) Time (s)

Fig. 15.22 Dynamics of a continued takeoff following an engine failure.

709
710 Operational Aircraft Performance and Flight Test Practices

Critical Engine Failure Speed VEF

The critical engine failure speed is defined by FAA Part 25 §25.107(a)(1)
as follows:
§25.107(a)(1) VEF is the calibrated airspeed at which the critical engine is
assumed to fail. VEF must be selected by the applicant, but may not be less
than VMCG determined under §25.149(e).

VMCG was described in Chapter 9; it is the minimum control speed on the

ground. We will also review testing for VMCG in more detail in the following
chapters.
Note that the regulation does not specify which engine is critical for
an airplane. It is the applicant’s responsibility to define the most
adverse failure condition and which engine produces that condition (see
Fig. 15.23). The airplane may have a traditional engine distribution with
all engines having the same thrust capability, or it may have an
unusual distribution—not all engines may provide the same thrust level.
(As an example, on a four-engine airplane, one engine may not have
any bleed extraction but the other three do, so that one engine may
produce slightly more thrust.) As well as looking at the thrust loss and
impact on the drag of the airplane, one must also look at the impact
on the airplane’s systems when one engine fails. For example, on an air-
plane with hydraulically powered flight controls, the loss of power from
one engine (loss of one hydraulic pump) may not impact the flight con-
trols in a symmetric way, thus the impact on the controllability of the air-
plane following an engine failure must be analyzed with respect to the
dynamic nature of a continued takeoff following the failure of the critical
engine at VEF .
In the end, as stated previously, it is the applicant’s responsibility to ident-
ify the critical engine and the critical engine failure speed VEF . That speed is
not a speed provided to the flight crew, but one that is used to define other
performance takeoff speeds.

Maximum Brake on Speed V1

Another very important speed for an engine failure event is the maximum
brake on speed V1 , defined by FAA Part 25, §25.107(a)(2):
25.107(a) V1 must be established in relation to VEF as follows:

(2) V1 , in terms of calibrated airspeed, is selected by the applicant; however, V1

may not be less than VEF plus the speed gained with critical engine inop-
erative during the time interval between the instant at which the critical
engine is failed, and the instant at which the pilot recognizes and reacts
to the engine failure, as indicated by the pilot’s initiation of the first
action (e.g., applying brakes, reducing thrust, deploying speed brakes) to
stop the airplane during accelerate-stop tests.
 Rolling moment

m
Sideslip fro
m et ri c th rust
asym

Outboard critical engine failure

CHAPTER 15
g engine
- Loss of lift outboard indmillin Inboard critical engine failure
Wake of w
- Rolling moment generated - Asymmetric flow over tail
- Loss of thrust outboard - Directional control
- Yawing moment generated - Pitch control

Takeoff Performance
Fig. 15.23 Selecting the critical engine.

711
712 Operational Aircraft Performance and Flight Test Practices

Per this description, the main use of the V1 speed is to define the
maximum speed at which the pilot must have initiated a rejected takeoff.
The analysis of a rejected takeoff is part of the discussion in future
chapters. The meaning of V1 in the context of continued takeoff is that
one must assume that if the pilot has not rejected that takeoff by the time
the airplane has reached V1 , then the pilot has elected to continue the
takeoff. We then analyze the takeoff performance of the plane based on
that assumption.
Prior to Amendment 42 (1978), V1 was regarded as a decision speed
where the pilot could decide to continue the takeoff or elect to stop the air-
plane. The wording for V1 at the time was:

§25.107(a) V1 must be selected by the applicant and must be at least the

minimum calibrated airspeed at which controllability by primary aerody-
namic controls alone is shown (during the takeoff run) to be adequate to
safely continue the takeoff, using normal piloting skill, when the critical
engine is suddenly made inoperative. [amendment 25-38, 1977]

This version did not clearly indicate that takeoff distance was not guaranteed
if a takeoff was rejected beyond that point V1 , so a clarification was made
soon after in Amendment 42. This is also the amendment level where VEF
was introduced (underline added by author):

§25.107(a)(2) V1 , in terms of calibrated airspeed, is the takeoff decision speed

selected by the applicant; however, V1 may not be less than VEF plus the
speed gained with the critical engine inoperative during the time interval
between the instant at which the critical engine is failed, and the instant at
which the pilot recognizes and reacts to the engine failure, as indicated by
the pilot’s application of the first retarding means during accelerate-stop
tests. [amendment 25-42, 1978]

NTSB/SIR-90/02: In July 1988, the crew of an Air France Boeing 747

initiated an RTO after a fire-warning horn for the no. 4 engine sounded at
V1 or an airspeed slightly higher than V1 during a takeoff from Delhi, India.
The airplane was destroyed when it struck a ditch beyond the end of the
runway; one passenger sustained minor injuries. NTSB root cause: The
crew’s reduction of power occurred as the airplane reached 167 kt; V1 was
156 kt. Author’s note: The airplane is designed for continued takeoff after an
engine failure, or in this case, a perceived engine failure.

In Amendment 92 (1998), the FAA removed the phrase decision speed from
the V1 regulation, so the words would now be focused on the execution of the
RTO rather than the crew reaction at or beyond V1 . We will discuss V1 more
in Chapters 17 and 18. For now, we will discuss the continued takeoff follow-
ing the failure of the critical engine at VEF .
 CHAPTER 15 Takeoff Performance 713

Modeling VEF to V35

The distance between VEF , the start of the rotation VR , the liftoff point
VLOF , and the point at which the aircraft reaches 35 ft above the runway
surface [per §25.113(a)] is a complex mix of T/W ratio (acceleration capa-
bility), pitch rate capability and pitch attitude vs time, general handling
with the developing asymmetric thrust, and target speed to be reached by
35 ft V35 (see Fig. 15.24). It is highly dependent on the AFM procedure
defined by the applicant [§25.101(f)] and how the crew executes that pro-
cedure. To model this dynamic segment, we are going to break it down
into a series of timed events, with Fig. 15.25 showing some expected beha-
viors to be addressed.

Events at VEF
The event that has the greatest impact on the distance for a continued
takeoff is the loss of one engine. Transport category airplanes will be demon-
strated to be safe following the loss of thrust on the most critical engine, but
the critical engine from a handling point of view may not always be the same
for the computation of takeoff distance; the applicant must determine which
engine will be more critical for continued takeoff performance.
Takeoff distance is based on reaching specified airspeeds, so the condition
that is most critical for performance will be the one that leads to the largest
loss of thrust in the most rapid way, typically a fuel starvation condition.
Under these conditions, the engine rapidly goes from producing forward
thrust to producing windmilling drag (aerodynamic forces making a gas
turbine or propeller turn; it is a retarding force). For most engines used for
Part 25 airplanes, the transition from one state to the next will occur over
the course of about 1 s. The model for the OEI condition will be a thrust
vs time model with 100% available thrust (based on current airspeed, altitude,
temperature, and thrust setting) just prior to VEF down to windmilling drag
in a time that is representative of the engine characteristics. The model can
be as simple as a linear variation with time such as is shown in Fig. 15.26, or it
can be more representative of the thrust decay vs time. The ultimate goal of
the model is to be able to best represent the airplane acceleration (rep-
resented by Nx in Fig. 15.26) vs time from engine failure to the end of the
takeoff run.

V35

35 ft
VEF V1 VR VLOF

Fig. 15.24 Sequence of events from VEF to 35 ft AGL.

 714
Operational Aircraft Performance and Flight Test Practices
KCAS (kt) N1-L (% max)
VEF VR VLOF
150 1.2
1.0

N1-L (% max)
100 130.5152 131.3089 133.206 0.8
KCAS

0.6
(kt)

128.8975 0.281911
50 0.4
0.9958212 0.5704172 0.2
0.6929667
0 0.0

Heading (deg) Rudder (% max)

40 37.30062
1.0

Rudder (% max)
36.86162 36.49105
38
36.7 0.8
Heading

0.6
(deg)

36 0.05126953 0.4
0.454834 0.2
34 0.524251
0.1960659 0.0
32 –0.2

Roll (deg) Aileron (% max)

2 0.1

Aileron (% max)
0
0.0006088429 –0.2 –0.06838368
1 0.0
Roll (deg)

0 –0.1
–1
–2 –0.01055416 –0.2
0.0006175613 –0.3
–3 –0.5940354
–4 –0.4

Elevator (% max) Pitch (deg)

0.0 10

Pitch (deg)
4.776142
Elevator
(% max)

–0.2 5
–0.2091316
–0.4 –1
–0.177012 –0.1770149
–0.6 0
–0.861615 –0.7 –0.6742876
–0.8 –5
60 61 63 64 65 66 67
38 1

3 2

95 3

04 4
99 e

38 e

08 e

94 e
.6 lin

.2 lin

.6 lin

.9 lin
Time (s)
Ev 8
61 nt

62 nt

62 t

64 nt
en
e

e
Ev

Ev

Ev
Fig. 15.25 OEI event details.
 CHAPTER 15 Takeoff Performance 715

Event Model
N1-L (%) KTGS (kt)
140 1.2
0.9958212 136.8843
138 1.0 Thrust
136

N1-L (% max)
0.7065704 factor
KTGS (kt)

0.8
134
135.6103 0.6
132

Forward thrust
130 0.4
128 0.2
126
0.0
Nx (g)
0.20
0.1689413
0.15
Nx (g)

0.06992935 ≈1s
0.10

Windmilling
0.05 Time
from VEF

drag
0.00
58 59 60 61 62 g 63 64
V EF 41 llin
Time (s) .69 mi 2
61 Wind 2037
.
62

Fig. 15.26 Making the loss of thrust model representative.

Propeller-driven airplanes that suffer an engine failure must contend with

much larger windmilling drag than turbojet and turbofan airplanes unless
they are equipped with a propeller feathering system (which most are) that
reduces the drag of the inoperative system to a much lower value. Even
with such a system, the transition time from windmilling to feathered propel-
ler can be relatively long compared to the continued takeoff event.
Figure 15.27 shows the typical time sequence from engine failure to wind-
milling drag Dtw and then a feathered condition Dtf that can be implemented
in the takeoff model.

Thrust
factor
Forward thrust

Coarse Fine
(high speed) (takeoff low speed)

Reverse
(after landing)

tw tf
Feathered
Time
(power failure)
from VEF
Windmilling
drag

Fig. 15.27 Windmilling and feathered propeller model.

716 Operational Aircraft Performance and Flight Test Practices

It should be noted that credit for feathering may be used only if it is

accomplished by an automatic system; if pilot action is required, then
no credit may be used for the purpose of computing takeoff distance per
§25.111.
§25.111 Takeoff Path

(c) During the takeoff path determination in accordance with paragraphs

(a) and (b) of this section—
(4) The airplane configuration may not be changed, except for gear
retraction and automatic propeller feathering, and no change in
power or thrust that requires action by the pilot may be made until
the airplane is 400 feet above the takeoff surface;

The next item that must be modeled is the pilot reaction to the asymmetric
thrust. As seen earlier in Fig. 15.25, the control surface deflection will tend to
occur in sequence as heading and then roll angle start changing. A sideslip
angle will also occur as the airplane lifts off the ground and the heading is
maintained to that of the runway. The impact of the deflection of flight con-
trols to these deviations is typically modeled as an increase in airplane drag
due to the asymmetry. Such effect was discussed in Chapter 13 under the
control drag CDctl that was proportional to the yawing moment Cna pro-
duced by the asymmetric thrust. One now needs to decide how this drag is
introduced in the takeoff model to best represent the speed change vs
time. The performance engineer can decide to introduce a ramp up of the
drag after a given pilot reaction time from engine failure (see Fig. 15.28),
or the drag can simply be applied to full value at the time of engine failure
(i.e., step input). The only requirement is to best match the speed change
and resulting distance.
Many systems depend on power generated by the engines, so the loss of
an engine may impact takeoff-critical systems such as flight control, includ-
ing the possibility of spoilers upfloat (partially deployed). Any impact on the
airplane’s ability to take off will also need to be modeled.

CDctl % CDctl
100
Fn

yeng

yeng

DW
Cna Crew Time or V
Ramp up
reaction from VEF

Fig. 15.28 Modeling control drag for OEI event.

 CHAPTER 15 Takeoff Performance 717

Landing gear drag represents an important part of the retarding force

during a takeoff. Retractable landing gears can take 5 to 15 s to fully
retract, which is usually much longer than the continued takeoff airborne
segment unless the airplane is at minimum excess thrust condition and the
takeoff distance is dictated by first segment climb. (As seen in Chapter 13,
for a twin-engine airplane, the minimum climb gradient out-of-ground
effects allowed by regulation is zero. The presence of ground effects may
help the airplane just enough to get off the ground and allow the pilot to
select gear up to be able to fly away from ground effects once the gear is
retracted.) To help with the continued takeoff, the landing gear must be
selected up as soon as the pilot sees a positive rate of climb. The FAA codified
this last requirement in §25.111(b) at initial amendment, 1965:
§25.111(b) During the acceleration to speed V2 , the nose gear may be raised
off the ground at a speed not less than VR . However, landing gear retraction
may not be begun until the airplane is airborne.

As with most regulations, this one was generated based on the industry and
agency experience in operation with such events as the 8 April 1963 nose gear
collapse of a Douglas DC-7 on takeoff at Rio de Janeiro-Galeão International
Airport.

The crew was engaged in a local training flight at Rio de Janeiro-Galeão Airport.
During the takeoff roll, just after V1 speed, the nose gear retracted. The aircraft
sank on its belly, lost engines number two and three, and slid for several
yards before coming to rest in flames. All seven crew members were slightly
injured, but the aircraft was damaged beyond repair. The probable cause was
poor flight preparation and lack of crew coordination, which led the crew to
retract the undercarriage prematurely before the aircraft had taken off.
Source: https://www.baaa-acro.com/aircraft/douglas-dc-7?page=4

Modern airplanes have some protection systems that will prevent a

landing gear retraction while weight on wheels (WOW) is detected. Some
pilots have come to rely on this function and select gear up while still on
the ground. This has led to some unfortunate accidents when the airplanes
encountered some bumps on the runway near liftoff that were just strong
718 Operational Aircraft Performance and Flight Test Practices

enough for the WOW switch to detect weight off wheels and allow the gear to
start retracting while insufficient lift was being produced to ensure
continued flight.

On 13 April 2018 at approximately 1745 Zulu (Z), 1045 Local (L) time, an
F-22A Raptor, T/N 07-4146, assigned to the 90th Fighter Squadron, 3rd
Wing, Joint Base Elmendorf-Richardson, Alaska, took off from runway 31
Left (31L) at Naval Air Station (NAS) Fallon, Nevada. The mishap pilot
(MP) initiated a military power (MIL) takeoff and rotated at 120 knots cali-
brated airspeed (KCAS). As the MP recognized his visual cues for the
mishap aircraft (MA) becoming airborne, he raised the landing gear handle
(LGH) to retract the landing gear (LG). Immediately after main landing
gear (MLG) retraction, the MA settled back on the runway with the MLG
doors fully closed and the nose landing gear (NLG) doors in transit. The
MA impacted the runway on its underside and slid approximately 6514 feet
(ft) until it came to rest 9,419 ft from the runway threshold. Once the MA
came to a complete stop, the MP safely egressed the aircraft. There were no
injuries, fatalities, or damage to civilian property.
The Accident Investigation Board (AIB) President found by a preponder-
ance of the evidence that the causes of the mishap were two procedural errors
by the MP. First, the MP had incorrect Takeoff and Landing Data (TOLD) for
the conditions at NAS Fallon on the day of the mishap, and more importantly,
he failed to apply any corrections to the incorrect TOLD. Second, the MP pre-
maturely retracted the LG at an airspeed that was insufficient for the MA to
maintain flight.
Source: https://www.kunsan.af.mil/Portals/6/180413-PACAF-JB%20Elmen
dorf-Richardson-Alaska-AIB%20NARRATIVE%20REPORT.pdf?ver=2018-11-
15-200849-187

Some airplanes are equipped with an automatic power reserve (APR)

system or an automatic takeoff thrust control system (ATTCS). These
systems are designed to detect the loss of power from one engine and auto-
matically increase the power on the remaining engine(s) by a few percent
(5 to 10%, typically) without pilot input. That system must comply with
the requirements of §25.904.

§25.904 Automatic takeoff thrust control system (ATTCS).

Each applicant seeking approval for installation of an engine power control
system that automatically resets the power or thrust on the operating
engine(s) when any engine fails during the takeoff must comply with the
requirements of appendix I of this part.

Figure 15.29 depicts a typical APR application. During the takeoff accelera-
tion, one engine fails, as seen by a rapid thrust loss. The system must
detect the failure and command the remaining engine to spool up to the
desired APR level. During this dynamic thrust variation timeframe, the
 CHAPTER 15 Takeoff Performance 719

Failure
APR
Thrust
Spool up Engine 2
Detect

Engine 1

Time
Windmill drag

Fig. 15.29 APR activation example.

handling of the airplane is carefully evaluated because it can have an impact

on the various minimum control speeds. From a performance point of view,
one needs to define the typical time intervals and thrust variation with time to
be able to reliably predict acceleration and distances.

Takeoff Speed Summary

As was seen in the previous discussion, the applicant is free to select the
speeds it believes will be best for the airplane being certified as long as it
respects all the regulatory constraints specified. To best visualize those
constraints on the accel-go part of the takeoff, this author developed
Fig. 15.30 for his class at Kansas University Continuing Education. It has

Selected by
applicant
Critical engine
Chapter 17 VEFmin VEF
failure speed
“Decision speed”
Chapter 17 V1min V1 High speed for start
of braking – RTO

≥1.05 VMC 25.107(e)(1)(ii) VRmin VR Rotation speed

≥1.10 VMU, AEO 25.107(e)(1)(iv) Liftoff speed

≥1.05VMU, OEI VLOFmin VLOF

25.107(b)(1) Takeoff
≥1.13 VSR V2min V2 safety
25.107(b)(3)
≥1.10 VMCA speed
25.107(c)
25.107(c)(3) ≤ 35 ft AGL
25.143(h) Maneuvering capability 25.107(c)(2) 25.121(b)
Climb gradient

Fig. 15.30 Takeoff speed regulations.

720 Operational Aircraft Performance and Flight Test Practices

proven very useful for students to more easily grasp the constraints on the
takeoff speeds.

Takeoff Distance, Acceleration-Go

To compute the takeoff (accel-go) distance, one must first establish the
takeoff speeds for a given weight, altitude, temperature, and configuration (see
Fig. 15.31). The following represents the process this author typically uses.
1. Select the minimum V2 .
a. Check that VSR  1.13 and verify maneuvering capability.
b. Check that VMCA  1.10.
c. Compute the specific excess thrust [(T – D)/W ] and check climb
gradients for all segments of takeoff.
2. Compute VR .
a. From the step 1 V2 and excess thrust, compute the first VR from the
speed spread model. (Figure 15.32 provides an example.)
i. Is VR  1.05 VMC ? (Include both VMCG and VMCA .)
1. If no, increase VR until it is.
3. Compute VLOF .
a. From the VR of step 2 and excess thrust, compute VLOF AEO with the
rapid rotation speed spread model.
i. Is VLOF AEO  1.10 VMU AEO?
1. If no, increase VR until it is.
b. With VR (possibly adjusted by VMU AEO), compute VLOF OEI with
the rapid rotation speed spread model.
i. Is VLOF OEI  1.05 VMU OEI?
1. If no, increase VR until it is.

VMCG limited VMU limited

VMCA limited
V2

V2min VR
V1
Airspeed

VSR

Takeoff weight

Fig. 15.31 Establishing appropriate takeoff speed prior to computing distance.

 CHAPTER 15 Takeoff Performance 721

Generic speed spread (VR to V35)

25

2.4% 2nd segment min

climb gradient
20
dV (KTAS), dt (s)

15

10

0
0 0.05 0.1 0.15 0.2 0.25
(T-D)/ W – runway gradient

Fig. 15.32 Generic speed spread model based on second segment drag.

4. Verify V2 .
a. If VR was adjusted by VMC (step 2) or VMU (step 3), use final VR and
excess thrust to compute new V2 .
b. New V2 should still be
i. Greater or equal to 1.13 VSR
ii. Greater or equal to 1.10 VMCA
iii. Providing maneuvering capability
c. Verify climb gradients with new V2 to ensure they meet minimum
requirements.
The generic data of Fig. 15.32 is represented by the following equations:
 
(T  D)
dV ¼ 3 þ 80  runway grad (15:22)
W
 0:3
(T  D)
dt ¼ 3:25  runway grad (15:23)
W
This is typically a good starting point to use and then adjust with results
from flight testing. It represents the ability of the plane to gain speed from VR
to V35 and climb to 35 ft above ground. The runway gradient is added to
account for the fact that the 35 ft altitude is an above ground level (AGL)
value, so if the runway slopes up by, say, 2%, that has an impact on horizontal
distance to climb to 35 ft AGL equivalent to a reduced excess thrust for the
722 Operational Aircraft Performance and Flight Test Practices

airplane. The final curve fit will depend on the takeoff procedure used by the
crew; if the OEM elects to use a simple VR to V2 spread of, say, 3 kt, the
resulting speed spread model may look totally different from the one pre-
sented in the figure.

Certification Regulations: Part 25 Transport Category

Airplanes
Transport Category, Part 25, certification rules are designed to ensure an
airplane will be able to continue climbing following the loss of the most criti-
cal engine and that enough runway is available to safely take off. There is no
single-engine airplane in Part 25. We reviewed the conditions for dry
runways in the preceding pages. Part 25 also offers regulations and guidance
for nondry runways and for clearway use.

Clearway
A significant portion of the takeoff distance can be spent in the airborne
distance from liftoff to 35 ft, especially for the OEI low thrust-to-weight
cases. Some airports have clearways at the ends of runways, which offer an
extended air distance with no obstacles that can be used for the purpose of
computing allowable takeoff weight as limited by takeoff run [§25.113(c)].
§25.113 Takeoff distance and takeoff run
(c) If the takeoff distance does not include a clearway, the takeoff run is
equal to the takeoff distance. If the takeoff distance includes a clearway—
(1) The takeoff run on a dry runway is the greater of—
(i) The horizontal distance along the takeoff path from the start of
the takeoff to a point equidistant between the point at which
VLOF is reached and the point at which the airplane is 35 feet
above the takeoff surface, as determined under §25.111 for a dry
runway; or
(ii) 115 percent of the horizontal distance along the takeoff path, with
all engines operating, from the start of the takeoff to a point
equidistant between the point at which VLOF is reached and the
point at which the airplane is 35 feet above the takeoff surface,
determined by a procedure consistent with §25.111.

To visualize these requirements, we show the following data for a dry runway
(see Fig. 15.33). The main advantage of the use of a clearway is that it offers
up to half the airborne distance of the takeoff to be done over the clearway
instead of the airplane reaching 35 ft above the runway. It allows an operator
to take off at a greater weight (higher VLOF ) when the weight of the airplane is
limited by the length of the available runway. It should be noted that when
the airplane is at the midpoint of the airborne distance, it is at an altitude
lower than 17.5 ft above the runway. We offer an example of an OEI event
at low thrust-to-weight condition in Fig. 15.34; note that at the midpoint,
the airplane is approximately 5 ft above ground. For this example, the
 CHAPTER 15 Takeoff Performance 723

Start AEO
V1 VLOF
35’
Mid-point
Ground roll Air distance
Distance to mid-point
Takeoff run = 1.15 × distance to mid-point (runway req’d) Clearway req’d
Takeoff distance = 1.15 × all engine distance to 35

OEI
Start
VEF V1 VLOF

35’
Mid-point
Ground roll

Takeoff run (runway required) Clearway

Takeoff run distance

Fig. 15.33 Correct usage of clearway for takeoff run on dry runway. Source: Extract
from AC25-7D.

takeoff distance is approximately 5770 ft (35 ft AGL for the OEI case), and
the liftoff point is about 4054 ft from brake release. The use of the clearway
for this example allowed the airplane to take off from a runway where only
about 4900 ft is available for the takeoff run [§25.113(c)(1)(i)] instead of
the full 5770 ft to get to 35 ft above ground. It also puts the airplane much
closer to the ground at the end of the takeoff run.
At this point, we introduce new acronyms, TORA and TODA, as defined
by FAA AC 150/5300-13A, Change 1, dated 26 Feb. 2014, titled “Airport
Design.” These acronyms are used to define the declared runway distances
used by airport owners for turbine-powered airplanes.
• Takeoff run available (TORA): The runway length declared available and
suitable for the ground run of an airplane taking off
• Takeoff distance available (TODA): The TORA plus the length of any
remaining runway or clearway beyond the far end of the TORA

The terrain at the end of a runway can be considered a clearway for the
purpose of computing takeoff distance if it meets the definition of FAA AC
25-7D, para 4.5.3.3:
Clearway is defined in 14 CFR part 1 as a plane extending from the end of the
runway with an upward slope not exceeding 1.25 percent, above which no
object nor any terrain protrudes. For the purpose of establishing takeoff
 724
Operational Aircraft Performance and Flight Test Practices
RADALT (ft) FPA (deg)
50 9
35.53794 8
40 7
6

FPA (deg)
RADALT (ft)

30 5
1.088465 4
1.424903
20 –0.07176801 5.01686 3
1.200967 2
10 1
0
0
DGPS X-Distacne (X-DGPS-1) ft
7000 4054.055 5770.117
Distance along

6000
runway (ft)

4898.54

5000

4000

62 64 66 68 70 72
OF oin
t
GL
VL 6077 Time (s)
.8 id-p 558 ft A 49
62 M .43 35 .075
66 70

Fig. 15.34 Use of clearway on an OEI low T/W case.

 CHAPTER 15 Takeoff Performance 725

distances and the length of takeoff runs, the clearway is considered to be part
of the takeoff surface extending with the same slope as the runway, and the
35 feet height should be measured from that surface.

An additional note from AC150/5300-13A (Airport Design), the clearway

must be at least 500 feet wide centered on the runway centerline. The
maximum declared clearway length may not be more than 12 the runway length.
Advisory Circular 25-7D offers the figure shown in Fig. 15.35a for
additional clarity on the definition of a clearway. Note that to count as a clear-
way, the area should be under the control of the airport. The clearway plane
extends above any obstacles at the end of the runway, including approach
lights.

Wet Runways
From an accelerate-go distance, a wet runway adds friction on the
ground, which can be modeled by an increase in the rolling friction force

Clearway plane
h
0.0% minimum ff pat
Takeo
35’

Clearway

Clearway plane h
ff pat
Takeo 35’
1.25% max.

Clearway

Fig. 15.35 Definition of clearway. Source: Advisory Circular 25-7D.

726 Operational Aircraft Performance and Flight Test Practices

≥V2

Start VEF V1 VLOF

35’
15’
Wet runway takeoff distance

Fig. 15.36 Wet runway with clearway. Source: Extract from AC25-7D.

coefficient mr . The wet runway also impacts the braking capability of the
plane and will often result in adjustment of V1 compared to a dry runway;
we will address the accel-stop and adjustment to V1 in Chapters 17 and 18.
The authorities do recognize the lower probability of both a wet runway
and an engine failure at the critical speed and does allow a distance credit
when performing an accel-go on wet by reducing the height at the end of
the takeoff run to 15 ft from 35 ft, per §25.113, but that distance cannot be
lower than that on a dry runway. Figure 15.36 provides a description of a
wet runway takeoff with a clearway.
§25.113 Takeoff distance and takeoff run
(b) Takeoff distance on a wet runway is the greater of—
(1) The takeoff distance on a dry runway determined in accordance with
paragraph (a) of this section; or
(2) The horizontal distance along the takeoff path from the start of the
takeoff to the point at which the airplane is 15 feet above the takeoff
surface, achieved in a manner consistent with the achievement of V2
before reaching 35 feet above the takeoff surface, determined under
§25.111 for a wet runway.
(c) If the takeoff distance does not include a clearway, the takeoff run
is equal to the takeoff distance. If the takeoff distance includes a
clearway—
(2) The takeoff run on a wet runway is the greater of—
(i) The horizontal distance along the takeoff path from the start of
the takeoff to the point at which the airplane is 15 feet above the
takeoff surface, achieved in a manner consistent with the
achievement of V2 before reaching 35 feet above the takeoff
surface, as determined under §25.111 for a wet runway; or
(ii) 115 percent of the horizontal distance along the takeoff path,
with all engines operating, from the start of the takeoff to a point
equidistant between the point at which VLOF is reached and the
point at which the airplane is 35 feet above the takeoff surface,
determined by a procedure consistent with §25.111.

Contaminated Runway
A wet runway is one that has up to 1/8 in. (3 mm) of water. If the runway
has a water depth of more than 1/8 in., it is considered contaminated, and the
performance numbers provided using the regulation shown earlier may not
 CHAPTER 15 Takeoff Performance 727

provide sufficient runway distance to ensure safe takeoff. As well, winter

operation introduces additional hazards for the takeoff and landing phases
of a flight:
1. Contaminant (ice, snow, sleet, and in some form the de-icing/anti-icing
fluids) on wing, fuselage, and tail will have some impact on the stall speeds
of the aircraft which may force the manufacturer to introduce special
procedures;
2. Contaminants on the runway will impact the acceleration and
deceleration capability of the aircraft:
a. Slush and standing water on the runway will increase the aircraft drag
and therefore reduce in acceleration during the take-off run;
b. Snow, and particularly ice, on the runway will decrease the aircraft
braking force during RTOs and landings.

The FAA does not require formal documentation for runways that are
contaminated, but accepts the guidance material generated in support of
certification for EASA CS25.1591, Performance Information for Operations
with Contaminated Runway Surface Conditions.
CS25.1591
(a) Supplementary performance information applicable to aeroplanes
operated on runways contaminated with standing water, slush, snow
or ice may be furnished at the discretion of the applicant. If supplied, this
information must include the expected performance of the aeroplane
during take-off and landing on hard-surfaced runways covered by these
contaminants. If information on any one or more of the above
contaminated surfaces is not supplied, the AFM must contain a
statement prohibiting operation(s) on the contaminated surface(s) for
which information is not supplied. Additional information covering
operation on contaminated surfaces other than the above may be
provided at the discretion of the applicant.
(b) Performance information furnished by the applicant must be contained
in the AFM. The information may be used to assist operators in
producing operational data and instructions for use by their flight crews
when operating with contaminated runway surface conditions. The
information may be established by calculation or by testing.
(c) The AFM must clearly indicate the conditions and the extent of
applicability for each contaminant used in establishing the
contaminated runway performance information. It must also state that
actual conditions that are different from those used for establishing the
contaminated runway performance information may lead to
different performance.

For Transport Canada certification, standard 525.1581(g) requires:

The Aeroplane Flight Manual shall contain information in the form of
approved guidance material for supplementary operating procedures and
performance information for operating on contaminated runways.

The guidance material of EASA AMC25.1591 is an acceptable means to

satisfy TCCA standard 525.1581(g).
EASA acceptable means of compliance (AMC) for 25.1591 provides a
methodology for computing the impact of various contaminants on the
runway. The method was created by reviewing test results from several test
728 Operational Aircraft Performance and Flight Test Practices

campaigns and dedicated research on the subject. It does recognize the limits
of the methodology and does mention that it should provide conservative
correction based on a series of assumptions. They include:
• The contaminant is spread over the entire runway surface to an even depth
(although rutting, for example, may have taken place).
• The contaminant is of a uniform specific gravity.
• Where the contaminant has been sanded, graded (mechanically leveled), or
otherwise treated before use, it has been done in accordance with
agreed-on national procedures.

The contaminants covered by this methodology include:

• Standing water: Water of a depth greater than 3 mm. (Method covers
depth of 3 to 15 mm.) A surface condition where there is a layer of water of
3 mm or less is considered wet for which AMC 25.1591 is not applicable.
• Slush: Partly melted snow or ice with a high water content, from which
water can readily flow, with an assumed specific gravity of 0.85. (Method
covers depth of 3 to 15 mm.) Slush is normally a transient condition found
only at temperatures close to 08C.
• Wet snow: Snow that will stick together when compressed, but will not
readily allow water to flow from it when squeezed, with an assumed specific
gravity of 0.5. (Method covers depth from 5 to 30 mm, no impact to
acceleration for less than 5 mm.)
• Dry snow: Fresh snow that can be blown, or, if compacted by hand, will fall
apart upon release (also commonly referred to as loose snow), with an
assumed specific gravity of 0.2. The assumption with respect to specific
gravity is not applicable to snow that has been subjected to the natural
aging process. (Method covers depth of 10 to 130 mm.)
• Compacted snow: Snow that has been compressed into a solid mass
such that the airplane wheels, at representative operating pressures
and loadings, will run on the surface without causing significant
rutting (not considered to impact acceleration).
• Ice: Water that has frozen on the runway surface, including the condition
where compacted snow transitions to a polished ice surface (not
considered to impact acceleration).

The main impact of a contaminant on the runway, from an accelerate-go

condition point of view, is the reduction of the airplane’s acceleration while
gears are on the ground. Figure 15.37 shows an example in which, during a
takeoff run, the airplane encountered a large amount of contaminant on
the runway and the sudden reduction in acceleration that goes away after
the contaminant is cleared. As one can observe, the impact on acceleration
can be significant, which is one of the reasons the conservative assumptions
listed previously are specified in developing models.
The guidance of EASA AMC 25.1591 provides mathematical means to
best estimate these impacts and does recognize that it can be difficult to
 KTGS (kt)
140
120
100
KTGS (kt)

80

60
40

20
0
Nx (g)
0.4

CHAPTER 15
0.3

0.2
Nx (g)

0.1

Takeoff Performance
0.0

–0.1
25 30 35 40 45
Time (s)

Fig. 15.37 An airplane encountering contaminants on runway during takeoff.

729
730 Operational Aircraft Performance and Flight Test Practices

reliably predict the impact of the contaminants over the actual takeoff run;
thus, it applies the conservative assumptions listed earlier. The impact on
acceleration is broken down into three types of ground (or precipitation) drags.
• Displacement drag (standing water, slush, wet snow)
• Spray impingement drag (standing water, slush, wet snow)
• Compression drag (dry snow)
The total drag is then the sum of the previous drag components.
Dprecip ¼ Ddispl þ Dspray þ Dcomp (15:24)

Displacement Drag
The displacement drag represents the work being done by a wheel going
through a contaminant and displacing it. This form of precipitation drag
applies to standing water (more than 3 mm thick), slush, and wet snow.
For dry snow and ice, this drag component is considered to be zero.
The model defined in the EASA guidance material is an empirical one that
is a good fit with multiple testing done as part of previous certifications or dedi-
cated testing. The drag applied to a single tire by the contaminant is expressed as
1
Ddispl ¼ r V 2 S CD (15:25)
2
where
r ¼ Contaminant density (water ¼ 999.8395 kg/m3 at 48C)
S ¼ Frontal area of the tire in the contaminant
V ¼ Ground speed through the contaminant
CD ¼ Drag coefficient of single tire in contaminant below aquaplaning
speed ¼ 0.75
One must pay attention that all units are appropriate for Eq. (15.25), just like
computing drag in air. The drag coefficient provided is the ground speed below
the aquaplaning speed VP . That speed, for this method, is simply computed as
pffiffiffi
Vp ¼ 9 P (15:26)
where P is the tire pressure in psi, and the resulting ground speed is in knots.
The area of the tire in the contaminant is the depth of the contaminant d
multiplied by the width of the tire b in the contaminant while under load
(weight applied to this particular wheel).
S ¼db (15:27)
The width of the tire in the contaminant is computed from the equation
provided by the AMC.
"    #0:5
dþd dþd 2
b ¼ 2W  (15:28)
W W
 CHAPTER 15 Takeoff Performance 731

W


Deformed tire
under load
R

d Contaminant

b

Fig. 15.38 Single tire in a contaminant.

where d is the tire deflection under load. This is provided by the manufac-
turer of the tire (load-deflection curve). W is the maximum width of the
unloaded tire, Fig. 15.38.
The AMC specifies that the drag of a dual wheel landing gear (typically
side-by-side) is twice the drag of the single tire. The AMC also specifies
that the drag of a four-wheel undercarriage is 4 times that of a single
wheel. Should the undercarriage have six wheels (e.g., Boeing 777, Airbus
A350-1000), the AMC specifies that the total drag should be 4.2 times that
of an individual tire.
The AMC specifies that the displacement drag reduces to zero at liftoff
from a peak value prior to aquaplaning speed per Fig. 15.39.

1.2

0.8
Factor

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
V/Vp

Fig. 15.39 Effects of speed on drag coefficients.

732 Operational Aircraft Performance and Flight Test Practices

Example 15.2
A Boeing 777-300ER is performing a takeoff when it encounters a 5-mm-deep
water pool on the runway while travelling at 80 kt ground speed. The esti-
mated weight on the nose landing gear at that time is 30,000 lb, and the
weight on each main landing gear is 250,000 lb. (That represents the airplane
weight minus the lift generated. The actual total airplane weight is 575,000 lb.)
The nose landing gear (NLG) tires are Bridgestone 43  17.5R17 with a
nominal pressure of 218 psi; the main landing gear (MLG) tires are Bridge-
stone 52  21.0R22 with a nominal pressure of 218 psi. There are two tires
on the NLG and six on each MLG. What is the aquaplaning speed? What is
the estimated displacement drag for this flight condition?

30,000 lb 250,000 lb/MLG

Solution: The NLG and MLG have the same tire pressure and will thus have
the same aquaplaning speed. The estimated speed for this condition, per the
AMC, would be
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VP ¼ 9 218 psi ¼ 133 kt ground speed
Because the airplane is traveling at a speed below aquaplaning speed, the dis-
placement drag factor will be 1.0, per Fig. 15.29.
The nose gear tire is 43 in. in diameter and 17 in. wide unloaded. The
manufacturer provides the following equation for deflection under load for
the NLG:
0:38 Z
d ¼ 0:12 W þ pffiffiffiffiffiffiffiffiffiffi
P WD
where Z is the load on an individual tire. Because the NLG has two tires, the
load on each tire is 15,000 lb. The deflection of each NLG tire is then
0:38 (15,000 lb)
d ¼ 0:12 (17 in:) þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 3 in:
(218 psi) (17 in:)(43 in:)
The width of the tire in the contaminant (5 mm deep ¼ 0.197 in.), per
Eq. (15.28), is then
"    #0:5
dþd dþd 2
b ¼ 2W  ¼ 13:3 in:
W W

(Continued)
 CHAPTER 15 Takeoff Performance 733

Example 15.2 (Continued)

The drag on each NLG tire is then
1
DdisplNLG ¼ r V 2 S CD ¼ 241 lb
2
The total drag on the NLG is twice this value, 482 lb.
Repeating the exercise for the MLG (52 in. diameter, 21 in. wide) using the
same load-deflection curve with a load on each tire of 41,667 lb would give us
a drag per tire of 323 lb, which translates to 1355 lb per MLG or 2710 lb for
both MLG.
The total displacement drag at the time of encountering the water pool
would then be 3192 lb—a significant force for less than a quarter of an inch
of water on the runway.

Spray Impingement Drag

An aircraft going through standing water, slush, or wet snow will see
sprays being generated by the tires that will hit the aircraft. These sprays
have complex shapes that depend on the aircraft landing gear geometry
and loading, the contaminant depth, and the aircraft ground speed (see
Fig. 15.40). The shape of the nose wheel tires will also have a large impact
on the actual shape of the spray. Part of the spray will impact the aircraft
structure at a normal angle and will result in the loss of momentum (dissipa-
tive force) of the fluid. Fluid running back on the wetted surface of the plane
will cause drag due to skin friction.
A conservative empirical estimate of the skin friction drag from the spray
coming off one landing gear is provided by the guidance material in the fol-
lowing form:
1
Dspray ¼ r V 2 S nwheels CDspray (15:29)
2
where:
r ¼ Contaminant density (water ¼ 999.8395 kg/m3 at 48C)
S ¼ Frontal area of the tire in the contaminant
(per previous method)
nwheels ¼ Number of wheels per landing gear in the contaminant
V ¼ Ground speed through the contaminant

Fig. 15.40 Water ingestion test and water pattern.

734 Operational Aircraft Performance and Flight Test Practices

LNLG

LMLG

Fig. 15.41 Length of spray touching airplane.

The drag coefficient takes on the following form:

CDspray ¼ 8 L (0:0025) (15:30)

The value L represents the wetted fuselage length behind the first contact
point of the spray (see Fig. 15.41). That length varies for each gear, being
longest for the NLG. The value of L changes with ground speed and can
usually be determined when the applicant performs dedicated water trough
testing for water ingestion.
Note that for traditional tricycle landing gear, only half of the spray drag
should be applied for the MLG if only the inner spray plume impacts the
fuselage.

Example 15.3
A Bombardier CRJ-1000 regional jet is going through a water trough with 12 in.
of water in it. The water spray shown in the following figure is produced.
After the test, you compute the following values: LNLG ¼ 110 ft and
LMLG ¼ 45 ft.

LNLG

LMLG

(Continued)
 CHAPTER 15 Takeoff Performance 735

Example 15.3 (Continued)

There are two wheels per gear. The tire width of the MLG at the contami-
nant surface is 8 in.; for the NLG it is 6 in. The ground speed of the plane was
110 KTGS. What is the spray drag of each MLG and of the NLG?

Solution: We first look at the long spray from the NLG. The drag coefficient
would be
h i
CDspray ¼ 8 (110 ft)(0:0025) ¼ 2:2
NLG
  
1 slug ft
Dspray NLG ¼ 2
1:94
ft3
185:6
s
(0:5 ft)(0:042 ft)(2 wheels) 2:2

¼ 3087 lb

The drag from one of the two MLGs would be as follows (note the 12 added
in front of the drag equation—only half of the spray coming off the MLG hits
the fuselage):
h i
CDspray ¼ 8 (45 ft)(0:0025) ¼ 0:9
MLG
    
1 1 slug ft
Dspray ¼
MLG 2 2 1:94 185:6 ð0:67 ft Þð 0:042 ft Þð 2 wheels Þ 0:9
ft3 sec
¼ 846 lb

The total spray drag would then be

Dspray ¼ 3087 lb þ 2  846 lb ¼ 4779 lb

Dry Snow Compression Drag

Dry snow, as compared to the other contaminants with a high liquid
water content, will create very little spray impingement drag, so this drag
component can be neglected. As for the precipitation drag, AMC 25.1591
suggests a slightly different approach than what was presented earlier. The
drag is broken down into a displacement drag DD and a compression drag
DC . The compression drag represents the work required to compress
snow, which has a specific gravity of 0.2. The AMC suggests these empirical
models. First, for tire pressure greater than 100 psi
DC ¼ 74,000 b d (Newtons) (15:31)
where b is the tire width, as computed in the displacement drag method dis-
cussed previously, and d is the snow depth. Both are in meters. For tire
pressure between 50 and 100 psi, the AMC recommends
DC ¼ 56,000 b d (Newtons) (15:32)
736 Operational Aircraft Performance and Flight Test Practices

The displacement drag for dry snow represents the work done to displace
the snow vertically down. With the snow having a specific gravity of 0.2, the
AMC suggests that the drag takes on the following form for tire pressure of
more than 100 psi:
 
56 9
DD ¼ þ b d2 Vg2 (Newtons) (15:33)
R d
where:
R ¼ Tire radius in meters
d ¼ Snow depth in meters
b ¼ Tire width per previous method in meters
Vg ¼ Ground speed in m/s

One can clearly see the empirical nature of the approach, and this is an
acceptable conservative approach to providing guidance to operators when
the previously stated assumptions are followed. For tire pressures between
50 and 100 psi, the displacement drag is
 
52 8
DD ¼ þ b d 2 Vg2 (Newtons) (15:34)
R d
These equations for compression and displacement drags apply only to
the front wheel of any gear (e.g., both wheels on a side-by-side dual wheel
landing gear, only the front two wheels for a four-wheel bogey), because
the snow is assumed to have been compressed and stays compressed after
the passage of the front wheel.

Contaminant Impact on Thrust

Finally, another possible impact of significant contaminant on the runway
is the risk of contaminant ingestion by the engine that could reduce the pro-
duction of thrust. The AMC suggests: “In considering the maximum depth of
runway contaminants it may be necessary to take account of the maximum
depth for which the engine air intakes have been shown to be free of ingesting
hazardous quantities of water in accordance with CS 25.1091(d)(2).”

Part 23
As compared to Part 25 airplanes, where the takeoff distance is computed
from brake release to a height of 35 ft above the runway, Part 23 (non-
commuter type) regulations require the airplane to reach 50 ft above the
runway surface. This reflects the fact that Part 23 does allow single-engine
airplanes, so the requirements did not specifically account for an engine
failure except for the commuter category airplanes (takeoff weights
between 12,500 lb and 19,000 lb) and the Part 23 jets (most):
 CHAPTER 15 Takeoff Performance 737

FAA Part 23 Pre-Amendment 64:

§23.51 Takeoff
(a) For each airplane. . .the distance required for takeoff and climb over a
50-foot obstacle must be determined with—
(1) The engines operating within approved operating limitations; and
(2) The cowl flaps in the normal takeoff position.
(d) For commuter category airplanes, takeoff performance and data as
required by Secs. 23.53 through 23.59 must be determined and included
in the Airplane Flight Manual.

§23.59 Takeoff distance and takeoff run

For normal, utility, and aerobatic category multiengine jets of more than
6,000 pounds maximum weight and commuter category airplanes, the
takeoff distance and, at the option of the applicant, the takeoff run, must
be determined.

(a) Takeoff distance is the greater of—

(1) The horizontal distance along the takeoff path from the start of the
takeoff to the point at which the airplane is 35 feet above the takeoff
surface as determined under Sec. 23.57 [OEI]; or
(2) With all engines operating, 115 percent of the horizontal distance
from the start of the takeoff to the point at which the airplane is
35 feet above the takeoff surface, determined by a procedure
consistent with Sec. 23.57.
In FAA Amendment 64, with the new assignment of categories of airplanes
based on their performance capabilities, the regulation for takeoff distance
for Part 23 became:
§23.2115 Takeoff Performance
(a) The applicant must determine airplane takeoff performance accounting
for—
(1) Stall speed safety margins;
(2) Minimum control speeds; and
(3) Climb gradients.
(b) For single engine airplanes and levels 1, 2, and 3 low-speed multiengine
airplanes, takeoff performance includes the determination of ground roll
and initial climb distance to 50 feet (15 meters) above the takeoff surface.
(c) For levels 1, 2, and 3 high-speed multiengine airplanes, and level 4
multiengine airplanes, takeoff performance includes a determination
the following distances after a sudden critical loss of thrust—
(1) An aborted takeoff at critical speed;
(2) Ground roll and initial climb to 35 feet (11 meters) above the takeoff
surface; and
(3) Net takeoff flight path.

The new approach with Amendment 64 and later for Part 23 is to get the
acceptable means of compliance (MOC) “accepted” by the FAA prior to the
collection of data. One such MOC for takeoff speeds is ASTM F3179,
where the speeds could be shown in a format similar to that presented in
Fig. 15.30 (see Fig. 15.42).
One major difference in this approach is the removal of VMU testing to
define minimum VLOF and the addition of a requirement on minimum VR
738 Operational Aircraft Performance and Flight Test Practices

to be at least 10% greater than stall speed out of ground effect. This is a reflec-
tion of the general difference in configuration between the majority of Part 25
airplanes and Part 23 airplanes. Typically, Part 25 airplanes (especially the
commercial airplanes) have slatted wings and a long fuselage, resulting in a
larger angle of attack range and lower tail clearance, whereas Part 23 air-
planes typically have no slats (smaller AoA range) and a shorter fuselage
(more tail clearance). So the latest version of Part 23 acceptable MOC puts
the emphasis on stall margin at VR .

V-Go vs V-Stop
For Part 23 and Part 25 airplanes, most certification authorities (FAA,
EASA, TCCA, ANAC, etc.) require the use of a single decision speed V1
for the crew to use in deciding to go or stop following the failure of an
engine (or other failures deemed critical by the crew). This is used in part
for ease of crew training; the takeoff procedures defined for those airplanes
do reflect the use of this single speed.
Some agencies and most military programs allow the use of two decision
V-speeds, one for go and one for stop. This often allows for maximization of
payload capability from a given airfield instead of being limited by the use of a
single V1 speed that may not always allow balancing of the takeoff distance
(see Fig. 15.43). Some military operations have allowed the go speed to be
defined as the minimum speed that will allow the airplane to become air-
borne (ground run, not distance to 35 ft AGL) within the available runway.

Selected by
Para 6.3.1 applicant
Critical engine
≥1 VMCG Para 6.3.1.1 VEFmin VEF
failure speed
“Decision speed”
≥1 VEF + Vpilot Para 6.3.1.2 V1min V1 Highest speed for start
of braking – RTO
≥V1 Para 6.3.2.1 Rotation speed
≥1.05 VMC Para 6.3.2.2 VRmin VR
≥1.10 VS1 Para 6.3.2.3
Liftoff speed
VLOFmin VLOF

Para 6.3.4 Takeoff

≥1.20 VS1 V2min V2 safety
Para 6.3.4 speed
≥1.10 VMCA
≤ 35 ft AGL Para 6.3.4
Para 6.3.2.4
climb gradient
AMC: ASTM F3179/F3179M – 18

Fig. 15.42 Part 23 acceptable MOC for takeoff speeds.

 CHAPTER 15 Takeoff Performance 739

V V
Liftoff
Vgo Vstop
VEF

Available

Available
runway

runway
X X

Fig. 15.43 Using two decision V-speeds.

Exercises
1. A commercial jet is going to take off on a runway with a clearway. What is
the minimum takeoff run required to allow the takeoff if the airplane has
an average T/W of 0.32 from brake release to VR on a flat dry runway with
no winds, the excess thrust at VR is 0.22, and the excess thrust at 35 ft is
0.20. Use available models in this chapter to compute the distance.
2. An airplane has MLG tire pressures of 185 psi. What is the expected
aquaplaning speed?
Chapter 16 Takeoff
Performance
Testing

Chapter Objective
Testing for takeoff distance involves careful preparation, early investigation
and validation of takeoff procedures, validation of the excess thrust model
of the airplane in the takeoff configuration, validation of the takeoff speeds
against certification requirements, and demonstrating that the safety
margins are sufficient. Part 25 ensures safe continued takeoff following an
engine failure by design. This must be verified by flight testing. We look at
some of the expected behaviors during takeoff testing and the risks involved
while collecting certification data.

Takeoff Distance Modeling

T
akeoff distance testing and modeling is an essential part of a certifi-
cation program. The data are used to produce the airplane flight
manual (AFM) Performance section. Before one can proceed to
takeoff distance testing, several activities must be completed and the data
validated so the airplane manufacturer can cost-efficiently execute the
takeoff distance measuring part of the certification program. Figure 16.1
offers an example of the timeline usually expected to develop and validate
takeoff distance information for the AFM.

Stall speed Takeoff procedure validation Test

testing Validation of ground accel and program
Validating VSR speed spread model timeline

First VMCG Takeoff VMCA VMU A/C config Takeoff distance AFM
flight testing trim setting testing testing frozen, testing
validation procedures Including abuse
set, speeds testing
validated

Fig. 16.1 Timeline to takeoff distance test.

741
742 Operational Aircraft Performance and Flight Test Practices

The importance of the validation milestone cannot be underestimated.

An airplane manufacturer may often deploy to a remote airport with a very
long runway, sufficient width, with no obstacles on either end of the main
test runway and very little air traffic into and out of the airport to be able
to expedite the testing. As well, some of the testing will involve validating
the baseline speeds and procedures by executing takeoff with significant devi-
ations from those speeds and procedures to show that the airplane can still
safely take off and climb away from the ground.
The first part of the preparation is the validation of the stall speeds of the
airplane for the various takeoff flaps. This is typically done very early in the
program and becomes the foundation of the preparation for the follow-on
testing because it sets minimum V2 speeds. This is an important first step
in the determination of the takeoff distance because it usually allows the air-
plane manufacturer to remove the initial takeoff speed adders used early in
the program as an additional margin of safety.

Takeoff Safety Speed V2 and Rotation Speed VR

With the minimum V2 speeds determined, the manufacturer turns to the
validation of the initial speed spread model. As a reminder, the takeoff speeds
must meet the requirements outlined in Fig. 16.2. So starting with the initial
estimates for VMCA , VMCG , and VMU and their impact on the V-speeds (VR
and V2 for continued takeoff), the manufacturer proceeds to analyze each
and every takeoff done to verify the speed spread model and the takeoff
procedure (next section). As the general test campaign progresses, the

Selected by
applicant

Chapter 17 VEFmin VEF Critical engine

failure speed
“Decision speed”
Chapter 17 V1min V1 Highest speed for start
of braking – RTO

≥1.05 VMC 25.107(e)(1)(ii) VRmin VR Rotation speed

≥1.10 VMU,AEO 25.107(e)(1)(iv)

VLOFmin VLOF Liftoff speed
≥1.05 VMU,OEI
25.107(b)(1) Takeoff
≥1.13 VSR V2min V2 Safety
25.107(b)(3)
≥1.10 VMCA Speed
25.107(c)
25.107(c)(3) ≤35 ft AGL
25.143(h) Maneuvering capability 25.107(c)(2) 25.121(b)
Climb gradient

Fig. 16.2 Takeoff V-speeds.

 CHAPTER 16 Takeoff Performance Testing 743

manufacturer’s performance engineer will have plenty of data to review and

will adjust the model. One of the first things that will need to be verified is the
ground airspeed indication calibration.

Airspeed Indication Calibration: Ground

To provide accurate takeoff distance, one must have a good understand-
ing of the relation between knots indicated airspeed (KIAS) and knots cali-
brated airspeed (KCAS) while in the ground acceleration phase of the
takeoff. Under FAA 14 CFR Part 25, §25.1323(b) provides the requirements
to be used by airplane manufacturers.
§25.1323 (b) ground speed calibration
(b) Each system must be calibrated to determine the system error (that is,
the relation between IAS and CAS) in flight and during the accelerated
takeoff ground run. The ground run calibration must be determined—
(1) From 0.8 of the minimum value of V1 to the maximum value of V2 ,
considering the approved ranges of altitude and weight; and
(2) With the flaps and power settings corresponding to the values
determined in the establishment of the takeoff path under §25.111
assuming that the critical engine fails at the minimum value of V1 .

To validate the airspeed indication, the airspeed indicator (ASI) value must
be compared to another truth source, similar to the general procedures
described in Chapter 4, but this source must account for the dynamic con-
dition of the takeoff phase. Two methods will be presented here. The first
compares the airplane’s air data system computed true airspeed against the
ground speed [from inertial or differential global positioning (DGPS)
system] plus the head wind component along the runway (from an indepen-
dent calibrated weather station) converted. The second method will compare
the airplane’s production air data system to a reference static pressure
(trapped static).
For the first calibration method, we will compare the airplane’s knots true
airspeed (KTAS) to a true airspeed derived from a ground speed source (typi-
cally an inertial reference system or a DGPS system) with a wind component
along the axis of the runway (from a calibrated weather station). This test
method requires a weather station reading winds at or near the same
elevation as the air data probe of the airplane. Then one must remember
that the weather station does not measure winds at the airplane, but at a
fixed location on the airport property (see Fig. 16.3). The desired calibration
must be most precise between min V1 and max V2 , so the ideal location for
the weather station is typically near the expected liftoff point of the airplane
for a continued takeoff. To minimize scatter during the analysis and improve
the comparison between the airplane’s system and the derived airspeed, a
calm day is best; one with zero winds is ideal.
Note that the airplane’s true airspeed is derived from the indicated air-
speed (measured total pressure and static pressure) and the airplane’s
 744
Operational Aircraft Performance and Flight Test Practices
Airspeed ground calibration check
KTAS (kt)
150
KTAS (kt)

100 138.2598
50

Wind
0
Headwind (kt) KTAS-KTGS (kt)
KTAS-KTGS (kt)

14
Headwind (kt)

7.882718
12 8.385562
10 Weather
8 station

way
6

Run
Pitch (deg)
25
Pitch (deg)

15
0.2967768
5
–5

n
Hp-DGPS (ft)

ctio
Hp-DGPS (ft)

–100

dire
–120 –135.7588

eoff
–140

Tak
43
9. R
0 20 40 60 80 100 120 140

V
87
KTGS (kt)

12
Fig. 16.3 Airspeed ground calibration check.
 CHAPTER 16 Takeoff Performance Testing 745

measured temperature. If uncertainty exists as to the reading of the airplane’s

air temperature sensor in this phase of flight, one can instead convert the true
airspeed derived from the weather station into a calibrated airspeed (using
weather station measured temperature and static pressure). The end goal
of this exercise is to validate the indicated airspeed to the crew so as to
reliably predict takeoff distance.
In Fig. 16.3, the airplane’s KTAS (computed using measured static
pressure, total pressure, and temperature) is plotted against the knots true
ground speed (KTGS, derived from an inertial reference unit—in this case,
an independent device to the air data system); an offset is noticed.
Then one computes the difference between the two signals (second strip,
KTAS – KTGS), which provides an equivalent headwind component; this
is compared to the wind measured by the independent weather station,
resolved against the direction of the takeoff (headwind component). In this
example, we see the two signals correlate very well (within about 1 kt).
Another method that can be used is the trapped static pressure method,
which can use either mechanical or electronic trapping. The method consists
of “trapping” a reference static pressure at brake release and computing a
reference airspeed by combining this reference static with the airplane’s
total pressure reading from the pitot probe, assuming negligible error on
the pitot reading. This reference airspeed is compared to the airplane’s air-
speed that is derived from the airplane’s pitot and static pressure sources
(see Fig. 16.4).
For the mechanical system, an airtight bottle that is sufficiently large is
inserted into the static air data system, and a means to block the air pressure
from escaping the system is provided to the test crew so they can trap the
static pressure at brake release. For the data reduction, one assumes good
total pressure and compares the evolution of airspeed during a ground accel-
eration on a level runway (need constant elevation). The airspeeds from both
systems are recorded (e.g., video, camera, data recorder) and compared, and

Plane’s altimeter
Plane’s static Plane’s ASI
Plane’s pitot

Airtight bottle
Test valve

Plane’s static Plane’s pitot

Test ASI
Test altimeter

Fig. 16.4 Trapped static source test configuration.

746 Operational Aircraft Performance and Flight Test Practices

the difference in airspeed is then attributed to the airplane’s static pressure

error reading during the ground run. The static pressure can also be electro-
nically trapped by setting a derived parameter in the data system equal to the
static pressure at the time of brake release. This value is then used in combi-
nation with the airplane’s pitot pressure signal to compute an airspeed. One
should verify the lag difference between the speed computed by the exper-
imental data system and the production system so as to not bias the
ground airspeed calibration.
If the runway is not perfectly flat, then one must compare the evolution of
the static pressure measured along the ground acceleration against the refer-
enced trapped pressure, converted to an equivalent altitude, to the expected
static pressure due to the geometric change in altitude (either from runway
survey and airplane position or from a DGPS height signal). The last strip
of Fig. 16.3 shows such a comparison. At the start of the ground run, the
difference in altitude between the DGPS geometric system and the altimeter
was approximately 140 ft. That difference reduced to 130 ft by midspeed and
was slowly getting back to the original 140 ft when the airplane reached VR .
For this example, the nominal test altitude was 11 ft. A 10-ft difference at
about 80 kt is equivalent to 1.4 kt calibrated difference. By VR (138
KTAS), the difference in altitude had changed to about 4 ft, resulting in a
calibrated-to-indicated airspeed difference of about 0.3 kt.
As a final note on ground airspeed calibration, the airplane’s display lag
(see Fig. 4.39 in Chapter 4) should also be verified by ground test (static air-
plane) initially and validated by continued takeoff testing. A telltale sign of
possible display lag is a consistent late crew input (elevator stick input) at
rotation speed VR when airspeed is recorded from the air data computer,
with a crew feedback that they are rotating “on condition.”

Takeoff Procedure Validation

The airplane’s performance for takeoff is based not only on the physical
properties of the design (excess thrust and lift capability), but also on the
pilot’s ability to execute the maneuver in a repeatable way. §25.101,
General, specifies:

(e) The airplane configurations may vary with weight, altitude, and
temperature, to the extent they are compatible with the operating
procedures required by paragraph (f) of this section.
(f) Unless otherwise prescribed, in determining the accelerate-stop
distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must be
made in accordance with procedures established by the applicant for
operation in service.
(h) Procedures established under paragraphs (f) and (g) of this section
must—
(1) Be able to be consistently executed in service by crews of average
skill;
 CHAPTER 16 Takeoff Performance Testing 747

(2) Use methods or devices that are safe and reliable; and
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.

Although the first part of the takeoff (brake release to VR ) may be relatively
straightforward, the next part involves pilot in the loop, which will influence
the end results. At the start of the flight test program, before proceeding
to formal distance measurement, the applicant will get a chance to
perform several nontesting takeoffs. The performance engineers should
review the data collected during those takeoffs as well as the crew comments
and work with the pilots to adjust the takeoff procedure to make it more
repeatable.
Things that will need to be verified include:
• Trim setting
• Target pitch attitude
• Target pitch rate
We will address these items in the next few sections.
The performance engineer should then review the resulting pilot forces
during the takeoff, the pitch attitude of VLOF for tail clearance, and the
ease of capturing target pitch. The engineer should also start validating
their basic models, including the lift coefficient at liftoff.

Takeoff Trim Setting

From our discussion in the previous chapter, we indicated that a proper
trim setting on takeoff will reduce the pilot workload and help the airplane
capture flight path (pitch) and airspeed by 35 ft above ground level (AGL).
The proper trim setting is a complex mix of force vectors changing with
speed and is dependent on airplane configuration (center of gravity location,
flaps used, and gear position; see Fig. 16.5).
The location of the center of gravity (CG) has a dominant effect. A more
forward CG will create a more nose-down pitching moment, which will
require more nose-up trim (see Fig. 16.6).

L L
Downwash T
V Downwash
V T

W W

Fig. 16.5 Some of the factors impacting proper takeoff trim setting.
748 Operational Aircraft Performance and Flight Test Practices

Nose-Up
Trim
Ref Ref More Ref
deflection
1.13 Vsr
Less
CG effects deflection VMC limited
Nose-Down Flap deflection Speed effect >1.13 Vsr
Trim
Fwd Aft Fwd Aft Fwd Aft
CG CG CG CG CG CG

Fig. 16.6 Various parameters’ impact on trim setting.

The flap configuration selected will have an impact on the nose-down

forces, with a typical behavior being that a larger flap deflection results in a
larger nose-down pitching moment for a given lift coefficient, thus requiring
more nose-up trim. Flaps downwash also impacts flow on the horizontal tail,
so the tail location with respect to the wing is also a very important factor
to consider.
At lower airspeed, an airplane with positive static longitudinal stability
(increased nose-down pitching moment coefficient with increasing lift
coefficient) will generate more nose-down pitching moment coefficient
(lower dynamic pressure) than a higher speed. Thus, a takeoff at V2min
(1.13VSR ) requires more nose-up trim than a V2 limited by other con-
ditions (and thus faster than 1.13VSR ). (See Fig. 16.2 earlier in the
chapter.)
Gear down will also typically generate a nose-down pitching moment
(drag from the gear below the center of gravity), which goes away as the
gear retracts. The applicant needs to decide if the trim setting for takeoff
will account for gear position. The end of the takeoff segment (35 ft AGL)
will most likely be with landing gear still down, possibly in transition for
the lower T/W conditions.
The thrust lapse rate, as seen in Chapter 8, is such that thrust reduces
with increasing airspeed, the magnitude of which is a function of the propul-
sion system used. Depending on thrust line, this lapse rate will impact the
nose-up/down pitch force.
Engine exhaust flow and propwash may also impact the downwash at the
tail, especially as the airplane rotates and the tail gets closer to the ground, as
discussed in Chapter 15, Fig. 15.13. A T-tail airplane (horizontal tail atop of
the vertical tail) will be less subject to this effect than an airplane with the
horizontal tail on the fuselage.
The selection of the trim setting can be resolved by analysis, accounting
for all the variables listed previously, and validated by flight test. We under-
stand that a proper takeoff trim setting greatly facilitates the pilot’s task of
capturing the correct target pitch rate and the proper takeoff safety air-
speed [§25.107(e)] by 35 ft AGL (see Fig. 16.7). The trim setting must
 CHAPTER 16 Takeoff Performance Testing 749

also be compatible with the takeoff procedure (i.e., a continuous pull to a

single target pitch or a segmented rotation with two pitch targets—one
near the ground and one later above 25 ft). So a typical test scope for
this will be to initially start by analysis pre–flight test, and then move on
to collect initial pitch forces during the nondedicated takeoffs and
mapping the results. Some trim shots at the proper takeoff speeds [V2
for one engine inoperative (OEI) and V2 þ XX for all engines operating
(AEO)], flaps, and CG with thrust at the takeoff setting can give a first vali-
dation of the analysis. As a flight test campaign progresses, the values are
clarified. Finally, during the dedicated takeoff distance performance testing,
including the testing of the OEI/Sim OEI conditions, the final settings
are validated.

140
Target V35 121
120
KCAS

Target VR 104
100
80
150
Wheel Spd

100
VLOF

50
0
20
10
AoA

0
–10
20 Target pitch 14
Pitch

10

0
–10
60 Pitch force
Stick force

40 ~15 lb
VR

20
0
–20
60
V35

40
RADALT

20
0

Fig. 16.7 Example of proper takeoff trim setting.

750 Operational Aircraft Performance and Flight Test Practices

Minimum Control Speed: Air VMCA

Because VMCA could define the takeoff speeds at low weight, one should
have completed VMCA testing and analysis prior to performing takeoff dis-
tance testing. This testing involves testing maximum thrust asymmetry at
the lightest weight possible. As a reminder, VMCA is the speed at which a
steady heading can just be maintained with the application of maximum
available rudder with a bank angle no greater than 5 deg (into the live
engine), with one engine inoperative and the remaining engine(s) developing
full takeoff thrust.
The aim of the test is to create the largest thrust asymmetry by shutting
down the most adverse engine. From Fig. 16.8, one can observe that this is
typically the most outboard engine. For jet-driven airplanes, a symmetri-
cally built airplane will typically not have a worst left or right engine,
and the manufacturer can elect to perform VMCA testing with either one
shut down.
Propeller-driven airplanes will have additional factors to consider when
planning VMCA testing. Most North American twin-engine airplanes have
clockwise-rotating propellers (when viewed from behind the airplane). A
propeller disk at a positive angle of attack will see a nonsymmetrical
loading of the propellers where the downgoing propeller will see a higher
angle of attack (generate more thrust), and the upgoing propeller will see a
reduced angle of attack (reduced thrust) (see Fig. 16.9).
On clockwise-rotating propellers, the right side of the disc will generate
more thrust (see Fig. 16.9) than the left side. (This effect is also referred to
as p-factor.) This will move the propellers’ aerodynamic center to the right
of the airplane centerline and induce an increasing yaw moment to the left
with increasing angle of attack or increasing power under normal operation.
It will also make the failure of the left engine more critical than the right
engine for symmetric engine distribution. The Airbus A400M, a four-engine
military transport, was fitted with alternating clockwise and counterclock-
wise propellers so as to make the distribution of propeller force more sym-
metric (see Fig. 16.9). Turbofan airplanes will have little to no effect due to
the airflow being turned by the inlet and the engine fan face essentially pre-
sented with the same, low angle of attack flow.
The lift distribution behind a propeller disk is highly complicated by the
rotation of the flow, creating locally increased angle of attack on the upgoing
propeller side and reduced local AoA on the downgoing side. At the same
time, the energy added to the flow by the propeller disk will create a zone
of increased dynamic pressure behind the disk. Impact can also be on the
horizontal tail if it is located in the propwash (Fig. 16.10), and pitch capability
may be impacted. On the other hand, turbofan-powered airplanes typically
have concentrated exhaust flow with minimal impact on the wing aerody-
namics. Part 25 airplanes, including props, are now equipped with multifunc-
tion spoilers that also help with roll control.
 Fn

yeng Fn
Fn

CHAPTER 16
yeng DW
DW
L

Takeoff Performance Testing

DW

Fig. 16.8 VMCA thrust asymmetry.

751
 752
Operational Aircraft Performance and Flight Test Practices
Clockwise
Ascending Descending Slow speed
blade blade High AoA Counterclockwise
Clockwise

Clockwise
Clockwise
Smaller Larger
blade blade
AoA AoA Counterclockwise

Fig. 16.9 P-factor.

 CHAPTER 16 Takeoff Performance Testing 753

Fig. 16.10 Prop wash impact on wing lift and tail aerodynamics based on geometry.

The relative size of the propeller disk to the wing/flap span will impact
how much of the lift is lost, and by extension, how much of an induced
roll can be created by one engine failing. Most of the wing lift is produced
inboard, especially with the flaps deployed. An engine failing to idle or wind-
milling can be worse than full feather (Fig. 16.11) in terms of loss of lift and
higher drag produced, a fact that was recognized by CAR 4b where the CAA
board needed to approve the feathering system.
Propwash can be relatively big compared to the airplane general size.
The swirl in the propwash may tend to move one propwash (typically the
left side) closer to the vertical tail, creating uneven flow distribution. Most
turbofans have stators behind the last rotor stage to help straighten the
flow and increase pressure, therefore removing most of the swirl from
the exhaust.
With all of this in mind, the tester must verify that the engine selected to
be shut down is the critical engine for VMCA testing. Some preliminary air-
speed reductions to near VMCA speed with either the left or the right
engine shut down and comparing the impact on flight controls deflection
(more rudder and/or ailerons required with one engine) will validate the
selection.
The first part of the VMCA testing is the static test, where the critical
engine is shut down, the airplane is in the proper configuration for the test
(takeoff or landing flaps), and it is trimmed at an airspeed above the expected
VMCA . Then the remaining engine (or engines) is moved to maximum thrust,
and the airplane’s airspeed is slowly reduced while the pilot maintains no
more than 5-deg bank into the live engine. The pilot focuses on executing

VMCA
J2
T
AoA J1 >J2 T VMCA
CL J1
Airfoil no prop AoA
or feathered CL/CD
T J2

Windmilling J1
prop Airfoil no prop
or feathered

D  Windmilling
 prop

Fig. 16.11 Complex flow on propeller with propwash.

754 Operational Aircraft Performance and Flight Test Practices

the maneuver until the point where the heading cannot be maintained with
maximum rudder available; this becomes the test-demonstrated VMCA .
During the test, one must check roll control (aileron travel) and the risk of
wing tip stall. Testing should not be performed slower than stall warning.
Figure 16.12 is an example of a static VMCA test condition for a jet-driven
airplane (little impact on aerodynamics). The airplane was trimmed at 110
KCAS with one engine off and at about 5 deg of bank angle. Then the remain-
ing engine was brought to maximum thrust, and the pilot started slowing
down. Note how quickly the altitude increases. As the airplane slows
down, the pilot tries to maintain 5-deg bank while applying just enough
rudder to maintain the heading. This is a very dynamic maneuver with mul-
tiple axes for the pilot to control.
The test difficulty is often compounded with the pilot losing sight of the
horizon because of the high pitch attitude during the test, which then makes
it harder to see the heading change unless the pilot is flying using the artificial
horizon; however, this requires the safety pilot to keep an eye on the outside.
Testing is over when the pilot clearly noticed loss of heading control.
As we will see in the data reduction section, the testing for VMCA should
be performed at minimum weight (most adverse); light weight plus
maximum thrust on the remaining engines lead to high rate of climb. The
test requirements mandate that testing be done with engines developing

Static VMCA
Max thrust set condition EOT
KCAS (kt) Altitude (ft)
115 3200
Altitude (ft)
KCAS (kt)

110 2800
105 2400
100 2000
Pitch (deg) Roll (deg)
25 2
Roll (deg)
Pitch (deg)

20 0
15 –2
10 –4
5° Bank
5 –6
Rudder (deg) Heading (deg)
40 –10
Heading (deg)
Rudder (deg)

Max 30 –12
rudder
20 –14
10 –16
0 –18
Yaw rate (deg/s)
Yaw rate (deg/s)

2
0
–2
0 10 20 30 40 50
Time (s)

Fig. 16.12 Example of static VMCA testing.

 CHAPTER 16 Takeoff Performance Testing 755

no less than 95% of maximum asymmetric thrust. We saw that maximum

thrust may reduce with increasing altitude, so time and low altitude
(reduced test safety margin) are the enemies of the test crew. The maneuver
must be executed as quickly and as precisely as possible. The testing may also
get more complicated if the airplane gets near maximum roll control and the
developing roll rate due to rudder input becomes harder to arrest.
A useful way to reduce the data collected from a VMCA test is to assume
the condition, as shown in Fig. 16.12, is obtained under steady unaccelerated
flight with equilibrium for side forces of rolling and yawing moments.

Side forces CL sinðfÞ ¼ Cyb b þ Cydr dr þ Cyda da (16:1)

Rolling moments 0 ¼ C‘b b þ C‘dr dr þ C‘da da (16:2)

Yawing moments Cna ¼ Cnb b þ Cndr dr þ Cnda da (16:3)

where:
b ¼ Sideslip angle
dr ¼ Rudder deflection angle
da ¼ Aileron deflection angle
f ¼ Bank angle
Cy ¼ Side force coefficient as a function of b, dr , da
C‘ ¼ Rolling moment coefficient as a function of b, dr , da
Cn ¼ Yawing moment coefficient as a function of b, dr , da
Cna ¼ Yawing moment coefficient due to asymmetric thrust

This yawing moment can be expressed as follows:

 
Fa yeng Fa yeng 
Cna ¼ ¼ CL (16:4)
1 2 W b
rSL VEAS Sb
2
The asymmetric net thrust Fa is equal to Eq. (16.5) for one engine wind-
milling or Eq. (16.6) for one engine at idle.

Fa ¼ T þ Dw (16:5)

Fa ¼ T  Tidle (16:6)

In Eq. (16.4), yeng is the distance of the thrust line or windmilling drag line
from the airplane centerline, as shown previously in Fig. 16.8, and b rep-
resents the airplane’s wingspan.
If one repeats the slowdown to minimum control condition at three or
more different bank angles for a given airplane configuration, one can
develop a reasonable model that defines the airplane’s aerodynamic capabili-
ties under VMCA conditions. One can solve the following equation to define
756 Operational Aircraft Performance and Flight Test Practices

the airplane’s characteristics:

8 9
8 9 > f >
<b= >
< >
=
 0
d ¼ ½ A1  CL (16:7)
: r; >
> Fna yeng >
>
da : ;
W b

where the matrix A defines the characteristics of the airplane configuration

tested.
2 3
Cyb Cydr Cyda
½ A ¼ 4 C‘b C‘dr C‘da 5 (16:8)
Cnb Cndr Cnda

This method is an acceptable approach to fit the data collected during a

flight test program to then perform extrapolation of the data to minimum
flight weight and maximum thrust condition. (One should not use wind
tunnel data to perform the certification analysis, but one can use wind
tunnel data to get an initial estimate of the coefficient values.) This approach
also assumes that the rudder can reach its maximum deflection angle because
it does not address aerodynamic limitations on the rudder displacement such
as reaching a hinge moment limit for the actuators or a rudder pedal force
limit (150 lb per §25.149). If the rudder does reach maximum deflection,
then this method has been shown to apply to jet- or propeller-driven air-
planes (irrespective of the type of propeller used – fixed pitch, variable
pitch, or constant speed).
The method also allows us to correct thrust to the 100% expected value in
service as long as the test is executed with at least 95% of the in-service
maximum. The 5% extrapolation is standard in that maximum thrust is
often under low altitude (standard sea level) and lower temperature. The
test must be done at a minimum safe altitude to complete the test. On a
twin-engine airplane, that involves shutting down one engine and setting
the remaining engine to maximum with the risk that this only available
engine may fail during the test. (Bird ingestion is always a risk.) On an air-
plane with three or more engines, the maximum asymmetry is obtained
when the two outboard engines and the nontest engine(s) can be used to
ease the test. (One could possibly even execute the test in near level flight,
but one must remember that the remaining engine may have an aerodynamic
influence on the airplane characteristics matrix A that changes with thrust.)
Finally, the method also allows the tester to correct the airplane weight to
the most adverse for VMCA , which is the minimum flight weight (see
Fig. 16.13). The test weight for the flight test airplane is never close to the
minimum flight weight. One must have a reasonable amount of fuel to
execute the test, there may be some ballast installed in the airplane to
achieve the proper center of gravity, and of course there is always the
weight of the instrumentation on board. So it is understood that the airplane
 CHAPTER 16 Takeoff Performance Testing 757

35
5 deg bank into the live engine

30 Max rudder (set to min tolerance)

25
Rudder required (deg)

Lower CL
20 Vsr Min flight weight
Stall warning
deflection limited)
15
Min test weight
VMCA (Rudder

10 MTOW
Higher CL

5
Test range
0
90 95 100 105 110 115 120 125
Airspeed (KIAS)

Fig. 16.13 Rudder required for most adverse CG and highest asymmetric thrust.

will not achieve the real VMCA during the test; however, if the testing is done
so that the airplane reaches an angle of attack (lift coefficient) near or greater
than the AoA for VMCA at minimum weight, then the aerodynamic charac-
teristic matrix A developed in flight testing will allow for the weight extrapol-
ation. This means testing to either maximum rudder deflection at 5 deg or
less (see Fig. 16.14) or to test to stall warning at 5-deg bank (see Fig. 16.13).
Using this approach to model the airplane behavior as it approaches
VMCA gives us a glimpse of the test risk but also the operational impact of
this speed. Figure 16.14 shows that at wings level condition, the minimum

Min test weight

35 15
Wings level
30 Max available rudder 10
Rudder required (deg)

Sideslip angle (deg)

25 3-deg bank 5
20
5-deg bank 0
15
Vsr Vsw –5
10
EOT
5 –10

0 –15
85 90 95 100 105 110 115 120 –5 –4 –3 –2 –1 0 1
Airspeed (KCAS) Bank angle (deg)

Fig. 16.14 Bank angle effect on rudder required and resulting sideslip angle.
758 Operational Aircraft Performance and Flight Test Practices

control speed can be much higher than it is at 5-deg bank into the live engine.
Figure 16.13 shows that VMCA is defined by the minimum flight weight of the
plane, a value defined by the applicant and not necessarily a zero fuel, no
payload condition. Figure 16.14 also shows the trend in sideslip angle
being generated by the engine-out condition, which impacts the control
drag defined in Chapter 13. The regulations of §25.107(b)(3) require that
V2 be no less than 1.1VMCA to provide some margin on takeoff while the air-
plane is not yet banked to 5 deg.
In recent years, the members of the Manufacturers Flight Test Council
(MFTC, a group this author has had a chance to be part of) have been able
to show the authorities that testing for static VMCA with one engine at idle
rather than shut down does provide enough data for the manufacturer to
compute the VMCA , engine shut down, while using proper analysis
methods. This approach adds to the safety of the test execution, especially
for twin-engine airplanes.

Dynamic VMCA
Once the static VMCA is established, the manufacturer must verify that
the speed still protects the airplane under dynamic conditions. This is
usually done by a test in which the airplane is set up at a speed near VMCA
and takeoff thrust is applied on all engines; when the engines reach correct
thrust, the most adverse engine is cut in the most adverse way (typically a
fuel cut). The pilot must react to the sudden loss of thrust (rapid slowdown
when airplane is flying at or near stall warning) and develop heading change
and most likely a roll angle increase (see Fig. 16.15). During this test, the
requirements of §25.149(d) stipulate that “During recovery, the airplane
may not assume any dangerous attitude or require exceptional piloting
skill, alertness, or strength to prevent a heading change of more than 20
degrees.” If during the testing the heading changes more than 20 deg, the
test speed is increased until the test passes. The speed increment is then
added to the static VMCA speed.
Figure 16.16 shows a typical sequence of events for a dynamic VMCA test
point, this one for a jet-driven airplane. Because this is a minimum control

L L

 V

.

.
 L

Fig. 16.15 Some variables to consider in a dynamic VMCA condition.

 CHAPTER 16 Takeoff Performance Testing 759

Engine
KIAS (kt) failure
140
KIAS (kt) 130
120
110
100
Pitch (deg) Flight path angle (deg) Elevator (deg)
30 10

Elevator (deg)
Pitch (deg)

5
FPA (deg)

20 0
10 –5
0 –10
–15
–10 –20
Heading (deg) Rudder (deg)
30
Heading (deg)

Rudder (deg)
–56
20
–60 10
–64 0
–68 –10
Roll (deg) Aileron (deg)

Aileron (deg)
30
Roll (deg)

5
0 20
–5 10
–10 0
–15 –10
10 15 20 25 30
Time (s)

Fig. 16.16 Dynamic VMCA test for jet-driven airplane.

case, the airplane will be at minimum test weight with maximum thrust being
produced, which will lead to larger flight path angles. The speed for the test
will be slow, which means a larger angle of attack. The combined larger flight
path angle and angle of attack will put the nose of the airplane high over the
horizon (large pitch angle). The pilot must rapidly lower the nose to ensure
the airplane does not stall. The nose-down pitching motion increases the
angle of attack. Any rolling motion that is generated as the airplane reacts
to the thrust loss (especially for propeller-driven airplanes, see Figs. 16.15,
16.17, and 16.12) will increase the chance of a wing stalling. The dynamic
VMCA test aims at verifying there is enough margin.
In Fig. 16.17, note how the angle of attack increases drastically once the
propeller-driven airplane has an engine failure. The risk of stall from one
wing is always present. Testing is always the preferred approach for
propeller-driven airplanes because the effects of the propeller transitioning
to the feather position (see Fig. 15.27 in Chapter 15) may be hard to model
as are the effects of rapid AoA change on both the pitching motion and
the loss of thrust from the propeller.
For jet-driven airplanes, many manufacturers (MFTC) have now started
performing analysis using a substantially flight test validated model in lieu
of testing. This is acceptable for twin-jet airplanes for which the engine
thrust has little impact on aerodynamics (such as tail-mounted engines) at
the VMCA conditions. The manufacturer can use this approach as long as
the modeling reveals no marginal condition. (A heading change of 18 deg
can be considered marginal.) This modeling approach improves flight
test safety.
760 Operational Aircraft Performance and Flight Test Practices

5000
4500
4000
Press Alt (ft)

3500

3000
2500 Engine failure
X = 00:00:37.6955
2000 X = 00:00:37.6000
115 1150
Y = 1000.06
110 950
KCAS NP1 NP2
105 750
KCAS (kt)

NP (RPM)
100 Static VMCA 550

95 350

90 150

85 –50
30
Pitch angle Angle of attack
25
Angle (deg)

20

15

10

5 Heading
5 X = 00:00:37.6000 345
Y = 332.12
0 340

Heading (deg)
Angle (deg)

–5 Test complete heading and sideslip stable 335

X = 00.01:1:44.5267
–10 Angle of sideslip 330
Rudder angle
–15 Magnetic heading 325

–20 320
–10 Heading
X = 00:01:23: 1990
5
Y = 321.65
Angle (deg)

0
Roll angle Aileron position
–5
–10
–15
–20
00:00:00 00:00:30 00:01:00 00:01:30 00:02:00
Time (hh:mm:ss)

Fig. 16.17 A dynamic VMCA test on a propeller-driven airplane.

VMCA Test Risks and Mitigations

The main risks of executing a minimum control speed test include:
• Low altitude with OEI
• Testing requires aiming for at least 95% of expected maximum thrust.
• Thermodynamically, engines’ maximum thrust is at sea level (or even
below).
• May need overboost if the engine can allow it.
• Loss of the remaining engine on a twin airplane
• Speed near stall warning and stall speed
• Risk of departure from controlled flight
 CHAPTER 16 Takeoff Performance Testing 761

The mitigations for this testing include:

• Consider performing the test at idle instead of OEI for
twin-engine airplanes.
• Agencies have allowed OEI to be replaced by idle on one engine under
specific conditions and after a full examination of the impact of the OEI
condition on other systems like the flight controls.
• Even if one intends to perform final VMCA testing with the engine
shut down, initial investigation with the engine at idle will reduce the risk
and provide low-speed, large-thrust asymmetry handling practice for
the crew.
• Overboost the engine if possible (work with the engine manufacturer) to
allow for higher test altitude.
• Allow enough altitude to relight the remaining engine.
• For twin-engine airplanes, the crew needs to be familiar with relight
procedures and the steps to protect airspeed (no stall) while executing the
procedure at low altitude.
• If installed, have the auxiliary power unit (APU) working to help relight
the engine.
• On an airplane with three or more engines, set the power on the nontest
engines to ensure continued flight following the loss of the test engine.
• Testing for VMCA should be completed after the testing for stall
characteristics is completed.
• Do not test at speeds below stall warning.

Final Note on VMCA

With the advent of electric engines and the greater ease in properly sizing
them (compared to the more complicated and expensive gas turbine engines,
for example), what if on a four-engine airplane the two outboard engines had
a smaller power rating? This approach would allow a spanwise variation in
power and a reduction of the impact of an engine failure on VMCA . One
should now consider this as part of the trade study.

Minimum Unstick Speed VMU

VMU testing is a challenging event for all manufacturers, and it results in
impressive takeoff attitudes. Most manufacturers do not hesitate to share pic-
tures and videos of this testing; an Internet search for VMU or minimum
unstick speed followed by a manufacturer’s name will result in multiple
hits. Minimum unstick speed is the lowest safe takeoff speed demonstrated.
As a reminder (underline added by author):
§25.107 (d) VMU is the calibrated airspeed at and above which the airplane
can safely lift off the ground, and continue the takeoff. VMU speeds must
be selected by the applicant throughout the range of thrust-to-weight
ratios to be certificated. These speeds may be established from free air
data if these data are verified by ground takeoff tests.
762 Operational Aircraft Performance and Flight Test Practices

CL When preparing for VMU testing, the

3 to 5 deg OEM must consider the airplane lift
curve and the impact of ground proxi-
mity on the airplane’s lift capability and
IGE OGE on the horizontal tail’s ability to generate
the proper pitching moment. Experience
MLG extended and analysis have shown that the airplane
Tail contact

lift capability at liftoff (Fig. 16.18) will

have the following:
• Increased lift coefficient for a given
 angle of attack
• Lower stall angle of attack [3–5 deg
lower in ground effects (IGE)
Fig. 16.18 Lift in ground effects. compared to out of ground effects
(OGE), with no propeller wash effects]
• Possibly lower CLmax
The applicant can therefore select a pitch attitude that is limited by tail
contact with the ground (geometry limit) or, if the angle at which the tail
would contact the ground (with landing gear fully extended and just touching
the runway) is above the stall angle of attack IGE, a smaller angle that would
just allow the use of a VR that would give the minimum V2 [per §25.107(e)].
The tail does not need to touch the ground for VMU testing. The aim of the test
is to demonstrate that the pitch attitude selected by the manufacturer results
in safe liftoff and climb away from the ground. If the manufacturer decides to
use a lower angle, such as one that would ensure no stall in ground effect (see
Fig. 16.19), then the takeoff speed will simply be based on this selected
pitch attitude.
A good starting point to estimate the critical angle of attack IGE is to start
with the OGE lift curve slope up to stall angle of attack, which will be vali-
dated by flight test early in the flight test program, and then remove some
capability (3–5 deg based on previous program experience, adjusted with
wind tunnel or computational fluid dynamics analysis) to estimate the capa-
bility of the plane at liftoff. If the manufacturer determines that the airplane
is geometry limited, then the OEM will expect tail contact with the runway

VMU Stall Tail

warn strike

Liftoff

Fig. 16.19 Limiting pitch angle during VMU testing.

 CHAPTER 16 Takeoff Performance Testing 763

and must prepare accordingly with the addition of tail protection systems
(something that will absorb not only some of the vertical load due to the
pitch rate, but also some of the friction loads when the tail is in contact with
the runway).
The next step is to plan the test scope. Typically, all takeoff flaps are
down, but an experienced OEM that has performed multiple VMU programs
can elect to offer as a means of compliance to test every other flap (e.g., Flaps
1, 3, and 5). This is a reasonable approach when the test conditions are
bounded with testing of least deflected flap and most deflected flap as well
as when the configuration selected by the OEM does not deviate
significantly from the company’s experience. As part of the test scope defi-
nition, the guidance material of FAA Advisory Circular AC 25-7D offers
the following:

(i) An applicant should comply with § 25.107(d) by conducting minimum

unstick speed (VMU ) determination tests with all engines operating and
with one engine inoperative. . . . During this demonstration, the takeoff
should be continued until the airplane is out of ground effect. The
airplane pitch attitude should not be decreased after liftoff.
(v) To conduct the VMU tests, rotate the airplane as necessary to achieve
the VMU attitude. . . . VMU is the speed at which the weight of the
airplane is completely supported by aerodynamic lift and thrust forces.
. . . Determining the liftoff point from gear loads and wheel speeds has
been found acceptable in past programs. After liftoff, the airplane should
be flown out of ground effect. During liftoff and the subsequent
climbout, the airplane should be fully controllable.

Figure 16.20 shows a test executed per this guidance. The airplane is first
accelerated to a predetermined airspeed (about 100 kt for this example), at
which point the pilot applies essentially maximum elevator input to start
the rotation. There is initially almost no pitch change other than possibly
nose landing gear (NLG) extension. As the airplane continues to accelerate,
the pilot will then feel a pitch rate developing, and the elevator control input
will be reduced to aim for a target pitch attitude (in this example, approxi-
mately the angle for tail contact). At this thrust setting and pitch attitude,
the airplane continues to accelerate at about constant pitch attitude, and
some wheel speed modulation can be seen as the main landing gear
(MLG) starts extending and the airplane bounces some before finally
lifting off completely. The VMU condition for this example is the point
when the last MLG lifts off the ground. After liftoff, the pitch attitude is main-
tained until the airplane is OGE (about one wingspan).
During the testing, the crew should comment on the general handling of
the airplane and note whether there are any tendencies to pitch and roll
beyond the command of the pilot. (Wing drops are an indication of local
wing stall.) Testing should be done at a lesser angle than the test final
target pitch attitude to allow the crew to become familiar with the high
pitch attitude handling qualities. During a given test, the test team should
764 Operational Aircraft Performance and Flight Test Practices

Elevator position
–0
–5

Elevator input
–10
–15
–20
–25

15 Pitch attitude (deg)

10

5 Pitch increase
as ldg gear
0 struts extend

–5
Wheel speed (KTGS) KTAS
160
140
120
100
Liftoff

80
60
RADALT (ft)
35
30
25
Tail Contact

20
15
10
5
0

Fig. 16.20 Geometry-limited VMU test.

monitor wind speed and wind direction (see Fig. 16.21). Ideally, the wind
should be near zero; however, in practice up to 5 kt of crosswind and 10 kt
of headwind might be tolerated, provided it is stable enough to ensure safe
testing. Note that crosswinds will expose the wing to asymmetric flow (side-
slip angle), and if the airplane is near stall angle of attack in ground effects,
one wing could stall first (as seen by the tendency for wing drops near the
liftoff point).
Note that the angle of attack source (often from a boom) may not be
reliable during rotation due to induced AoA during rotation or sensitivity
 CHAPTER 16 Takeoff Performance Testing 765

Elevator (deg) Pitch (deg)

10
Maintain pitch
Elevator (deg)

13.30512
Pitch (deg)

0
–13.76013
–10

–20

Wheel speed (kt) TAS (kt)

150
140 118.4363
Wheel speed (kt)

130
TAS (kt)

120
110
100
90
109.6716
80
70

Pressure altitude (ft) FPA (deg)

1600 6
Pressure altitude (ft)

1550 1315.579 5

FPA (deg)
1500 4
1450 –0.07533848 3
1400 2
1350 1
1300 0
.6 off
02

40 45 55 60 65 70
50 Lift
14

Time (s)

Wind speed (kt) Wind direction (deg)

180
14

Wind direction (deg)

12 155.4247 170
Wind speed (kt)

10
160
8
6 150
4
Keep an eye on winds 140
2
3.453544
0 130

CAS (kt) AoA (deg)

150 20
140 13.36115

130 15
AoA (deg)
CAS (kt)

120
110 10
100
114.7881 5
90
80
0

Roll (deg) AoSS (deg)

4
0.683864
2
AoSS (deg)
Roll (deg)

–2
Check for roll activity
0.1125289
–4
40 45 50 55 60 65 70
Time (s)

Fig. 16.21 Data of interest during VMU testing.

766 Operational Aircraft Performance and Flight Test Practices

to wind change and crosswind. The pitch indicator is often a more reliable
source during this dynamic maneuver, so mathematically the AoA at liftoff
is equal to the pitch angle at this same condition. Note that the flight path
angle sensor may also not be reliable prior to liftoff (see Fig. 16.21) and
may provide a false indication of positive angle prior to the liftoff of the
last wheel.
Testing for one-flap configuration should be performed at several weights
and thrust-to-weight (T/W) conditions as recommended by the guidance
and then repeated for other flaps. Part of the thrust will produce a lifting
component, so there will be a natural test scatter of liftoff angles with the
higher T/W resulting in lower angles and the near-weight-altitude-
temperature (WAT) limit (minimum OEI thrust equivalent) producing the
larger angles, possibly limited by tail contact.
The aim of the test is to be able to model the liftoff capability of the plane
under varying thrust conditions. The lift coefficient of the airplane at liftoff
can be expressed as
W
CLLO ¼ (16:9)
1
r sV2 S
2 SL LO
This lift coefficient includes the net aerodynamic lift from the airplane
(wing lift minus tail download) and a thrust component (see Fig. 16.22),
where the thrust component acts at a slightly different angle (engine inci-
dence iT ) than the airplane’s angle of attack. We can remove this thrust com-
ponent to find the minimum unstick lift coefficient

W  T sinðaMU þ iT Þ
CLMU ¼ (16:10)
1 2
r s VLO S
2 SL
Note that Eq. (16.10) does not remove the influence of the engine
thrust on the lift; it simply removes a vertical force due to the thrust
produced.
When performing the data reduction, one should always keep an eye on
the relation between the IGE-derived lift curve (tested to a target pitch

iT L
T

MU

Fig. 16.22 VMU analysis.

 CHAPTER 16 Takeoff Performance Testing 767

attitude/AoA at liftoff) and the OGE CLMU

lift curve. It is often useful to plot 3 to 5 deg
the data together as a reminder (see
Fig. 16.23). Testing for VMU verifies
at LO
that CLmaxjIGE is not lower than OGE
the expected (demonstrated) liftoff
angles. In the end, the model created

VMU test limit

will be based on the demonstrated
ability of the airplane to safely lift off
and fly away from the ground. One
can also create a VMU model as a
function of OGE stall speed ratio, 
but that model should be based on
the most adverse thrust condition
(i.e., zero thrust). Fig. 16.23 Curve fitting test results.

Example 16.1
You perform a VMU test with a target pitch attitude of 13 deg and a relatively
high thrust-to-weight ratio. The pilot initiates rotation at approximately 105
KCAS (see Fig. 16.24). Due to the high T/W, the airplane lifts off at a pitch
attitude (8.58) lower than the target as seen by the wheel speed spin down.
The estimated T/W at the time of liftoff was 0.278, wing loading was
64.95 psf, and engine incidence was þ2 deg. Determine the CLMU (no
thrust) for this test point and compare it to the OGE lift curve provided.

150 2.2
Airspeed

100
2
IGE zero
50
150
1.8 thrust
Wheel Spd

100
CL
50 1.6
0 LO = 1.502
20

10 1.4 MU = 1.426
AoA

OGE
0
1.2 zero thrust
–10
20

10
1
Pitch

Pitch/AoA = 8.5 deg

0

–10 0.8
110 112 114 116 118 120 122 124 126 128 130 0 5 10 15 20 25 30

Fig. 16.24 VMU test example.

(Continued)
768 Operational Aircraft Performance and Flight Test Practices

Example 16.1 (Continued)

Solution: From the test data, we determine that the airspeed was approxi-
mately 113 KCAS at the time of liftoff. Using the approximation that KCAS 
KEAS at low altitudes and low airspeeds, we can compute a total lift coefficient
at liftoff of
ðW =S Þ
CLLO ¼ ¼ 1:502
1 2
rSL VEAS
2

We know there is a thrust component to this lift coefficient, and it needs to

be removed to compute the CLMU .
 
T
sinðaLO þ iT Þ
ðW =S Þ W
CLMU ¼    ¼ 1:426
1 2 1 2 1
rSL VEAS r V
2 2 SL EAS W =S

VMU is, by its very nature, a high-risk test. Careful preparation and pretest
analysis including wind tunnel will ensure that the test team is approaching
the test with all the knowledge possible to mitigate the risks. The main risks
for performing such a test include:
• Tail strike (almost a given)
• Inability to lift off and/or accelerate to V2
• Stall/roll off on takeoff
• Stick pusher (if installed) activation on liftoff
• Engine problems due to higher than normal angle of attack and
ground effects
• Runway tracking (high pitch attitude, loss of runway sight, crosswind)
• Overrotation
Common mitigations for this type of testing include:
• Use a tail bumper.
• Use a runway of sufficient length.
• Keep landing gear down after liftoff.
• Pilot is prepared to lower the pitch attitude on liftoff if there are signs of stall
(e.g., shaker activation, wing drop, buffeting).
• Minimize crosswinds.
• Target fix pitch attitude for maneuver.
• Pilot is prepared to increase thrust to maximum available if a reduced thrust
takeoff is performed.
• Testing is performed with symmetric thrust (simulate OEI with lower AEO
thrust).

(Continued)
 CHAPTER 16 Takeoff Performance Testing 769

Example 16.1 (Continued)

The guidance material recognizes the high risk involved with performing such
a test. The airplane essentially is stalled for most of the maneuver prior to
takeoff, so one should not add asymmetric conditions (asymmetric thrust
and flight controls deflections) to the maneuver while data are being collected
to define the safety model.
The wording of FAA AC 25-7D includes:
In lieu of conducting one-engine-inoperative VMU tests, the applicant
may conduct all-engines-operating VMU tests if all pertinent factors
that would be associated with an actual one-engine-inoperative
VMU test are simulated or otherwise taken into account. To take
into account all pertinent factors, it may be necessary to adjust the
resulting VMU test values analytically. The factors to be accounted
for should include at least the following:
• Thrust/weight ratio for the one-engine-inoperative range.
• Controllability (may be related to one-engine-inoperative free
air tests, such as minimum control speed in the air (VMCA )).
• Increased drag due to use of lateral/directional control systems.
• Reduced lift due to use of devices such as wing spoilers for
lateral control.
• Adverse effects of use of any other systems or devices on
control, drag, or lift.

Ground Run Distance Modeling

Following all the preliminary testing identified in the previous section
and the validation of the speeds and procedures, the airplane manufacturer
now turns its attention to takeoff distance measurement and validation. We
focus our discussion on continued takeoff with either AEO or OEI.
Figure 16.25 shows a typical AEO takeoff for a tricycle landing gear con-
figuration using an NLG. The pilot will first set maximum thrust with
brakes applied [thrust lever angle (TLA) moved, here increased load on
NLG recorded]. When the engines reach the proper thrust setting, the
brakes are released [note the recorded flight path angle (FPA) oscillation
in Fig. 16.25 and the reduction of the NLG load] and the airplane acceler-
ates, NLG on ground, until reaching VR . During the acceleration, the
NLG and MLG will see some load reduction (reduced compression
recorded in Fig. 16.25) as the lift increases. At VR , the pilot pulls on the
control column to initiate rotation (elevator input). The pilot adjusts the ele-
vator input to achieve a target pitch rate and pitch attitude. The airplane lifts
off the ground at the last MLG contact (seen as a reduction in wheel speed).
The takeoff distance test is over at 35 ft AGL. The test maneuver can be ter-
minated at a safe altitude (typically 400 ft AGL), at which point the crew can
simply concentrate on flying the airplane.
770 Operational Aircraft Performance and Flight Test Practices

KIAS (kt) Wheel Spd (kt) Elevator (deg) VR

200 5

150 0
Wheel Spd (kt)

Elevator (deg)
KIAS (kt)

100 –5

50 –10

0 –15
MLG Shock Strut (in)
NLG Shock Strut (in)

MLG Shock Strut (in) NLG Shock Strut (in) Pitch (deg) FPA (deg)
20

Pitch (deg) FPA (deg)

14 Target
12 pitch
15
10 10
8 5
6
0
4
2 –5
0 –10
10 20 30 Time (s) 40 50 60
TLA Brake NLG MLG
moved release LOF
LOF

Fig. 16.25 Typical takeoff sequence.

We start the analysis of the takeoff distance modeling with the ground
run segment. For convenience, we rewrite Eq. (15.14) here:
1  
T  mr W  rSL s V 2 S CD,TO  mr CL,TO  W sinðuÞ
2
W dV
¼ ðV  VW Þ (16:11)
g dX
We can then break it down into segments to focus the discussion on test
preparation and execution.

Runway Survey
The local runway slope adds a force to the takeoff equation that is pro-
portional to the weight times the sine of the local slope [W sin(u)]; that is,
now, part of the aircraft’s weight will act along the runway (see Fig. 16.26).

D T
f

W
N

Fig. 16.26 Local runway slope effect on acceleration forces.

 CHAPTER 16 Takeoff Performance Testing 771

Prior to performing takeoff distance testing, one should perform a

runway survey. This can be done with traditional survey means where the
runway elevation is mapped and provided to the test team prior to testing,
or a more rapid method can be used where the airplane’s onboard DGPS is
used to map the runway. The general approach (see Fig. 16.27) is to be
able to reliably predict the impact of the local slope (slope of the runway at
a given distance along the runway) on the airplane’s acceleration during
the test.
The procedure to model the runway can use the airplane’s onboard DGPS
can be as follows:
1. Align the airplane on the runway centerline at the beginning of the runway
and apply the brake (zero ground speed). This point should be considered
the beginning of the takeoff distance for runway testing.
2. After the point is recorded, release the brakes and taxi the airplane along
the centerline of the runway at low speed (10 to 20 kt) to the nearest
1000-ft runway marker. The pilot can then event (trigger a signal in the
data recording system to identify condition) the position or stop at
the marker.
3. For the follow-on segment, taxi the airplane at low speed still, but this time
the pilot will deviate one block to the left of the centerline (runway
concrete blocks are typically 25 ft in width), return to the centerline, and
then deviate right one block, before returning to the centerline and
stopping at the next 1000-ft marker.
4. The pilot can then continue to the end of the runway, still at low speed and
along the centerline, and continue to event at every 1000-ft marker. At the
end of the runway, the pilot should come to a complete stop on the point
considered the furthest point along the runway for testing (which can be
the start of the runway should the testing proceed along the opposite
heading).
Elev Elev Elev Elev
6581’ 6565’ 6421’ 6368’

05
213’

23
0% 9% 18.5% 12.5%
105 m 55 m
238 m 137 m
535 m
1755’

Fig. 16.27 Runway survey, Courchevel, France.

772 Operational Aircraft Performance and Flight Test Practices

3605

3604
Smoothed
GPS height above reference SL (ft)

3603

3602

3601 As recorded

3600

3599

3598

3597

3596

3595
0 1000 2000 3000 4000 5000 6000 7000 8000
Distance along the runway (ft)

Fig. 16.28 DGPS recorded height during taxi runway survey.

This procedure will allow the tester to collect data on the geometric
height of the runway along runway segments and derive a slope that can
be applied. As well, it will help validate the general data reduction mathemat-
ical models for runway heading, lateral deviation, longitudinal position, and
ground speed. It should be remembered that the precision of the crew with
the event/stop or event and the precision of the location of the runway
markers may lead to distances between pilot events not being exactly
1000 ft . It is not the purpose of this method to validate the data within
1 ft , but rather to validate the general data reduction method to be discussed
next. An example of the output of such a test is shown in Fig. 16.28.
Note the scale of the GPS height in Fig. 16.28; the runway average slope is –
0.064 deg. The DGPS signal naturally drifts a little; as the signal is captured by
an antenna on the airplane, it is subject to the oscillation of the fuselage during
the ground run. For the data to be useful for runway testing data reduction, one
will need to smooth out the collected data and possibly even segment the data
(average slope over a given distance along the runway). There is no perfect
approach; the user should review the data collected and use what best helps
the data reduction for the takeoff distance.

DGPS Runway Survey Analysis Method

The Global Positioning System (GPS) uses a model of the Earth (WSG-84
Earth Gravitational Model) to provide position information (longitude, lati-
tude, height above mean sea level). This is an Earth-centered, Earth-fixed
 CHAPTER 16 Takeoff Performance Testing 773

(ECEF) X, Y, Z model (see Fig. 16.29). For Pole

that model, the Earth is not perfectly
round, and the major axes are of the fol-
r b
lowing dimensions:
a ¼ 6,378,137.0 miles ¼ N
20,926,667.497 ft  a
b ¼ 6,356,752.314 miles ¼
20,856,504.342 ft
Equator

The flatness factor f of this model is

ab 1
f ¼ ¼ (16:12)
a 298:257223563
The local radius of curvature N in the Fig. 16.29 Earth model.
prime vertical (constant longitude) at a
given latitude can be written as
a
N ðfÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð16:13Þ
1  e2 sin2 ðfÞ
where:
f ¼ latitude
e ¼ eccentricity of the Earth

The eccentricity is defined as

e2 ¼ 2 f  f 2 (16:14)
When dealing with small distances, such as the takeoff distance, one can
simplify the analysis to a quasi-two-dimensional flat Earth condition. One
first sets the latitude equal to the midpoint of the runway and computes
the local Earth radius using Eq. (16.13). Then, one can compute the
change in distance along the north–south axis (constant longitude) as
Dy ¼ N ðfÞ sinðDfÞ (16:15)
where:
Dy ¼ change in distance north–south
Df ¼ change in latitude

The same general approach can be repeated for the change in distance
along the east–west axis. For this, we first compute the Earth’s minor
radius r for the midrunway point (latitude).
r ¼ N ðfÞ cosðfÞ (16:16)
774 Operational Aircraft Performance and Flight Test Practices

This then allows us to compute the change in distance east–west Dx as a

function of change in longitude Dl
Dx ¼ r sinðlÞ (16:17)
The position of the airplane along the runway can be computed from the
ECEF Cartesian coordinates in the following way:
x ¼ ðN þ hÞ cosðfÞ cosðlÞ
(16:18)
y ¼ ðN þ hÞ cosðfÞ sinðlÞ
where h is the height of the airplane above local sea level (per GPS model), a
value provided by the GPS unit. Note that h was not accounted for in Eqs.
(16.16) and (16.17) because the small altitude range of test runways (typically
less than 10,000 ft above mean sea level) is extremely small compared to the
Earth’s radius, and the overall impact on distance is less than 0.05%. In prac-
tice, it could also be dropped from Eq. (16.18).
Using these coordinates and the DGPS runway survey with the airplane
procedure described earlier, one can then derive the runway heading with
respect to true north as being equal to
 
1 y2  y1
cgps ¼ tan (16:19)
x2  x1
where the start point (x1 , y1 ) corresponds to the beginning of the runway and
the end point (x2 , y2 ) is the end of the runway as collected during the taxi test
(see Fig. 16.30). Note that the y-coordinates are often referred to as northing
coordinates (along the north–south axis) and the x-coordinates as easting
(along the east–west axis).
The distance d along the runway centerline can be computed as the dis-
tance from the starting point (x1 , y1 )
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
y (northing) d ¼ ðx  x1 Þ2 þ ðy  y1 Þ2
(x2, y2) (16:20)
da
where (x, y) is the position of
the airplane on the runway at
d a given time. Equation (16.20)
can then be checked against
(x1, y1) the 1000-ft markers collected
during the runway survey taxi.
We can also develop the
x (easting)
equations for computing the
airplane’s lateral deviation
Fig. 16.30 Easting–northing coordinate system, from the runway centerline.
distance along (d ) and across (da ) the runway. The centerline, as defined by a
 CHAPTER 16 Takeoff Performance Testing 775

straight line between (x1 , y1 ) and (x2 , y2 ), can be represented by an equation

of the following form:
y¼mxþb (16:21)
where
y2  y1
m¼ (16:22)
x2  x1
and
x 2 y1  x 1 y2
b¼ (16:23)
x2  x1
Then, the distance da of point (x, y) 90 deg away from the centerline can
be computed as follows:
ymxb
da ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (16:24)
m2 þ 1
where the sign convention for the deviation is positive to the right of the
centerline. Again, the data collected during the taxi runway survey can be
used as a check of the data and the implementation of the equations.

Modeling Excess Thrust During Takeoff Ground Run

Understanding the variation of thrust with airspeed (the thrust lapse rate)
during a takeoff is a critical factor in creating an acceptable takeoff distance
model. A typical thrust variation on Part 25 airplanes is one where the thrust
decays with increasing airspeed (see Fig. 16.31). This lapse rate can be seen
when one plots the airplane acceleration during takeoff vs the airplane air-
speed. Looking at the equation for the airplane’s acceleration, Eq. (16.25),

AEO Takeoff accel 40,000

Nx (g) VR VLOF 35,000
0.5 T
30,000
0.4
25,000
Rotation speed
Foreces (lb)

Liftoff

0.3 20,000
Nx (g)

Excess
15,000 Thrust
0.2
10,000
0.1 (f + D)
5000

0.0 0 f
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
Airspeed (kt)
KTGS (kt)

Fig. 16.31 Verifying ground acceleration against expectations, jet-driven airplane.

776 Operational Aircraft Performance and Flight Test Practices

one sees
1 2  
g T  mr W  W sinðuÞ  rSL s VTAS S CDTO  mr CLTO
2
a¼ ¼ g Nx
W
(16:25)
There are many parameters that influence the acceleration; one way to
extract the thrust model out of Eq. (16.25) is to test one flap configuration
(fixing CDTO and CLTO ) and performing takeoffs at multiple weights at the
same airport (same general pressure altitude). Testing can also be done at
multiple thrust settings (mix of weight variation and thrust/weight for the
same expected drag and ground rolling friction coefficients).
For propeller-driven airplanes, one often sees a plateau in the acceleration
in the lower speed region (below about Mach 0.10) where the propeller thrust
is sorting itself out from static (no forward speed) condition to in-flight con-
dition (where the thrust becomes more or less proportional to 1/V 2 ). One
assumption this author likes to use is to assume the propeller has established
forward flight conditions by Mach 0.10 and that the thrust provided at Mach
numbers less than 0.10 is equal to the thrust model at Mach 0.10 (see
Fig. 16.32). This simplified model must be validated by flight test results,
but it does provide a good starting point.

Rolling Friction Force Coefficient

The rolling friction force coefficient mr has a relatively small impact on
acceleration [Eq. (16.25)], with a typical value on asphalt or concrete

Nx – Fus (g) Mach M = 0.1

0.4 0.20
Power
set to TO

0.3 0.15
Nx – Fus (g)

Mach

0.2 0.10

0.1 0.05

0.0 0.00
90 95 100 105 110 115 120
Time (s)

Fig. 16.32 Ground run acceleration modeling for propeller-driven airplane.

 CHAPTER 16 Takeoff Performance Testing 777

runways of about 0.02 (2% of the airplane weight on wheels). On such sur-
faces, tires with higher tire pressures may have a lower value (0.018), and
the tires with lower pressures a slightly larger value (0.022). Under most
testing conditions, this small adjustment (0.001 to 0.002 change in mr ) in
rolling friction force may be lost in the scatter of the testing.
If one simplifies the acceleration equation to a more basic format [we
rewrite Eq. (15.15) from Chapter 15 here], we can see that the impact of
an adjustment on the rolling friction force coefficient will be greater when
testing at low thrust-to-weight conditions.
T
ag  mr  sinðuÞ (16:26)
W
Under low T/W takeoff from brake release to VR , the runway distance
required will be very long. So when one looks at the fitting of the test data
against the model results for the same atmospheric conditions and airplane
configuration (see Fig. 16.33), one can perceive the need to adjust mr if the
data deviate from the best fit curve at the longer distances while being a
good fit at the smaller distances (higher T/W).

Wind Considerations
Winds can cause havoc in the creation of a takeoff model. Winds, even if
steady (not gusty), will rarely be of constant magnitude and direction along
the entire length of the takeoff; it is therefore suggested to minimize winds
during the testing. Ideally, winds should be less than 5 kt with no crosswind
component. (Crosswind introduces sideslip and increased drag.) In practice,
because one does need to execute the testing in a reasonable timeframe,
winds of up to 10 kt along the runway and 5 kt across will produce acceptable
scatter in the test results.
Guidance material such as FAA AC 25-7D suggests: “Wind Limits.
For takeoff and landing tests, a wind velocity limit of 10 knots (from
any direction) or 0.11VSR1 (whichever is lower) at the height of the
mean-aerodynamic-chord (MAC), as determined with the airplane in a
static ground attitude, has been considered the maximum acceptable.” For
smaller planes than typical Part 25 airplanes, one looks to the guidance
material of FAA AC 23-8C (underline added by author):
Wind (takeoff and, landing tests, not for crosswind component testing): As
low as possible, but not to exceed approximately 12 percent VS1 or 10
knots, whichever is lower, along the runway measured at a height of 6 feet
above the runway surface. At higher wind velocities, the data may be unreli-
able due to wind variations and unsmooth flight conditions.

The minimum 6 ft height above ground is required because wind measured

closer to the ground is not considered reliable.
A dedicated and calibrated weather station with continuously recorded
parameters such as wind speed and direction, temperature, and static
778 Operational Aircraft Performance and Flight Test Practices

5000

Conditions where
4500 r has most impact

4000

3500

3000
Model distance (ft)

2500

2000

1500

1000
+/– 200 ft
tolerance
500

0
0 1000 2000 3000 4000 5000
Test distance (ft)

Fig. 16.33 Comparing model vs test distance for mr .

pressure that can be located at the midpoint of the takeoff run is highly rec-
ommended because it will help with the data reduction. This weather station
can guide the test team during the execution and allow them to best make a
decision to continue or stop testing.
§25.21 Proof of compliance

(d) Parameters critical for the test being conducted, such as weight, loading
(center of gravity and inertia), airspeed, power, and wind, must be
maintained within acceptable tolerances of the critical values during
flight testing.
(f) In meeting the requirements of §§ 25.105(d), 25.125, 25.233 and 25.237,
the wind velocity must be measured at a height of 10 meters above the
surface, or corrected for the difference between the height at which the
wind velocity is measured and the 10-meter height.
 CHAPTER 16 Takeoff Performance Testing 779

45

40

35

30
Height (ft)

25

20

15
Height of measurement
10

5
Wind speed not reliable
0
0 2 4 6 8 10 12 14
Wind speed (kt)

Fig. 16.34 Wind profile using power of 1/7; 10-kt wind measured at 10 ft.

One acceptable way to convert winds from recorded height to the 10-m
height (32.8 ft) is to use the following equation:
 1=7
h2
Vw2 ¼ Vw1 (16:27)
h1
This approach assumes a given wind profile close to the ground (see
Fig. 16.34). If the winds are kept low during the testing, the difference in wind
computed using Eq. (16.27) and the actual will be small within the region of
interest for takeoff (height below 35 ft) and well within normal test scatter.

Takeoff Ground Run Testing Procedure

The ground run part of the takeoff usually is the one that results in the best
fit for the performance model because there is little variability (the wind being
the largest one). For best results, the test procedure should require:
1. Airplane starts from zero ground speed and engine thrust set to the target
for the test.
2. Airplane accelerates to target VR + 2 KIAS.
3. Pilot pitch input should be distinct (near step input to provide distinct
feedback of pilot input).
The data collected for the ground run testing can be plotted in a manner
similar to Fig. 16.33 to show that the applicant has collected sufficient data to
780 Operational Aircraft Performance and Flight Test Practices

validate the model. This can be important when one considers that manufac-
turers typically concentrate their runway performance testing at one airport
to ease logistics and reduce execution time.
Historically, the extrapolation of the takeoff data was limited to 6000 ft
above and 3000 ft below the altitude at which the data were collected. This
meant that an applicant, seeking to certify an aircraft for takeoff anywhere
from sea level to 10,000 ft pressure altitude, needed to test at altitudes
between 3000 and 4000 ft. We, as an industry, keep learning. One just needs to
look back at an old FAA advisory circular, now cancelled (AC 25-2, Extrapolation
of Takeoff and Landing Distance Data over a Range of Altitude for Turbine-
Powered Transport Airplanes, 1964) that said in the discussion segment:
Since the introduction of the transport category requirements, the takeoff and
landing distance performance of reciprocating engine airplanes has been
extrapolated from one set of basic flight test data obtained at one airport
elevation over an altitude range extending from sea level to 8,000 feet. The cal-
culation methods were standardized to a great extent, and many subsequent
years of operating experience on the older airplanes has indicated no particu-
lar grounds for doubting the soundness of this procedure. When approval for
operation at altitudes higher than 8,000 feet was desired, then additional tests
at a high altitude were conducted.
Upon the introduction of turbine-powered airplanes, the allowable range of
takeoff and landing distance performance extrapolation was reduced. The
reasons for the reduction were that the turbine-powered airplanes had
many new design features that were considerably different from those on pre-
vious type certificated reciprocating engine airplanes. The engine features that
were different included larger ambient air effects on maximum power, idling
power characteristics, engine acceleration and operating characteristics, etc.
The airplane features that were different included sweptwings, high lift
devices, control systems, wing spoilers, bogey landing gears, takeoff and
landing speeds, ground control and acceleration characteristics, etc. In the
early stages, the effects of these various engine and airplane features on per-
formance were not substantiated to the necessary degree of accuracy for the
turbine-powered airplanes.
The overall objective of CAR 4b.100(a) should be observed in applying per-
formance calculation methods. The stated objective is to allow the use of cal-
culation methods based on the minimum amount of direct flight testing that
will result in performance information accuracy that is equivalent to direct
flight test performance measurements.

The certification agencies have now recognized that experience has been
gained with performance modeling and that the drag models are well
behaved at all altitudes. They have also recognized that the soundness of
the extrapolation is mainly due to the modeling of the propulsion system.
AC 25-7D allows dedicated takeoff distance testing at any altitude provided
the applicant performs engine performance testing to within 3000 ft of the
minimum and maximum field altitudes.

VR to V35
The phase of takeoff from VR to V35 (reaching 35 ft AGL and a
specified airspeed) is a complex set of events involving acceleration,
 CHAPTER 16 Takeoff Performance Testing 781

Curved path
KCAS (kt) Wheel Spd (kt) RADALT (ft) Ground Speed (kt)
200 100
190 90
Ground Speed (kt)
180 80
Wheel Spd (kt)
170 70

Radalt (ft)
KCAS (kt)

160 60
150 50
140 40
130 30
120 20
110 10
100 0
Nx (g) Nz (g)
0.30 0.4
0.25 Nz 0.3
Nx
0.20 0.2

Nz (g)
Nx (g)

0.15 0.1
0.10 0.0
0.05 –0.1
0.00 –0.2
Pitch (deg) Pitch rate (deg/s) Elevator (deg)
25 0
Pitch rate (deg/s)

Elevator (deg)
20 –5
Pitch (deg)

15
–10
10
5 –15
0 –20
66 68 70 72 74 76 78
VR Time (s) VLOF V35

V V
t t

Fig. 16.35 Events from VR to V35 , AEO takeoff.

climb, and pitching motion, all executed over a relatively short time
period with risks of stall and tail contact with the runway. It involves
(most of the time) a pilot in the loop that must execute this maneuver
precisely following a clear set of instructions to ensure repeatability
and safety.
The goal of the performance engineer is to develop a set of takeoff pro-
cedures that will be easily accomplished by the crew and lead to a repeatable
takeoff distance while achieving the proper takeoff speeds. If one observes the
typical AEO takeoff of Fig. 16.35, one can see the behavior of the airplane
described in Fig. 15.11 in Chapter 15. We will break down this part of the
takeoff into two segments and see what drives each.

AEO: VR to VLOF
This part of the takeoff starts when the pilot applies a nose-up pitching
moment via the elevator at the manufacturer’s defined rotation speed VR .
As the elevator comes in, a pitch rate will develop (see Fig. 16.35). The air-
plane will lift off when the net lift produced plus the vertical thrust com-
ponent are equal to the airplane weight—Eq. (16.28) and Fig. 16.36. This is
achieved at a pitch attitude that is dependent on:
• Lift curve in ground effect.
• Thrust to weight of the airplane when the rotation is initiated.
• VR .
• Average pitch rate commanded by the pilot. The industry average is 3 deg/s.
782 Operational Aircraft Performance and Flight Test Practices

CLMU 3 to 5 deg

L
at LO
OGE
 = T

VMU test limit

T

Estimated pitch at liftoff Lt




Fig. 16.36 Dynamics of liftoff.

Note the difference between the average rotation rate (3.25 deg/s from
start of elevator input to reaching liftoff pitch attitude) and the peak rate
(5 deg/s) in Fig. 16.35. That rate is dependent on pilot input and having an
airplane properly trimmed for takeoff. The time from all gear on ground to
reaching liftoff pitch attitude will also depend on these variables. During
that time, the airplane airspeed also keeps increasing rapidly, being depen-
dent on the excess thrust at VR . [The drag will not increase much at this
stage with MLG still on the ground, and the acceleration will not reduce
much, Eq. (16.29)].

ðL  Lt Þ þ T sinðaT Þ ¼ W (16:28)

ðT  DÞ
DV  Dt (16:29)
W VR

To find the desired VR for the airplane, it is suggested the performance

engineer pick the desired zero-thrust liftoff attitude (which translates into
an angle of attack at liftoff and a given lift coefficient in ground effect) and
then work back to VR based on expected T/W at VR and the crew procedure
that dictates average pitch rate (see Fig. 16.37). The desired liftoff attitude
must provide robustness (see the “Abused Takeoff” section) to in-service
takeoff procedure deviations.
Of course, the selected VR needs to respect all certification requirements
(see Fig. 16.2 earlier in the chapter). If the VR selected from this approach
falls below a minimum value, it must then be set to this value. As well, this
approach does not go against the method proposed in Chapter 15 that
defines VR by starting from V2 and using a speed spread model. The
method used in this section provides a means of defining a desired tail
clearance (defined pitch attitude at liftoff) to define a minimum VR
(see Fig. 16.38). One must note that the minimum tail clearance may occur
after liftoff (see Fig. 16.39).
 CHAPTER 16 Takeoff Performance Testing 783

LO

Increasing pitch rate

Increasing T/W

VR

Fig. 16.37 Variables impacting liftoff pitch attitude.

AEO: VLOF to V35

The next segment is the acceleration from liftoff speed VLOF to the speed
that must be reached by the time the airplane reaches 35 ft above ground V35 .
Under AEO conditions, the airplane will have a fair amount of excess thrust,
and the pilot must manage the load factor increase to curve the flight path
and capture the proper speed by the time the airplane reaches 35 ft. While
OGE, the V35 has a minimum margin of 1.15 g to stall warning (30-deg
bank, as seen in Chapter 14) for OEI conditions and 1.3 g under AEO con-
ditions. So, for AEO takeoff, one should expect to see a load factor in the
1.15 to 1.25 g range (see Fig. 16.35), so as not to overspeed the airplane by
35 ft AGL and reduce the takeoff distance. During the maneuver, the
induced drag will increase (increasing load factor, airplane going out of
ground effects) and the lift coefficient will increase (increased load factor).
One should monitor the airplane angle of attack a during the maneuver
(see Fig. 16.40) and monitor for any signs of loss of lift (one wing dropping
in particular). The airplane will be moving OGE and there will be significant
changes in the airplane lift curve (see Fig. 16.41).

Pitch
rate

Low point

Fig. 16.38 Monitor tail clearance during takeoff testing.

 784
Curved path
KCAS (kt) Wheel spd (kt) Ground Spd (kt) RADALT (kt)

Operational Aircraft Performance and Flight Test Practices

150 60
140 50

Ground spd (kt)

Wheel spd (kt)
130 40

RADALT (kt)
KCAS (kt)
120 30
110 20
100 10
90 0
Pitch (deg) Pitch rate (deg/s) Elevator (deg/s)
12 0
10 –5
Pitch rate (deg/s)

Elevator (kt)
–10
Pitch (deg)

6
4 –15
2 –20
0 –25
–2
Tail clearance (in)
50
Tail clearance

40
30
(in)

20
10
VLOF
0
80 81 82 83 84 85 86 87 88 89 90
VR Time (s) V35

V V
t t

Fig. 16.39 Monitoring tail clearance during takeoff.

 CHAPTER 16 Takeoff Performance Testing 785

Curved path
KIAS (kt) Wheel spd RADALT (ft)
160
140
Wheel spd (kt) 150 120
100

RADALT (ft)
KIAS (kt)

140
80
130 60
40
120
20
110 0
Pitch (deg) FPA (deg)
15

10
Pitch (deg)

Pitch
FPA (deg)

5 
FPA
0

–5
40 41 42 43 44 45 46 47 48 49 50
VR VLOF Times (s) V35
V
t

Fig. 16.40 Monitoring AoA from liftoff to OGE.

IGE stall development (post-accident)

CL

Free air
Ground + 130 in.
Ground + 70 in.
Ground + 35 in.
Ground

8 10 12 14 16 18
Angle of attack (degrees)

Fig. 16.41 IGE lift coefficient change with AoA [1].

786 Operational Aircraft Performance and Flight Test Practices

During AEO testing, it must be verified that the airplane reaches the
target V35 (typically V2 þ 10) by 35 ft within a reasonable margin (typically
+2 KCAS) while the pilot follows the prescribed procedure. If the speed is
deviating more than reasonable, one will need to adjust either the takeoff dis-
tance model or the takeoff procedure (adjust pitch). Note that under the con-
dition of highest excess thrust, the airplane will most likely not have reached
the target pitch attitude when 35 ft is crossed, and the landing gear may not
even have been selected up by the crew.
We discussed in Chapter 15 that the minimum V2 (V2min ) for an
airplane is either 1.13VSR or higher. There is a provision in the regulations
to reduce this minimum value if there is a means to substantially improve
the airplane lift capability with power on that will not be meaningfully
impacted by an engine failure. For propeller-driven airplanes with more
than three engines, that margin can be reduced to 1.08VSR (VSR being with
zero thrust) because the application of power can substantially reduce the
stall speed (even with one engine inoperative) (see Fig. 16.11 earlier in the
chapter). This extra margin to zero thrust stall angle can be verified by
takeoff testing [and maneuver margin verification, §25.143(h)].
Other items that need monitoring during the takeoff are the pull forces
and the elevator travel during the takeoff. For normal takeoff, even at
the most adverse CG, the control forces should not be excessive, and the
elevator travel should not be limited. Figures 16.35, 16.39, and 16.42 show
examples of elevator travel in the takeoff phase; note that no elevator
bottomed out (large constant angle of deflection for several seconds). As
well, the pilot should not have to push on the stick to stop the nose from
increasing when the target pitch and airspeed are reached.
The testing should cover the entire weight range of the aircraft and the
T/W ratio range. Finally, during the testing, distance data are collected so
as to update the takeoff model. It is, after all, the aim of the testing.

OEI Testing
Note: Under the new Part 23
Testing for OEI condition is a necessary part of (Amendment 64 and up), an
certification testing because all takeoffs are acceptable means of
assumed to have a probable risk of full engine compliance for performance
is ASTM 3179 which states:
failure. That statement is true for reciprocating “Loss of Thurst—For
engines and gas turbines. In more recent years, elec- conventional aeroplanes
tric propulsion has been coming along, and some (reciprocating or turbine
multi–electric motor airplanes may actually benefit engine-powered), loss of
thrust means one engine
from the more “graceful” degradation of electric inoperative. For other
motors. (Multiphase motors may not lose all phases aeroplanes, the amount of
and can still provide some power.) But at the time thrust loss shall be proposed
of writing this chapter, the regulations had not by the applicant and
accepted by the CAA.”
totally caught up with this approach.
 Minimum
curved path
KCAS (kt) Wheel Spd (kt) Ground Speed (kt) RADALT (ft)
200 100

180 80

Ground Spd (kt)

Wheel Spd (kt)

RADALT (ft)
KCAS (kt) 60
160

140 40

120 20

100 0
Nx (g) Nz (g)
0.4 0.6
0.4
0.3
NZ 0.2

Nz (g)
0.2
Nx (g)

CHAPTER 16
0.1
–0.2
0 –0.4
NX
–0.1 –0.6
Pitch (deg) Pitch rate (deg/s) Elevator (deg)
12

Takeoff Performance Testing

Pitch rate (deg/s)

10 0

Elevator (deg)
Pitch (deg)

8
6 –5
4 –10
2
–15
0
–2 –20
70
Idle
80 90 VR VLOF 100 110
Time (s) V35
chop
V V
t t

Fig. 16.42 OEI takeoff following chop to idle.

787
788 Operational Aircraft Performance and Flight Test Practices

The objectives of a continued takeoff following the loss of one engine

are unchanged from those of the AEO takeoff. The pilot must initiate a
rotation at a prescribed VR , the same VR as for the AEO case
[§25.107(e)(2)] and reach a target pitch attitude (most of the time a
lower target than for AEO, unless the AEO is conservatively low). The
major difference between the two events is that the acceleration following
an engine failure will rapidly reduce, so the excess thrust available for the
continued takeoff will be much less (see the longitudinal acceleration Nx in
Figs. 16.42 and 16.43).
Before proceeding to OEI testing, a test group will typically validate
the baseline takeoff procedures and baseline takeoff model under AEO
conditions. When the team is confident they understand the airplane
enough, they will first proceed to investigate the OEI condition by
simulating a thrust loss by reducing one of the engines to idle thrust
(engine still available should it come to this condition; see Figs. 16.42
and 16.43). The test team must select the VEF speed (simulated engine
failure speed for the idle test); for the chop to idle, this VEF speed
may be artificially low to allow the engines to fully spool down prior
to reaching VR . This long exposure to a reduced thrust with its
reduced acceleration will lead to the use of a lot more runway then a
normal OEI test—the test in Fig. 16.42 had almost 19 s at reduced
thrust with a ground speed of over 260 ft/s, which represents a distance
of more than 4900 ft only for the chop to idle part of the test. The team
should plan accordingly by selecting a very long runway for this test. For
the OEI (idle or fuel cut) test, the target V35 speed must be V2 as defined
by §25.107.
If the engine failure occurs at a low speed (near VMCG or below, depend-
ing on the actual test point thrust), the rudder alone may not provide suffi-
cient directional control. If nose wheel steering is available, the system
should be on for OEI testing as a risk mitigation.
The loss of one engine, with a continued takeoff, will result in an
asymmetric thrust condition with its associated increased directional
control input from the pilot and the generation of a sideslip angle
(increased drag) and roll angle. If the rolling moment generated is
sufficient and the airplane is equipped with roll spoilers, the drag of
the airplane will increase as the spoilers deploy, further reducing the air-
plane acceleration capability (see the step down in Nx after liftoff in
Fig. 16.43).
The OEI testing (idle or fuel cut) allows the manufacturer to test the
airplane down to the minimum climb capability (WAT) case. Observation
of Eq. (16.30) shows that as the excess thrust in this phase of flight nears
minimum climb value (g35 , climb angle at 35 ft AGL as a reference), the
time to climb will increase rapidly. The height change dh for the
takeoff distance computation is fixed (35 ft), so one must minimize
the change in airspeed dV in the climb in order to minimize the distance
 Minimum
curved path
KCAS (kt) Wheel spd (kt) Ground speed (kt) RADALT (kt)
200 100
190 90

Ground speed (kt)

180 80

Wheel spd (kt)

170 70

RADALT (ft)
KCAS (kt)
160 60
150 50
140 40
130 30
120 20
110 10
100 0
Nx (g) Sideslip (deg)
0.4 4
Sideslip 3
0.3

Sideslip (deg)
2
1
0.2
Nx (g)

0
0.1 –1
–2

CHAPTER 16
0 –3
NX –4
–0.1 –5
Pitch (deg) FPA (deg)
12
10 Pitch

Takeoff Performance Testing

Pitch (deg)
FPA (deg)

8
6
4
2 FPA
0
–2
70 Idle chop 80 90 VR VLOF 100 110
Time (s) V35
V V
t t

Fig. 16.43 OEI takeoff following chop to idle, sideslip present, low FPA.

789
790 Operational Aircraft Performance and Flight Test Practices

traveled in this segment of the takeoff.

V V
dh þ dV dh þ dV
g g
DtLOF35 ¼  (16:30)
V ½ðT  DÞ=W  V g35

In Figs. 16.42 through 16.44, note how under the low thrust conditions
(these takeoffs representing near minimum climb capability for Part 25),
the speed increase beyond liftoff is essentially zero. The airplane reaches
the lower target pitch attitude before reaching 35 ft and maintains that
value. The airplane is now in a climb mode to reach 35 ft as fast as it can.
This approach is, of course, a compromise because the airplane will also
cover more distance on the ground prior to reaching VLOF . The performance
engineer must pick the right balance that ensures ease of takeoff execution
and minimum takeoff distance.
Once enough testing is done at idle, a few points may need to be
repeated with the engine off. FAA AC 25-7D stipulates that for jet-driven
airplanes, the speed spread should be “substantiated by at least a limited
number of fuel cuts at VEF .” It goes on to say: “For derivative programs
not involving a modification that would affect thrust decay characteristics,
demonstrations of fuel cuts may be unnecessary.” Figure 16.44 shows the
rapid decay in longitudinal acceleration Nx when the critical engine is cut
at VEF .
The critical “fuel cut” condition can be especially important for
propeller-driven airplanes because the testing will show how the distance
model is impacted by the way the propeller attains its final position
(ideally feathered) following a sudden engine failure. FAA AC 25-7D spe-
cifies: “The number of tests that should be conducted using fuel cuts
depends on the correlation obtained with the throttle chop data and sub-
stantiation that the data analysis methodology adequately models the
effects of a sudden engine failure.”
Testing for OEI takeoff, whether a throttle chop to idle or a fuel cut,
always involves increased risk. The flight test community sadly lost a test
crew following the accident of a Gulfstream G650 back in 2011—see the acci-
dent box “NTSB/AAR-12/02” in the “Takeoff Performance Testing: Risks
and Mitigations” section later in this chapter.
The largest obvious risk of performing fuel cut testing (one engine fully
inoperative) on a twin-engine airplane is obvious. Once the test engine is
cut and the aircrew has committed to continuing the takeoff (about 1 s
time lapse between the two events), then the airplane is at risk of an accident
should the remaining engine fail (ingestion of birds is always a risk) before
the airplane reaches a safe height, then an accident will follow with the
risk of losing the airplane and the crew. That minimum safe height for a jet-
driven airplane is often 2000 to 2500 ft above ground; at this point the
 Minimum
curved path
KCAS (kt) Wheel spd (kt) Ground Spd (kt) RADALT (kt)
180 100

160 80

Ground spd (kt)

Wheel spd (kt)

RADALT (ft)
KCAS (kt)
140 60

120 40

100 20

80 0
Nx (g) Nz (g)
0.3 0.3
0.2
0.2 NZ
0.1

Nz (g)
Nx (g)

CHAPTER 16
0.1 0
–0.1
0
NX –0.2
–0.1
Pitch (deg) Pitch rate (deg/s) Elevator (deg)

Takeoff Performance Testing

10 0
Pitch rate (deg/s)

Elevator (deg)
–5
Pitch (deg)

6
4
–10
2
0 –15
–2
–4 –20
55 60 VEF VR VLOF 65 V3570 75
Time (s)
V V

t t

Fig. 16.44 OEI takeoff following a fuel cut.

791
792 Operational Aircraft Performance and Flight Test Practices

crew would have enough time to relight the test engine following the loss of
the nontest engine.
The members of the MFTC have come forward to show the limits of
this approach to the authorities. It was highlighted that FAA Order
4040.26B, Flight Test Safety, specifies: “Accept no unnecessary risk. An
unnecessary risk is any risk that, if taken, will not contribute meaningfully
to the task.” The question becomes: Is the test condition really needed in
its present form (fuel cut) to provide the information required to certify
the airplane? Are there no other flight test conditions that could be com-
pleted to an adequate level of safety to collect the information required for
analysis? Can simulation (desktop or pilot in the loop simulator) be used
to show the same capability? The MFTC group requested that the flight
testing be subject to the same “single failure” rule as certified airplanes
where continued safe flight and landing is assured. Some members of
the council have adopted the approach of OEI continued takeoff testing
with only chop to idle and follow-on simulation with a model “substan-
tially validated by other flight test maneuvers” as stipulated in FAA AC
25-7D.

Speed Spread Model

A speed spread model (Fig. 16.45) is a good way to visualize the behavior
of the airplane during takeoff. As we have seen in Figs. 16.35 to 16.44, the

25
2.4% 2nd segment min
climb gradient

20
dV
dV (KTAS), dt (s)

15

10

dt
5

0
0 0.05 0.1 0.15 0.2 0.25
(T-D)/W – runway gradient
dt dV

Fig. 16.45 Adjusting speed spread model with flight test data.
 CHAPTER 16 Takeoff Performance Testing 793

acceleration of the airplane during the transition from VR to V35 is by no

means constant; it is highly dependent on pilot inputs (ability to follow the
takeoff procedure, speed capture, target pitch capture, pitch rate achieved).
But for each segment of the takeoff (VR to VLOF , VLOF to V35 , VR to V35 )
we can measure a speed increase and a delta time throughout the testing
over a large weight range and T/W range and see how both the airplane
characteristics and ease of flying the takeoff procedure are integrated. They
become the average acceleration (a ¼ DV/Dt) capability of the airplane.
This speed spread model can be configuration dependent or the same for
all flaps settings, whichever is the best way to model the test distance; it rep-
resents a pilot-in-the-loop takeoff distance. It also has the advantage of
clearly showing to the authorities that the model matches the test data
well. It is this author’s preferred tool to show compliance.
A good speed spread model should start from a generic one (or one from
a past program for a similar airplane) adjusted by analysis of pitch rate capa-
bility and target pitch capture prior to first flight to ensure good tail clearance
and margins to stall. Then, as data are gathered through the program, the
model should be adjusted in order to be as good as possible before the dedi-
cated takeoff distance testing, where it will then be validated.

Sources of Scatter in the Results

The sources of scatter in the test results are numerous, with the most
common discussed previously and repeated here as a reminder:
• Takeoff procedure, ease of use
• Pitch rate, actual vs target vs time
• Most direct impact on VLOF
• Pitch force (related to trim setting)
• Most direct impact on pitch rate
• Target pitch capture
• Most direct impact on V35
• Reaction time at VR
• Wind fluctuation (speed and direction)
This author experienced an example of significant scatter during the
testing of the Sino-Swearingen SJ30-2 (now the Syberjet SJ30) back in 2005
when the company testing was completed and the speed spread and distance
models were “validated.” When the FAA crew first came to validate the
company testing, the test team reported a much longer takeoff distance
than expected. Upon review of the data, it was shown that the crew was
adopting the proper target pitch attitude but at a much lower pitch rate,
which meant that the airplane airspeed was increasing beyond the target
V35 before reaching 35 ft AGL, thus more distance was covered to get to
35 ft AGL. The takeoff procedure was clarified, and testing was repeated to
show that the initial model developed was indeed valid.
794 Operational Aircraft Performance and Flight Test Practices

Air Fance and Lufthansa adjust runway length at Bogota airport

(BEA2017-0148): French BEA investigators have disclosed the carriers’
actions in an analysis of a serious departure incident from the Colombian
capital Bogota in March 2017. The A340-300 started rotating at 142kt some
2,760 m from the threshold of runway 13R, which is 3,800 m long and has a
clearway of 300 m. But it did not lift off for another 11s at which point it
was only 140 m from the threshold of the opposite-direction runway 31L,
crossing it at a height of just 6 ft.
Air France and Lufthansa, both of which served Bogota with A340-300s,
conducted a joint analysis of rotation techniques in the aftermath of the inci-
dent. The studies showed an average continuous rotation rate of 1.98/s rather
than the 3.18/s required by the A340-300 performance model. For the com-
bined Bogota study the two airlines showed similar average rotation rates of
around 1.88/s and average rotation times of around 7 s resulting in an
additional 200m takeoff distance compared to the certified model. This was
the main contributing factor to the longer take-off distance.
Both Air France and Lufthansa have introduced a restrictive measure for
Bogota departures, an artificial reduction of the runway length for perform-
ance calculations in order to provide additional safety margins.

Rapid Rotation
To support the data expansion with VMU , some takeoff at the scheduled
rotation speed VR will need to be performed with rapid rotation (higher than
normally used). This will allow one to verify that VR is sufficiently distant
from VMU and respect §25.107(e)(1)(iv):

§25.107(e) VR , in terms of calibrated airspeed, must be selected in

accordance with the conditions of paragraphs (e)(1) through (4) of this
section:

(1) (iv) A speed that, if the airplane is rotated at its maximum practicable
rate, will result in a VLOF of not less than—
(A) 110 percent of VMU in the all-engines-operating condition, and 105
percent of VMU determined at the thrust-to-weight ratio
corresponding to the one-engine-inoperative condition;

Although there are often various interpretations of what constitutes a

“maximum practical pitch rate,” one interpretation, that the pitch rate gener-
ated should not require exceptional piloting skills to precisely capture the
target pitch angle, has been accepted by the authorities. A small but sufficient
number of rapid rotations should be done at various thrust-to-weight con-
ditions to come up with a delta over the normal VR to VLOF takeoff model.
The data collected should cover both AEO and OEI thrust range (see
Fig. 16.46).
 CHAPTER 16 Takeoff Performance Testing 795

12

10

8 V
V (KTAS), t (s)

V
6 id)
(Rap

t
2
t
(Rapid)
0
0 0.05 0.1 0.15 0.2 0.25
((T – D) / W) @ VR

Fig. 16.46 Impact of rapid rotation on speed spread between VR and VLOF .

Instrumentation
The parameter of prime importance for takeoff distance performance
testing is the distance! The use of INS integrated acceleration or DGPS distance
or even video recording of the takeoff with ground references are acceptable
means of building a model, provided enough data are collected to support the
approach used. Another aspect of the takeoff performance testing is to show
repeatability in the execution of the takeoff for capturing safe speeds. The
minimum instrumentation required to document the testing typically involves:
• Means to measure distance along runway
• Weather station information to correct for winds during the data reduction
• Pitch force and/or elevator position to determine pilot action to
initiate rotation
• Stabilizer (or pitch trim) position to verify that it was correctly set
• RADALT and/or DGPS to determine height above ground
• Airspeed, altitude, and temperature
• Aircraft weight
• Wheel speeds (facilitates finding the liftoff point)
• Angle of attack, pitch, pitch rate, and flight path angle
• Engine parameters: anything that is required to run the engine deck and
re-create the thrust lapse rate for each test point
A note on wheel speed vs ground speed: Most manufacturers will cali-
brate the wheel speeds to closely match the airplane ground speed;
796 Operational Aircraft Performance and Flight Test Practices

KCAS (kt) Wheel Spd (kt) RADALT (ft) Ground Speed (kt)
200 100
190 90
180 80

Ground Speed (kt)

Wheel Spd (kt)
170 70

RADALT (ft)
Well calibrated

KCAS (kt)
160 60
150 50
140 40
130 30
120 20
110 10
100 0

KCAS (kt) Wheel Spd (kt) Ground speed (kt) RADALT (ft)
200 100

180 80
Ground Speed (kt)
Wheel Spd (kt)

RADALT (ft)
KCAS (kt)

160 60
Underpredicting
140 40
ground speed
120 20

100 0

KCAS (kt) Wheel Spd (kt) Ground speed (kt)

170

160
Ground Speed (kt)

150
Wheel Spd (kt)
KCAS (kt)

Overpredicting 140

ground speed 130

120

110

100

Fig. 16.47 Wheel speed calibration.

however, a wheel speed is nothing more than a wheel RPM counter using an
average rolling radius to compute a horizontal speed. That rolling radius is
dependent on the weight on wheel (airplane weight and CG dependent
and impacted by lift). It should not be considered the primary source of
ground speed for the purpose of computing distance. It does provide a
good indication of MLG wheel liftoff point (see Fig. 16.47), because one
will see the speed decay. The signal can be somewhat noisy and make it
harder to pick a reliable (within a tenth of a second) liftoff point, but it will
provide a reliable source once the performance engineer decides on the
best means to read the information (e.g., in the last trace of Fig. 16.47,
liftoff is the midpoint in the plateau of the wheel speed).
Whatever ground speed source is selected as a reference, it should be
cross checked against the along-runway ground distance (see Fig. 16.48),
especially if the distance and the ground speed come from two different
sources. If the ground speed or ground distance sources are noisy, one
may need to apply filtering (see Fig. 16.28 earlier in the chapter), which
may introduce small differences.

Certification Regulations: Robustness of Takeoff Procedure

The requirements of Part 25 (and of most of Part 23) are designed to
ensure an airplane will be able to safely continue climbing following the
loss of the most critical engine on takeoff. There is no single-engine airplane
certified to Part 25. The requirements also ask the applicant to validate that
 CHAPTER 16 Takeoff Performance Testing 797

Ground speed (kt) Wheel Spd (ft) Distance along runway (ft)
7000
140
100.7
6000
120

Distance along runway (ft)

5000
Ground speed (kt)

100 60.2
Wheel Spd (kt)

4000
80
3000
60
40.2
2809
40 2000
X = 1025 ft
20 t = 7.5 s 1000
1479 1784 Vaverage = (1025/7.5) = 136.7 ft/s
= 81 KTGS
0 s s s 0
40 50 53.8 57
.4 60 64
.9 70 80
Time (s)

Fig. 16.48 Crosscheck of distance and ground speed.

the takeoff procedure established by test provides robustness to in-service

takeoff procedure deviations by the crew operating the airplane.

Abused Takeoff
The final steps to validate the takeoff performance is to do abused takeoff
testing. Under normal operation, one can expect the airplane to be subjected
to several abnormal takeoff configurations (in addition to the OEI case
already tested at the normal takeoff speeds). These include:
• Overrotation on takeoff
• Early rotation that could be due to miscalculation of the rotation speed
(weight computation error or AFM chart reading error)
• Mistrim takeoff where the trim setting is not appropriate for the CG
location (could also be due to a miscalculation of the weight and
CG or simply an oversight during the pretakeoff checks) that could lead
to either:
• Very high pull force being required to lift the nose
• The airplane rotating by itself and the pilot requiring a push force to
stop the rotation rate prior to overrotation and possibly stall
The verification is done to meet the requirements of §25.107:
§25.107(e) (4) Reasonably expected variations in service from the established
takeoff procedures for the operation of the airplane (such as over-rotation of
the airplane and out-of-trim conditions) may not result in unsafe flight
characteristics or in marked increases in the scheduled takeoff distances
established in accordance with §25.113(a).
798 Operational Aircraft Performance and Flight Test Practices

The guidance material of FAA AC 25-7D provides accepted testing method-

ologies to verify that the airplane is still safe to operate under normally
expected in-service deviation. Note that the AC provides test conditions
for “traditional” deviations in service. If an airplane has special features
that could impact takeoff, they should be evaluated appropriately.

Early Rotation
An early rotation, prior to reaching the appropriate manufacturer-defined
VR , can result from several factors including:
• Miscomputation of the airplane weight resulting in using speeds that are
too low for the actual weight
• AFM chart reading error
• Transcription error
• Trying to take off because of a sudden obstacle on the runway

NTSB DEN90LA046: United Airlines B757-222, 16 Jan. 1990

While preparing for flight in a Boeing 757, the first officer inadvertently com-
puted the takeoff data by using the Boeing 767 data. V1 , VR , and V2 speeds
were calculated to be 115, 118, and 129 kt, respectively. The correct speeds
should have been 145, 148, and 152 kt, respectively. The captain did not
confirm the V-speed computation; both pilots set their airspeed bugs at the
improper speeds. During liftoff, the aircraft was overrotated and the lower
aft fuselage contacted the runway. The crew noted a “jolt,” but elected to con-
tinue to their destination after discussing the situation with maintenance per-
sonnel. After landing, damage was noted to the lower aft skin, aft pressure
bulkhead, and associated structure.

Boeing 737-800 tail strike, 28 July 2018

A Blue Air Boeing 737-800 had a tail strike at Birmingham (UK) airport on 28
July 2018, when a crew error led to the use of the wrong takeoff weight to
compute the takeoff speeds.
The UK Aviation Accident Investigation Board (AAIB) stated:
During the pre-flight preparation, the commander read the ZFW from
the load sheet to the co-pilot, instead of the TOW, and this was
entered into the EFB. The takeoff performance data calculated conse-
quently used a takeoff mass of about 12 tonnes less than the aircraft’s
actual weight. This produced takeoff speeds that were more than 10 kt
less than those required. Having flown the takeoff using these slower
speeds the aircraft suffered a tailstrike.
 CHAPTER 16 Takeoff Performance Testing 799

The UK AAIB bulletin noted that the crew had used the zero fuel mass instead
of takeoff mass for takeoff performance computation, which resulted in
V1 ¼ 140 kt, VR ¼ 140 kt, and V2 ¼ 143 kt being computed instead of the
correct V1 ¼ 152 kt, VR ¼ 153 kt, and V2 ¼ 157 kt. As a result, the crew
rotated the aircraft at 143 KIAS and attained a nose-up pitch angle of
11.95 deg with a peak pitch rate of 4.2 deg/s until the aircraft became airborne.
The crew suspected the tail might have contacted the runway surface.
Nonetheless, the crew decided to continue the flight, climbed to FL350 en
route and landed safely in Bucharest about 3 h after departure.

The manufacturer must validate that both the OEI and AEO conditions
have acceptable rotation speeds to address the requirements of §25.107. For
the OEI case, when an engine fails after V1 , the pilot must continue the
takeoff. If the speed spread between V1 and VR is relatively large, the pilot
will see the aircraft acceleration diminish significantly and may become
weary of the approaching end of the runway, and thus start the rotation
early. The requirements of §25.107(e)(3) defines this as an rotation occurring
5 knots slower than the planned rotation speed.

§25.107(e)(3) It must be shown that the one-engine-inoperative takeoff

distance, using a rotation speed of 5 knots less than VR established in
accordance with paragraphs (e)(1) and (2) of this section, does not exceed
the corresponding one-engine-inoperative takeoff distance using the estab-
lished VR . The takeoff distances must be determined in accordance with
§25.113(a)(1).

This requires only a 5-kt reduction because an early rotation under

OEI condition still needs to occur above V1 for the takeoff to continue.
During such an event, one should expect that the airplane will become air-
borne sooner (higher AoA) and will adopt a higher pitch attitude sooner
closer to the ground, resulting in less tail clearance (see Fig. 16.49).

Rotating toward Rotating toward

target pitch target pitch
Min tail clearance Min tail clearance

Tail height Tail height

Gear height Gear height

VR-5 ≈VLOF–5 VR VLOF

Higher pitch/aoa
on liftoff
Early rotation Rotation at VR

Fig. 16.49 Events in an early rotation.

800 Operational Aircraft Performance and Flight Test Practices

In-Service Deviation that Leaves Its Mark: Singapore Transportation

Accident Report 03-003
On 12 March 2003, a B747-400 from SIA had a severe tail strike where the tail
scraped the runway for nearly 1600 ft (»490 m) before the aircraft became air-
borne with resulting severe damage to the tail structure. A postflight review of
the takeoff by investigators revealed that an error in the takeoff weight tran-
scription (weight used was 247.4 t, but should have been 347.4 t) led to a
takeoff data miscalculation, resulting in excessively low takeoff speed (rotation
speed was nearly 33 kt lower than the normal 163 kt) and low takeoff thrust.
During this accident, the airplane lifted off at nearly stall speed and the shaker
fired for nearly 6 s throughout the maneuver.

FLAP B747-400 RWY

SQ 286 DATE
20 23L
APT NZAA ELEV +23’’

ATC TO / TO1 / D- 52 ATIS

T/O EPR 1. 34

V1 1 23

VR 1 30

V2 1 43
STAB 6.6
TRIM TOW 247.4 OAT 22 C

QNH CORR 2.0

389 1009
POB QNH
ADJ TOW 249.4

OM 27.12.99 NFW 231.0 FORM SOPS 111A

[fuel] 116

Figure 1 Figure 4
Bug card take-off data 9V-SMT damaged tail section

The test technique for an OEI early rotation should be similar to the
normal takeoff procedure in terms of target pitch attitude, pitch rate, and
trim setting; therefore, it can be expected that the airspeed at 35 ft above
ground can be lower than V2 . AC 25-7D suggests that the speed should
not be lower than V2 – 5 kt. The guidance material also suggests that inten-
tional tailskid or tail strike is not considered acceptable but suggests
occasional, minor/nondamaging contacts are allowed. Some airplanes that
are more prone to tail contact events are equipped with tail skids.
This test can be done with an actual engine failure (fuel cut) or simulated
engine failure (reduced to idle). If the former is used, then the engine is cut at
or near takeoff procedure V1 , allowing the engine to be fully off by VR – 5. If
instead, an idle chop is used, the engine must typically be reduced to idle
several seconds prior to the target speed of VR – 5 to allow the engine to
spool down by the target speed, similar to Fig. 16.42. So, when simulating
an engine failure, care must be taken that the selected engine cut speed
 CHAPTER 16 Takeoff Performance Testing 801

respects the established VMCG speed or that other mitigating actions are
taken (such as the use of nose wheel steering).
To minimize the risk of a tail strike under OEI conditions, one can select a
target pitch attitude of no more than the geometric tail clearance with
landing gear fully extended. Another way is to recommend slower pitch
rate to capture the target pitch attitude to allow time for the airplane to accel-
erate, thus lowering the angle of attack at liftoff.
The probability of an early rotation without engine failure (AEO con-
dition) is greater than one with OEI, so the testing done to validate the
takeoff speeds for AEO is done under more severe abuse conditions (see
Fig. 16.50). Under §25.107(e)(4) (author’s underline):

§25.107(e) (4) Reasonably expected variations in service from the established

takeoff procedures for the operation of the airplane (such as over-rotation of
the airplane and out-of-trim conditions) may not result in unsafe flight
characteristics or in marked increases in the scheduled takeoff distances
established in accordance with §25.113(a).

The guidance material of AC 25-7D suggests performing this test at near-

maximum sea-level takeoff weight and with a rotation speed that is 10 kt
or 7% lower than the normal VR for the test weight (whichever results
in the higher rotation speed). From this reduced VR , the test should be per-
formed at a rapid rotation rate and to a takeoff target pitch attitude 2 deg
higher than normally scheduled (not combined, two separate test conditions).
The 2-deg abuse (see Fig. 16.50) is typically more critical for airplanes
with few AoA margins to stall at V2 . This is typical of an airplane that is
not geometry limited at liftoff (tail clearance greater than stall warning
angle) with a higher aspect ratio wing, a hard leading edge where the
margin between V2 AoA and stall AoA may only be 5 to 6 deg out of
ground effects. This combination of VR – 10 and 2-deg target pitch abuse
may lead the airplane to be at a higher angle of attack near the ground (in
ground effect) where the AoA for CLmax is lower. For the tests using overro-
tations, the resulting increased pitch attitude (2 deg above normal target
pitch) should be maintained until the airplane is out of ground effect (one
wingspan) to validate that there are no adverse effects. One must remember
that VR was selected so that the airplane VLOF is at least 10% above VMU
using rapid rotation, so this test should not be an issue for tail strike or
adverse handling; it is a validation of the normal takeoff procedure. One
should also expect that the airplane may reach slightly more draggy configur-
ation than a normal takeoff because when the testing is initiated at VR – 10,
one should expect the airspeed to be about V2 (10 kt slower than V35 , which
is normally V2 þ 10) at 35 ft. One must check the resulting distance against
the AFM expected distance for the configuration tested.
For the rapid rotation (see Fig. 16.50), the airplane will achieve a higher
AoA sooner but will not exceed the target pitch attitude per the takeoff
procedure. This will result in higher AoA at liftoff and possibly lower tail
802 Operational Aircraft Performance and Flight Test Practices

VR-10, 2 deg abuse

Elevator (deg) Wheel speed (kt)

5 200
147.631

Wheel speed (kt)

0 150
Elevator (deg)

–6.816073
–5 100

–10 50

–15 0

Pitch (deg) Pitch rate (deg/s) AoA (deg)

–20
8.340406
9.267693
–15
Pitch rate (deg/s)
Pitch (deg)

AoA (deg)

–10 3.884791

–5

RADALT (ft) Tail clearance (ft)

30
Tail clearance (ft)
RADALT (ft)

20
3.418074
–0.323796
10

0 1
ne
95 100 t li 1 105 110 115
ven 2.54
E 0
1 Time (s)

Liftoff Min tail clearance

VR-10 rapid rotation

Elevator (deg) Wheel speed (kt)

5 136.8753 160
0 140

Wheel speed (kt)

–9.664332 120
Elevator (deg)

–5 100
–10 80
–15 60
40
–20 20
–25 0

Pitch (deg) Pitch rate (deg/s) AoA (deg)

20
7.870076
9.795231
15
Pitch rate (deg/s)
Pitch (deg)

AoA (deg)

10 5.440601

–5
RADALT (ft) Tail clearance (ft)

30
Tail clearance (ft)
RADALT (ft)

20 1.826497

–0.75
10

0 1
ne
75 80 t li 85 90 95
ven .001
E 84 Time (s)

Liftoff Min tail clearance

Fig. 16.50 Two speed abuse conditions.

 CHAPTER 16 Takeoff Performance Testing 803

clearance (see Fig. 16.37 earlier in the chapter). A rapid rotation, just like for
VMU , is one where the pilot can achieve the test target pitch attitude with
no exceptional piloting skills, not necessarily the maximum capability of the
airplane. In this case, the rapid rotation test should also be a nonevent with
VLOF being equal to or greater than the VMU speed for the test condition.
From an airplane flight characteristics point of view, FAA AC 25-7D
specifies that the activation of any stall warning devices, or the occurrence
of airframe buffeting during the takeoff speed abuse testing, is unacceptable
because it may lead to the crew not following the procedure that ensures
obstacle clearance (takeoff flight path) by reducing the pitch angle (lowering
AoA). This fact needs to be considered when selecting the pitch attitude at
liftoff. The guidance material does specify that tail strikes during these
abuse test demonstrations are acceptable if they are minor and do not
result in unsafe conditions.
From a takeoff distance point of view, §25.107(e)(4) states that during
the abuse testing, the resulting takeoff distance should not result in a
“marked increase” as compared to the AFM value. FAA AC 25-7D states
that a marked increase in the takeoff distance represents an increase of 1%
over the AFM scheduled distance for the same condition. The AFM sched-
uled distance, as we will discuss more in Chapter 18 in the section on
AFM takeoff distance, is the longer of the OEI takeoff distance or 115% of
the AEO takeoff distance.

Mistrim Takeoff
Finally, for abuse takeoff test, a CG miscalculation or flight crew
inattention resulting in the trim setting being incorrectly set should also be
expected in service. To limit the risk of a dangerously out of trim condition
on takeoff, OEM provides indications of acceptable trim range in the form
of a “green band” (see Fig. 16.51). If the trim setting were to be set outside
this acceptable range on takeoff, the airplane’s system will provide a configur-
ation warning to the crew so that they can abort the takeoff and set proper
limits.

Fig. 16.51 Example of acceptable green band displays.

804 Operational Aircraft Performance and Flight Test Practices

This reduced approved trim range for takeoff helps prevents dangerous
takeoff configurations, but this range must be validated by testing. The criti-
cal condition for takeoff performance is the forward CG mistrim condition.
For this test, one sets the airplane CG at the most critical forward location
while setting the trim in the full, authorized, nose-down condition (trim
set for most adverse aft CG).
For the forward CG mistrim test, the airplane is set up in the normal
takeoff configuration, using normal target VR and V2 , and AEO, but with
the trim set for full nose-down position. The trim setting should be at the
limit that will be just shy of the configuration warning. Because the pilot
will get less help from the trim system to pitch the nose up, one should
expect high pitch forces, possibly reaching maximum elevator travel, and
slow to no rotation during the test.
The flight test data of Fig. 16.52 represent an example of a forward CG
takeoff with the airplane mistrimmed to the aft CG trim setting. The pilot
applies a pitch force at the target VR . The elevator reaches its travel limits,
and the pilot pulls against the control stop (high pull force) while the airplane
initially fails to rotate. As the airspeed increases, the dynamic pressure
becomes large enough to start the rotation, and the airplane finally lifts off the
ground. The airplane pitch forces are still high by the time the aircraft reaches
35 ft AGL.
For this test, the forces used must meet the requirements of §25.143(d)
short-term application. For the case of the airplane in Fig. 16.52, those
forces were reached as the elevator reached the travel limits; any additional
force did not contribute to lifting the nose because the pilot was pulling
against a control stop. As well, for this test, there must be no “marked
increase” in the takeoff distance, as discussed in previous sections.

Nose-Down Mistrim: Aero Commander 100 (N3628X), 8 Jan. 2003,

Geneva, AL
NTSB Identification: ATL03LA032. The accident occurred during the initial
takeoff from a private grass airstrip in Geneva, Alabama. According to the
pilot, as the airplane accelerated down the runway during the attempted
takeoff, the airplane would not climb and he could not explain why. According
to the pilot, he elected to abort the takeoff with approximately 400 ft of runway
remaining. When the pilot elected to abort the takeoff, the airplane veered off
the right side of the runway and collided with trees. Examination of the air-
plane revealed that the elevator trim was observed in the full nose-down pos-
ition. The right wing was damaged at the root, and the right wing spar was
bent. The pilot did not report any mechanical difficulties prior to the accident.
According to the pilot’s operating handbook, the pretakeoff checklist requires
the elevator trim be set for takeoff.
 CHAPTER 16 Takeoff Performance Testing 805

140
120
Target VR
100
80
150

100

50
30
20
10
0
–10
200
150
100
50
0
20
10

0
–10
60
40
20
0
124 126 128 130 132 134 136 138 140

Fig. 16.52 Forward CG mistrim test point.

Although not directly tied to takeoff distance performance, the test team
must also validate the opposite trim setting, the full nose-up condition
(trimmed for maximum forward CG) with the airplane tested at the most
critical aft CG configuration; the takeoff trim band must be certified to be
safe. Figure 16.53 shows such an example. Here, as the airplane accelerates
through VR , the pilot initiates a rotation by applying a pull force. Soon
after, the pilot must relax the pull force and even start applying a push
force (normally unacceptable characteristics for a normal takeoff). The
characteristics are acceptable if the pilot does not have to apply more than
the temporary maximum push force of §25.143(d) while performing the
test to the target normal takeoff speed.
806 Operational Aircraft Performance and Flight Test Practices

Fig. 16.53 Aft CG mistrim test point.

Nose-Up Mistrim: DC-8-54F (N8053U), 11 Jan. 1983, Detroit, MI

NTSB Identification: DCA83AA014. According to witnesses, the takeoff roll
was normal, and the aircraft rotated to a takeoff attitude about 12 to 23 of the
way down the runway. After liftoff, the aircraft’s pitch attitude steepened, and
the plane climbed to about 1000 ft AGL. The aircraft then rolled to the right,
entered a steep descent, crashed, and burned. The horizontal stabilizer was
found at 7.5 units aircraft nose-up (ANU). About 65 s before takeoff, the first
officer and second officer changed duty positions. The board found that the
flight crew failed to follow procedural checklist requirements and to detect
and correct a mistrimmed stabilizer before the aircraft became uncontrollable.
 CHAPTER 16 Takeoff Performance Testing 807

Takeoff Performance Testing: Risks and Mitigations

Until the applicant can establish a safe and repeatable AFM takeoff pro-
cedure, any takeoff testing will involve different levels of risk.
• Stall on takeoff
• Inability to climb due to high drag attitude
• Engine failure on takeoff due to high angle of attack
• Tail strikes
• Loss of directional control during OEI and/or nosewheel steering off tests
• Single engine operation at low altitude
• Runway overrun
The risk mitigations for normal AEO and OEI testing following the pre-
scribed takeoff procedure (target pitch and speeds) are:
• Stall testing completed prior to dedicated takeoff tests
• Chops to idle prior to engine cuts
• VMC testing completed
• VMU testing completed
• Establish WAT limits prior to takeoff testing
• Runway of sufficient length
• Build up rotation rates
• Buildup in target pitch attitude
• Test high T/W prior to low T/W
• Test AEO prior to OEI
• Engine ignition on for initial testing

Use of Telemetry
Another risk mitigation factor for takeoff performance is real-time teleme-
try monitoring. This allows a series of engineers and ground crew to monitor
several more parameters during a takeoff than what the flight crew can do. The
flight crew should concentrate on the takeoff technique, target speeds, atti-
tudes, pitch rate, and so forth. The telemetry crew can provide feedback on
the soundness of the data collected soon after the takeoff is completed and
identify the need to immediately repeat the test point if necessary, thereby
decreasing the risk of data scatter. The telemetry crew can also monitor critical
parameters such as brake temperature and engine parameters, determine
liftoff angle and closeness of the tail to the ground (tail strike risk), and so
on. Overall, telemetry reduces test risk and the number of repeat flights.
The basic risks associated with speed abuse takeoff tests are, of course, the
same as for normal testing, but the following risks should be carefully con-
sidered by the test team due to the higher pitch and AoA near the ground:
• Tail strike
• Engine problems due to the higher angle of attack
• Stall on takeoff (low speed, overrotation, IGE)
808 Operational Aircraft Performance and Flight Test Practices

• Loss of directional control (OEI test, one wing stalling first)

• Inability to climb (WAT tests)
Some of the risk mitigation for the abuse testing includes:
• VMC testing completed.
• Nose wheel steering on, if available.
• VMU testing completed.
• Get a clear understanding of the lift capability in ground effects.
• Runway of sufficient length to be used.
• Testing shall not be done at weights exceeding the WAT limit.
• Normal takeoff testing and final (preliminary) takeoff speeds are to be
established prior to abuse speed testing.
• No obstacles near the edge of the runway.
• AEO abuse tests prior to OEI abuse testing.
The mistrim testing can be limited to the two most adverse conditions,
one forward CG and one aft CG. The risks associated with mistrim takeoff
testing include:
• Inability to rotate the aircraft leading to possible runway overrun
(nose-down trim and forward CG)
• Inability to stop the rotation on takeoff and stall (nose-up trim and aft CG)
The risk mitigations for this testing include:
• Nose-down mistrim
• Use of runway of sufficient length.
• If rotation has not started at a given, pretest-determined distance,
initiate a rejected takeoff or use the pitch trim system to help with
the rotation.
§ The preferred approach must be agreed to prior to the test. It should
not be left to the pilot to decide during the test, because this could lead
to moments of hesitation and increase the risk of an accident.
• Mistrim testing using a midpoint trim setting can be done to validate
pretest analysis and estimates.
• Nose-up mistrim
• Mistrim testing using a midpoint trim setting can be done to validate
pretest analysis and estimates.
As a final discussion on test risk, we briefly discuss the accident involving a
Gulfstream G650 in Roswell, New Mexico, in April 2011. We are not doing this
to assign blame, but instead to show the real risk involved with takeoff perform-
ance testing. We must learn from our past experiences as a test community,
and we cannot forget. This author lost a friend and ex-colleague in this acci-
dent. We also use these data because they are publicly available in the NTSB
report (see accident box on NTSB/AAR-12/02).
From the report, we find the data shown in Fig. 16.54. The figure
compares important parameters from the accident takeoff and the takeoff
 DCA11MA076: GAC G540, NG52GD, Roswell, NM, 04/02/2011
Comparision of flight 153 flaps 10 OEI runs
10
Calibarated airspeed, knots

145
8 Accident

Pitch rate, degree/second

140 Prior TO 6
4
135
Accident 2 Prior TO
130 0
–2
125 –4 Run 6C3
Run 7A1 –6 Run 7A2 (accident)
120 Run 7A2 (accident)
–8
115 –10
–1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 –1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

CHAPTER 16
16
14 G: Gear handle selected up Accident 40
L: Left main WOW = AIR 35
12
Pich angle, degrees

R: Right main WOW = AIR R

L 30
10

Radio altitude, feet

G
L R 25 Run 7A1
8 Run 7A2 (accident)
6
Prior TO 20 Accident
15

Takeoff Performance Testing

4
Run 7A1
10
2 Run 7A2 (accident)
0 5
0
–2 Prior TO
–1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 –5
Time elapsed from start of column pull, seconds –1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Time elapsed from start of column pull, seconds

Fig. 16.54 G650 accident takeoff vs the previous takeoff, parameters compared. Source: NTSB/AAR-12/02.

809
810 Operational Aircraft Performance and Flight Test Practices

that occurred just before (Prior TO). For both conditions, the pilot put in a
control input at about the same VR (within test tolerance, time 0 on charts).
Note, this was not an abuse test but a normal OEI test. The airplane reached
the same initial target pitch attitude of 9 deg for both conditions at the same
time (thus the same pitch rate, as shown). The target pitch was reached prior
to liftoff. The main difference between the two conditions is that, for the acci-
dent takeoff, the pitch angle continued to increase by a few more degrees and
the airspeed stagnated (high drag attitude, low excess thrust), which resulted
in a higher angle of attack while the airplane was in ground effects; this was
effectively the equivalent of a 2-deg abuse test condition. Soon after liftoff,
one wing lost some lift and the wing tip hit the ground. The airplane and
the crew were lost during the follow-on crash.

NTSB/AAR-12/02: Crash during experimental test flight, Gulfstream Aero-

space Corporation GVI (G650), N652GD, Roswell, New Mexico, 2 April 2011.
The accident occurred during a planned one-engine-inoperative (OEI)
takeoff when a stall on the right outboard wing produced a rolling moment
that the flight crew was not able to control, which led to the right wingtip con-
tacting the runway and the airplane departing the runway from the right side.
After departing the runway, the airplane impacted a concrete structure and an
airport weather station, resulting in extensive structural damage and a post-
crash fire that completely consumed the fuselage and cabin interior.

. . .In preparing for the G650 field performance flight tests, Gulfstream
considered ground effect when predicting the airplane’s takeoff performance
capability but overestimated the in-ground-effect stall AOA. Consequently,
the airplane’s AOA threshold for stick shaker (stall warning) activation and
the corresponding pitch limit indicator (on the primary flight display) were
set too high, and the flight crew received no tactile or visual warning before
the actual stall occurred.
The accident flight was the third time that a right outboard wing stall
occurred during G650 flight testing. Gulfstream did not determine (until
after the accident) that the cause of two previous uncommanded roll events
was a stall of the right outboard wing at a lower-than-expected AOA. . . .
 CHAPTER 16 Takeoff Performance Testing 811

During field performance testing before the accident, the G650 consist-
ently exceeded target takeoff safety speeds (V2). . . . Rather than determining
the root cause for the V2 exceedance problem, Gulfstream attempted to
reduce the V2 speeds and the takeoff distances by modifying the piloting tech-
nique used to rotate the airplane for takeoff. . . .
The NTSB determines that the probable cause of this accident was an
aerodynamic stall and subsequent uncommanded roll during an OEI takeoff
flight test, which were the result of (1) Gulfstream’s failure to properly
develop and validate takeoff speeds for the flight tests and recognize and
correct the V2 error during previous G650 flight tests, (2) the G650 flight
test team’s persistent and increasingly aggressive attempts to achieve V2
speeds that were erroneously low, and (3) Gulfstream’s inadequate investi-
gation of previous G650 uncommanded roll events, which indicated that
the company’s estimated stall AOA while the airplane was in ground effect
was too high.

Exercise
1. Compute the net lift at liftoff for this example where the airplane weight is
155,000 lb and the thrust to weight at liftoff is 0.295. Assume the thrust
line is aligned with the airplane angle of attack. Compute the average
acceleration of the airplane from VR to VLOF . What was the distance
covered during this segment?

KCAS (kt) Wheel Spd (kt) Ground Speed (kt)

170
160
150
Ground Speed (kt)
Wheel Spd (kt)

140
KCAS (kt)

130
120
110
100
Pitch (deg) Pitch rate (deg/s) Elevator (deg)
20 0

15
–5
Pitch rate (deg/s)

Elevator (deg)
Pitch (deg)

10
–10
5
–15
0

–5 –20
38 39 40 41 42 43 44 45 46 47 48
Times (s)

Reference
[1] Howard, J., and Donovan, P., “A Tragic First—Gulfstream G650 Flight Test Accident,”
Gulfstream presentation material, Flight Test Safety Committee 2013 Workshop,
New Orleans, 23 April 2013.
Chapter 17 Rejected
Takeoff
Performance

Chapter Objective
In Chapters 15 and 16, we covered the material for the continued takeoff,
either AEO or OEI. There might be conditions that develop during a
takeoff roll where the crew may elect to reject the takeoff and stop. To
ensure this can be done safely, certification requirements define a
maximum reject speed (V1 ) that is shown to provide a safe maximum speed
and also ensures that the stop can be done within the available runway
length. Our discussion in this chapter will setup the framework for safe
rejected takeoff (RTO).

Why Reject a Takeoff?

A
rejected takeoff (RTO) is not part of normal operations; the pilot
does not plan on rejecting a takeoff when the airplane gets on the
runway. A rejected takeoff occurs when an emergency arises, and
the crew decides to not take off. Part 25 transport category airplanes’
takeoff distance will always cater to the possibility of a rejected takeoff
(accelerate-stop distance) in the instructions provided for the crew,
§25.101(f):
25.101(f) Unless otherwise prescribed, in determining the accelerate-stop
distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must be
made in accordance with procedures established by the applicant for oper-
ation in service.

Experience has shown that RTO events are few and far between (see
Fig. 17.1), which makes it more of a surprise to pilots when it does occur.
Regular training for RTO will help with the recognition of the need to
perform one and the steps required to ensure it is done properly.
So what would cause the crew to reject a takeoff? The reasons are numer-
ous; some of the leading causes are:
• Engine failure or troubles
• Fire
• Loss of thrust, slower than required acceleration

813
814 Operational Aircraft Performance and Flight Test Practices

80 Takeoffs, RTOs, and overruns

76% RTO overrun Through 2003 Typical recent year
60
Percent of total

accidents Takeoffs 18,000,000

430,000,000
principally come
40 from the 2% of the RTOs (est.) 143,000 6000
RTOs that are
high speed RTO overrun 97 4*
20 18% accidents/incidents
4%
2% 1 RTO per 3000 takeoffs
0
80 knots 80 to 100 to Above 1 RTO overrun accident/incident per 4,500,000 takeoffs
or less 100 knots 120 knots 120 knots *Accidents/incidents that would occur if historical rates continue.

Fig. 17.1 RTO statistics [1].

• Physical loss of engine

• Brakes/tires
• Tire failure with possible loss of heading control
• Brake seizure with slower than normal acceleration
• Jammed flight control—elevator jam preventing rotation
• Aircraft configuration
• Wrong flap setting
• Unstowed spoilers
• Wrong stabilizer trim setting
• Unlatched cockpit window
• Loss of instrumentation
• Electrical failures
• Loss of displays
• Pilot/copilot airspeed disagreement—which is the right one? (See
incident box ATSB AO-2013-212.)
• Weather, including impact on runway surface conditions
• Obstructions on the runway
• Equipment on the runway
• Wildlife on the runway
• Runway incursion by other aircraft
• Crew errors
• Unauthorized takeoff
• Taxiway takeoff
• Off-runway takeoff
• Loss of aircraft control
• Loss of directional control due to rudder jam

Irrespective of what the emergency is, it is the pilot’s decision to reject or con-
tinue a takeoff; although pilots do not plan to RTO, they should always be
prepared to do so. The pilot’s job can be helped by using automatic
systems monitoring equipment that can clearly announce to the pilot that
a critical system has failed.
 CHAPTER 17 Rejected Takeoff Performance 815

More and more airplanes are fitted with “dark cockpits” where annun-
ciator lights appear only when the pilot’s attention is required, thereby
allowing the pilot to concentrate on the takeoff. The takeoff is a
task-intensive phase of flight and requires all of the pilot’s attention. False
warnings or misinterpretation of warning lights have been known to cause
unnecessary RTOs.

ATSB AO-2013-212: On 21 November 2013, after a flight from Singapore, an

Etihad Airways Airbus A330, A6-EYJ landed at Brisbane airport and was
taxied to the terminal. Approximately 2 hours later, the aircraft was pushed-
back from the gate for the return flight to Singapore.
The captain rejected the initial take-off attempt after observing an airspeed
indication failure on his display. The aircraft taxied back to the terminal where
troubleshooting was carried out, before being released back into service.
During the second take-off roll, the crew became aware of an airspeed dis-
crepancy after the V1 decision speed and the take-off was continued. Once air-
borne, the crew declared a MAYDAY and decided to return to Brisbane where
an overweight landing was carried out.
Engineering inspection after the overweight landing found that the Cap-
tain’s pitot probe was almost totally obstructed by an insect nest, consistent
with mud-dauber wasp residue. The pitot obstruction had occurred during
the 2-hour period that the aircraft was on the ground at Brisbane and was
not detected during troubleshooting after the initial rejected take-off.

A6-EYJ A330 21 November 2013

Rejected takeoff (RTO) AO-2013-212
Yes
Yes Flight director 2 engaged No
No Flight director 1 engaged
Yes FO switched to ADR3 Yes
No Captain switched to ADR3 No
Engine thrust (N1) right (%)

Autothrust (ATHR) engaged

Engine thrust (N1) left (%)

Yes
No
100
80
60
100 40
90 Airspeed from ISIS (uses standby pitot/static system) 20
80 0
Groundspeed (kts)

70
CAS AD1 (kts)
CAS AD2 (kts)
CAS ISIS (kts)

60
50 Airspeed source for FO's PFD - ADR2
40
30 Groundspeed
20
10
0
Airspeed source for the Captain’s PFD - ADR1
12:03:35 12:03:43 12:03:50 12:03:58 12:04:05 12:04:13 12:04:20 12:04:28 12:04:35
FDR/DAR Plot: RTO EST (hh:mm:ss)
Revised: April 08, 2015 Australian Transport Safety Bureau (ATSB)
816 Operational Aircraft Performance and Flight Test Practices

RTO Distance Definition

An RTO can be broken down into segments for the purpose of analysis
(see Fig. 17.2). We will first consider the baseline case of an engine failure
during takeoff. The first part of the RTO sequence, up to the point an
event occurs (here, an engine failure), is essentially identical to the
all-engines-operating (AEO) takeoff acceleration covered in Chapters 15
and 16. From this point, the acceleration of the airplane is reduced due to
the one-engine-inoperative (OEI) condition. Once the situation is recog-
nized, the pilot decides to reject the takeoff, the first step is the application
of the first deceleration device, typically the brakes. The maximum airspeed
at which this can occur is V1 . The pilot then sequentially applies other decel-
eration devices, typically a throttle chop and spoilers. The rejected takeoff is
complete when the airplane reaches zero airspeed or a speed low enough to
taxi off the runway.
From a Part 25 transport category airplane certification point of view, the
requirements for rejected takeoff are spelled out in §25.109:
§25.109 Accelerate-stop distance
(a) The accelerate-stop distance on a dry runway is the greater of the
following distances:
(1) The sum of the distances necessary to—
(i) Accelerate the airplane from a standing start with all engines
operating to VEF for takeoff from a dry runway;
(ii) Allow the airplane to accelerate from VEF to the highest
speed reached during the rejected takeoff, assuming the
critical engine fails at VEF and the pilot takes the first
action to reject the takeoff at the V1 for takeoff from a dry
runway; and
(iii) Come to a full stop on a dry runway from the speed reached as
prescribed in paragraph (a)(1)(ii) of this section; plus
(iv) A distance equivalent to 2 seconds at the V1 for takeoff from a
dry runway.

AEO
V35
V1
VEF
Airspeed

Throttle chop
Spoilers
Brakes

Distance

Fig. 17.2 Typical RTO sequence for OEI condition.

 CHAPTER 17 Rejected Takeoff Performance 817

(2) The sum of the distances necessary to—

(i) Accelerate the airplane from a standing start with all
engines operating to the highest speed reached during
the rejected takeoff, assuming the pilot takes the first
action to reject the takeoff at the V1 for takeoff from a dry
runway; and
(ii) With all engines still operating, come to a full stop on dry
runway from the speed reached as prescribed in paragraph
(a)(2)(i) of this section; plus
(iii) A distance equivalent to 2 seconds at the V1 for takeoff from a
dry runway.

As we can see from §25.109, one must account for the worst (longest)
condition of either an OEI (engine failure) or an AEO, acknowledging that
rejected takeoffs can be initiated for other reasons than an engine failure
leading to a loss of thrust from the most adverse engine. We will expand
on this subject in this chapter.

Example 17.1
A 127-ft-long jetliner has an average acceleration capability of 0.25 g from
brake release to a reject speed V1 of 110 kt at ISA sea level with no winds
on a flat runway condition. At V1, the pilot rejects the takeoff and achieves
an average deceleration of 0.46 g to full stop. What is the actual total distance
from brake release to full stop? What would be the aircraft flight manual
(AFM) published distance? What is the braking distance? If the braking capa-
bility increases such that the average deceleration increases by 5%, what would
be the reduction in published AFM distance?
Solution: This is a simplified problem to illustrate the change in airspeed vs
time and distance (see Fig. 17.3). The starting point is the ISA sea level con-
dition with no winds, which means the calibrated airspeed, true airspeed,
and ground speed are all the same. Furthermore, because the acceleration
and deceleration phases are quoted in terms of average acceleration values,
one can easily resolve the problem using kinematic equations.
• The first question, actual total distance, becomes equal to 3307 ft. This is
composed of 2143 ft to accelerate to V1 and 1164 ft to stop.
• For the second question, what would be the AFM RTO distance, one must
use the actual distance computed in question 1 and add 2 s at V1 (110 kt,
185.6 ft/s), which equates to 371 ft additional (almost 32% of the stopping
distance in this case), for a total of 3307 ft.
• What is the braking distance was answered previously: 1164 ft.
• If the braking force is increased so as to increase the average deceleration by
5% (from 0.46 g to 0.483 g), the stopping distance would reduce by 55 ft

(Continued)
818 Operational Aircraft Performance and Flight Test Practices

Example 17.1 (Continued)

120

100
127 ft

80
Airspeed (kt)

60

40

20

0
0 5 10 15 20 25 30 35 40
Time (s)
120

100
127 ft

80
Airspeed (kt)

60

40

20

0
0 500 1000 1500 2000 2500 3000 3500
Distance (ft)

Fig. 17.3 Airspeed vs time and distance for Example 17.1.

(about 4.8% reduction), but the overall distance from brake release to V1 to
full stop would change by only 1.51%; the acceleration part of the maneuver
(the largest segment) and the safety margin (2 s at V1 ) are unaffected.

(Continued)
 CHAPTER 17 Rejected Takeoff Performance 819

Example 17.1 (Continued)

• Figure 17.3 was used to show the evolution of the maneuver as a function of
time and distance. Note how much distance is covered at the higher
airspeed for a given acceleration time. Rejecting a takeoff at high speed
requires a lot of runway.
What if the airplane had rejected the takeoff earlier, or later? Figure 17.4
shows the impact of a rejected speed variation by +30 kt. The impact is, of
course, nonlinear, being closer to a proportionality of the ground speed
square.

160.0

140.0

127 ft
120.0

100.0
Airspeed (kt)

80.0

60.0

40.0

20.0

0.0
0 1000 2000 3000 4000 5000 6000
Distance (ft)

Fig. 17.4 Impact of speed deviation on RTO distance.

Review of RTO Critical Speeds

Because there is a benefit, from a total distance point of view, to reject
a takeoff at a lower airspeed, one must consider the certification
options available. From a flight manual point of view, planning for
minimum published distance must account for the possibility of a rejected
takeoff. That RTO distance will be defined in terms of airspeeds
provided by the crew (crew procedure). The takeoff speeds specified under
§25.107 state:
820 Operational Aircraft Performance and Flight Test Practices

§25.107 Takeoff speeds

(a) V1 must be established in relation to VEF as follows:
(1) VEF is the calibrated airspeed at which the critical engine is assumed
to fail. VEF must be selected by the applicant, but may not be less
than VMCG determined under §25.149(e).
(2) V1, in terms of calibrated airspeed, is selected by the applicant;
however, V1 may not be less than VEF plus the speed gained with
critical engine inoperative during the time interval between the
instant at which the critical engine is failed, and the instant at which
the pilot recognizes and reacts to the engine failure, as indicated by
the pilot’s initiation of the first action (e.g., applying brakes, reducing
thrust, deploying speed brakes) to stop the airplane during
accelerate-stop tests.

§25.149 Minimum Control Speed

(e) VMCG, the minimum control speed on the ground, is the calibrated
airspeed during the takeoff run at which, when the critical engine is suddenly
made inoperative, it is possible to maintain control of the airplane using the
rudder control alone (without the use of nosewheel steering), as limited by
150 pounds of force, and the lateral control to the extent of keeping the wings
level to enable the takeoff to be safely continued using normal piloting skill.
In the determination of VMCG, assuming that the path of the airplane
accelerating with all engines operating is along the centerline of the runway,
its path from the point at which the critical engine is made inoperative to the
point at which recovery to a direction parallel to the centerline is completed
may not deviate more than 30 feet laterally from the centerline at any point.

Thus, the minimum published RTO distance per §25.109 will be defined by
the minimum published V1 speed as limited by the value of VMCG (through
VEF ). We label this minimum speed as V1MCG . This is not to say that an
airplane will not reject a takeoff at a lower speed, just that from a flight
manual point of view, we will not take credit for lower speeds from an
RTO distance point of view. We also note that V1 and VEF are applicant
defined (can be larger than the minimum values limited by §25.107).

V1 Critical Speed: Evolution from Decision Speed

V1 prior to FAA Part 25, Amendment 42 (1978) was defined as the
decision speed (author’s underline):
§25.107 Takeoff speeds. (prior to amendment 42)
(a) V1 must be selected by the applicant and must be at least the minimum
calibrated airspeed at which controllability by primary aerodynamic
controls alone is shown (during the takeoff run) to be adequate to safely
continue the takeoff, using normal piloting skill, when the critical engine
is suddenly made inoperative.

It was a speed where the pilot could either elect to continue takeoff or decide
to stop. The speed allowed for safe continued takeoff with aerodynamic con-
trols alone (essentially minimum control speed on the ground). The problem
with this approach is that an RTO distance computed with this airspeed
could assume start of braking at that point, but crews could actually interpret
 CHAPTER 17 Rejected Takeoff Performance 821

this speed as one where the pilot makes a decision (finite time involved) and
thus lead to braking at speeds higher than V1 , leading to a longer stopping
distance than computed (see accident box NTSB SIR-90/02, Air France
747-200). With Amendment 42, the concept of critical engine failure speed
limited by VMCG was introduced, and the focus of V1 was placed on the
application of the first deceleration device, thus the start of the RTO
braking phase; the pilot reaction time, as demonstrated by flight test, was
now included in the definition of V1 . Amendment 42 also introduced a con-
tinued acceleration for 2 s beyond V1 in §25.109 (underline added by author).
§25.107 Takeoff speeds. (Amendment 42, 1978)

(a) V1 must be established in relation to VEF as follows:

(1) VEF is the calibrated airspeed at which the critical engine is assumed
to fail. VEF must be selected by the applicant, but may not be less
than VMCG determined under Sec. 25.149(e).
(2) V1 , in terms of calibrated airspeed, is the takeoff decision speed
selected by the applicant; however, V1 may not be less than VEF plus
the speed gained with the critical engine inoperative during the time
interval between the instant at which the critical engine is failed, and
the instant at which the pilot recognizes and reacts to the engine
failure, as indicated by the pilot’s application of the first retarding
means during accelerate-stop tests.

§25.109 Accelerate-stop distance. (Amendment 42, 1978)

(a) The accelerate-stop distance is the greater of the following distances:
(1) The sum of the distances necessary to—
(i) Accelerate the airplane from a standing start to VEF with all
engines operating;
(ii) Accelerate the airplane from VEF to V1 and continue the
acceleration for 2.0 seconds after V1 is reached, assuming the
critical engine fails at VEF; and
(iii) Come to a full stop from the point reached at the end of the
acceleration period prescribed in paragraph (a)(1)(ii) of this
section, assuming that the pilot does not apply any means of
retarding the airplane until that point is reached and that the
critical engine is still inoperative.
(2) The sum of the distances necessary to—
(i) Accelerate the airplane from a standing start to V1 and continue
the acceleration for 2.0 seconds after V1 is reached with all
engines operating; and
(ii) Come to a full stop from the point reached at the end of the
acceleration period prescribed in paragraph (a)(2)(i) of this
section, assuming that the pilot does not apply any means of
retarding the airplane until that point is reached and that all
engines are still operating.]

Note that the words takeoff decision speed were removed at FAA Part 25,
Amendment 92 (1998, also at JAA JAR 25.107, Change 15, similar timeframe)
to try to help emphasize that the speed was meant to be the one for the appli-
cation of the first deceleration device. Also in Amendment 92, the continued
acceleration at V1 was replaced by a fixed distance equal to 2 s at V1 , as
822 Operational Aircraft Performance and Flight Test Practices

shown earlier in this chapter. Operationally, the changes introduced in

Amendments 42 and 92 translate into a clear statement that V1 is the
highest speed at which the pilot must have initiated an RTO. If the pilot
has not applied the first deceleration device by the time the aircraft
reaches V1 , the pilot has by default elected to continue the takeoff (from
an available runway distance point of view)!
The FAA Final Rule for Amendment 92 tells the story of the need for the
clarification of the meaning of V1 [2]:

The diversity displayed in the comments illustrates a great deal of misunder-

standing and disagreement regarding the definition and use of the V1 speed.
In general, inconsistent terminology used over the years in reference to V1
has probably contributed to this confusion. As noted by the commenters,
V1 has been referred to at various times as the critical engine failure speed,
the engine failure recognition speed, and the takeoff decision speed.
Special Civil Air Regulation No. SR-422, effective August 27, 1957, orig-
inally referred to V1 as "the critical engine failure speed." These same stan-
dards, which were later recodified into part 25, defined the accelerate-stop
distance as the distance to accelerate to V1 and then to stop from that
speed. Although an allowance was required for any time delays that may
reasonably be expected in service, SR-422 did not explicitly state where or
how the time delays should be introduced relative to V1. For certification pur-
poses, the FAA considered V1 to be the speed at which the pilot took the first
action to stop the airplane. Time delays for recognition and reaction to that
failure were applied prior to V1, and delays in accomplishing each subsequent
action for stopping the airplane were applied after V1. Allowing for the time
delays, the actual engine failure was therefore assumed to occur prior to V1.
With Amendment 25-42, effective March 1, 1978, the FAA amended the
airworthiness standards to clarify and standardize the method of applying
these time delays. V1 was referred to as the "takeoff decision speed," which
turned out to be ambiguous in that it could be interpreted to mean either
the beginning or the end of the pilot’s decision process. The preamble to
Amendment 25-42, however, states that "V1 is determined by adding to V1
[the speed at which the critical engine is assumed to fail] the speed gained
with the critical engine inoperative during the time interval between the
instant at which the critical engine is failed and the instant at which the test
pilot recognizes and reacts to the engine failure, as indicated by the pilot’s
application of the first retarding means during accelerate-stop tests." This
same definition was codified as Sec. 25.107(a)(2). Not only is V1 intended to
occur at the end of the decision process, but it also includes the time it takes
for the pilot to perform the first action to stop the airplane.

Civil aviation aircraft accident summary for the year 1988 (DGAC India)
and NTSB SIR-90/02: Delhi-Indira Gandhi International Airport, India, 24
July 1988. An Air France Boeing 747-200 suffered a takeoff accident when
the airplane overran the runway end. The copilot was flying the plane and,
based on the weight, the V1 speed was 156 kt. As the airplane was executing
the takeoff, 2.5 s beyond V1, the No. 4 Engine Fire warning came on. The
copilot elected to reject the takeoff at 167 kt. The airplane was destroyed
when it struck a ditch beyond the end of the runway; one passenger sustained
minor injuries.
 CHAPTER 17 Rejected Takeoff Performance 823

19 January 1997, Learjet 36A: Due to a hydraulic leak, the nose-gear steering
malfunctioned during the takeoff roll. The airplane exited the runway onto
snow-covered terrain during the follow-on RTO.

On 19 Feb. 2008, an ATR-72 from Air Bagan ran off the runway during a
takeoff roll when a failure indicator showed that one of the engines was not
working and the pilot aborted the takeoff. At that stage the nose gear had
become airborne and the aircraft was about 3700 ft (1130 m) down the
7000-ft-long runway.

Modeling an RTO: Transition Phase

We break down the acceleration-stop (RTO) maneuver into segments to
compute the distance. The first segment is the acceleration to a given air-
speed (VEF , V1 , or other). For this segment, the analysis follows the steps dis-
cussed in Chapter 15.
The next segment is the transition phase where multiple time-based
events occur. As a starting point, we write down the force equation along
the runway surface, Eq. (17.1) and Fig. 17.5.

  W dVGS
T  f  D  W sin urwy ¼ ð17:1Þ
g dt
where

f ¼ mB ðW  LÞ ð17:2Þ

mB represents the braking force coefficient applied to the wheel. During

the transition segment, this value will increase with time from the rolling
friction value (about 0.02 on hard surfaces) to maximum braking value

W N

Fig. 17.5 Forces in a rejected takeoff.

824 Operational Aircraft Performance and Flight Test Practices

(0.35 to 0.5 on dry hard surfaces). We will expand on this in the braking
section of this chapter. (W 2 L) represents the airplane weight minus
the airplane lift, depending on the center of gravity location and dynamic
condition. Part of this force acts on the main landing gear where the
brakes are generally located.
For the transition phase, the pilot reacts to an event (engine failure,
warning in the cockpit, etc.) and decides to reject the takeoff. The pilot
must then reduce the thrust and apply the brakes and any other approved
deceleration devices to stop the airplane within the remaining runway dis-
tance. This must be done swiftly, especially at high speed. Part 23 and
Part 25 certification require the crew to use an approved procedure (one
that has been demonstrated to provide reliable and consistent results).
§25.109 Accelerate-stop distance

(a) Except as provided in paragraph (f)(1) of this section, means other than
wheel brakes may be used to determine the accelerate-stop distance if
that means—
(1) Is safe and reliable;
(2) Is used so that consistent results can be expected under normal
operating conditions; and
(3) Is such that exceptional skill is not required to control the airplane.
(b) The effects of available reverse thrust—
(1) Shall not be included as an additional means of deceleration when
determining the accelerate-stop distance on a dry runway; and
(2) May be included as an additional means of deceleration using
recommended reverse thrust procedures when determining the
accelerate-stop distance on a wet runway, provided the requirements
of paragraph (e) of this section are met.
(c) The landing gear must remain extended throughout the accelerate-stop
distance.

As well, the pilot actions must account for reasonable reaction time between
events:
§25.101 General

(f) Unless otherwise prescribed, in determining the accelerate-stop

distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must be
made in accordance with procedures established by the applicant for
operation in service.
(h) The procedures established under paragraph (f) and (g) of this section
must—
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.

Let’s use an engine failure condition as the baseline for our discussion, with
the failure occurring at VEF . After a certain delay time, depending on how
suddenly the engine loss occurs (a large heading rate of change that requires
a large rudder deflection is a clear indication) and on the kind of cockpit
annunciation, the pilot will recognize that he or she has lost the thrust
 CHAPTER 17 Rejected Takeoff Performance 825

⌬tact1 ⌬tact2 ⌬tact3

VEF V1

Fig. 17.6 Transition time modeling.

from one engine, decide to reject the takeoff, and apply the first deceleration
device. This reaction time is defined as Dtact1 . The first action is
usually the application of brakes (but again, it is defined by the applicant
through an approved procedure to be used by the crew). Any follow-on
pilot action takes a finite amount of time, which can be defined as time
from the previous action of time or as time intervals from the first action
(Fig. 17.6).
Dtact1 ¼ The demonstrated time between the engine failure and
the first action by the pilot. (For AFM expansion, this
value must not be smaller than 1 s; we will discuss this
more in Chapter 18.)
Dtact2 , Dtact3 ¼ The subsequent demonstrated time intervals (for thrust
cuts and deploying spoilers).
Note that these pilot event time intervals (Dtact2 , Dtact3 ) typically remain
the same whether the rejected takeoff is from an AEO condition or an OEI
condition. The pilot should not have to memorize different reject procedures
for an OEI or AEO condition.
In the transition phase of an RTO, from VEF to full braking configuration,
the thrust from the failed engine will decay rapidly. This can be modeled as a
ramp response from the thrust just before the engine failure to windmilling
drag in a given amount of time (see Fig. 17.7). This decay, for the most adverse
thrust loss, is very rapid and often of the order of 1 s.
The other engine(s) remain at maximum thrust until the pilot elects to
stop and brings the engines to idle by throttle chop (rapid movement of
the throttles). The throttle chops for the other operating engine(s) can be
modeled in a similar way as the engine failure case, but in this case, the
thrust decay is a reduction from the maximum thrust to idle and over a
longer period of time (slower decay); see Fig. 17.7. Note that for propeller-
driven airplanes, the idle setting may actually produce negative thrust
(discussed further later in this chapter). Whichever model is used, one
must best match the change in distance/acceleration with time.
The next system we usually include in the RTO is the speedbrakes/
spoilers. The first one increases drag only (can be located on the fuselage),
whereas the latter increases drag and reduces lift. These devices generally
826 Operational Aircraft Performance and Flight Test Practices

100

Failed engine model

80

60
% Thrust

40
Actual thrust decay
20
Thrust
0
Drag
Model
–20
0 1 2 3 4
Time
100
90
Throttle chop model
80
70
60
% Thrust

50
Actual thrust decay
40
30
20
10
Model
0
0 2 4 6 8 10
Time

Fig. 17.7 Example of failed engine and throttle chop modeling.

fall into the acceptable category for use in a rejected takeoff when they meet
the requirements of §25.109(e):
§25.109(e) Except as provided in paragraph (f)(1) of this section, means
other than wheel brakes may be used to determine the accelerate-stop dis-
tance if that means—
(1) Is safe and reliable;
(2) Is used so that consistent results can be expected under normal
operating conditions; and
(3) Is such that exceptional skill is not required to control the airplane.
To see the impact of these devices on the airplane’s forces affecting accelera-
tion, we propose an example of the deployment of spoilers (which impact
 CHAPTER 17 Rejected Takeoff Performance 827

both lift and drag) 1 s after a model airplane has crossed 120 KTAS in an
RTO (thrust already at idle, in this case still producing 2000 lb of forward
force); see Fig. 17.8. Spoilers can easily deploy at 30 to 60 deg/s; for this
example, we selected 45 deg/s with a maximum deflection angle of 60 deg.
We look at one case where the pilot has yet to apply the brakes (rolling
m ¼ 0.02) and another where braking has started (braking m ¼ 0.40). For
this example, the baseline airplane model takes on the following form:
W ¼ 100,000 lb; lift and drag coefficients in the takeoff configuration ¼ 0.5
and 0.06, respectively; impact on spoilers on lift and drag coefficients ¼ –
0.3 and þ0.02, respectively.
What we observe:
• The time to deploy the spoilers is relatively short (1.25 s in this case), so
modeling the transient aerodynamics as a ramp (as it was here) or step
input may have limited impact on the overall distance model. (Spoilers’
impact on aerodynamics is not linear in the first place; the final model is
matched to the demonstrated test distance.)
• Although no brakes are applied, the drag increase due to the spoilers
deploying improves deceleration by about 0.5 ft/s2 .
• With brakes applied, overall deceleration improves by 2 ft/s2 .
• We note that 0.5 ft/s2 is contributed by the increased drag; the
remainder is from the increased weight on the wheels improving the
braking force.
• Note the magnitude of the braking force increase in Fig. 17.8c. Increased
weight on wheels will greatly increase deceleration forces on most
runway surfaces.
One feature that also provides a means of slowing the airplane down is
reverse thrust. This consists of changing the direction of the thrust force
by redirecting the air flow. On turbojet and turbofan airplanes, this is done
using cascades, clamshells, and/or buckets (see Fig. 17.9). On propeller-
driven airplanes, reverse thrust is achieved by changing the blade pitch
angle from positive to negative.
The additional retarding force from thrust reversers will reduce the stop-
ping distance for the airplane compared to not using them. The requirements
of §25.109 state that no credit on dry runway should be used for the purpose
of computing the RTO distance; however, credit can be used on wet (and
contaminated) runways:
25.109(f) The effects of available reverse thrust—

(1) Shall not be included as an additional means of deceleration when

determining the accelerate-stop distance on a dry runway; and
(2) May be included as an additional means of deceleration using
recommended reverse thrust procedures when determining the
accelerate-stop distance on a wet runway, provided the requirements of
paragraph (e) of this section are met.
 828
Operational Aircraft Performance and Flight Test Practices
Brakes not applied Brakes not applied
a) 5000 b) 125 0
120 –0.2

True airspeed (kt)

4000 115
Forces (lb) –0.4

Decel (fps/s)
3000 110 –0.6
105 –0.8
2000 100
95 –1
1000 90 –1.2
85 –1.4
0 80 –1.6
0 1 2 3 4 5 0 1 2 3 4 5
Time (s) Time (s)
Drag Friction force Thrust True airspeed Decel (fps/s)

c) Brakes applied d) Brakes applied

40,000 125 0
35,000 120 –2

True airspeed (kt)

30,000 115
Forces (lb)

Decel (fps/s)
110 –4
25,000 –6
20,000 105
100 –8
15,000 95
10,000 –10
90
5000 85 –12
0 80 –14
0 1 2 3 4 5 0 1 2 3 4 5
Time (s) Time (s)
Drag Friction force Thrust True airspeed Decel (fps/s)

Fig. 17.8 Impact of spoilers’ deployment on retarding forces.

 CHAPTER 17 Rejected Takeoff Performance 829

Cascades Clamshells

Forward
flight

Reverse thrust
Buckets

Fig. 17.9 Turbojet/turbofan reverse thrust mechanisms.

Of course, credit can only be used if the requirements of §25.109(e) are met.
One concern, of course, is asymmetric thrust conditions under OEI (one of
the reasons to RTO). Will the airplane maintain directional control? The
loss of directional control authority can come in part from tail blanking
(for fuselage-mounted engines especially) where the reverse thrust plume
alters the aerodynamic authority of the tail. This may force the manufacturer
to adopt a reduced thrust value and possibly even have a cutoff speed below
which reverse thrust is not authorized above idle. This effect can be worse for
AEO where both sides of the tail are blanked. FAA AC 23-8C notes:

Demonstration of full single engine reverse controllability on a wet runway

and in a 10-knot adverse crosswind will be required. Control down to zero
speed is not essential, but a cancellation speed based on controllability can
be declared and credit given for use of reverse thrust above that speed.
The use of reverse thrust on one engine on a wet runway requires that the
reverse thrust component be equally matched by a braking component
and rudder use on the other side. If brake modulation is required and
allowed for steering, then the reduced braking must be accounted for in
the stopping performance.

Reverse thrust may also create issues with nose-up pitching moment (engine
thrust line above vertical CG) and loss of nose wheel ground contact may
become a limiting issue for the maximum reverse thrust that can be used.
It should be noted that it takes a lot of time to spool down the engines
from takeoff power to idle prior to selecting reverse thrust and spooling
the engines back to max reverse thrust. The benefit of reverse thrust on
830 Operational Aircraft Performance and Flight Test Practices

dry runways (including the possible limitation of maximum value due to

handling) is often not significant. The impact of reverse thrust on wet
runways can be important and on contaminated runways even more (see
Fig. 17.10) due to the longer time required to slow down (smaller average
deceleration). Proper rejected takeoff procedure should define how the
reverse thrust can be used.

Thrust

Takeoff

Forward
idle Time
Reverse
Max selected
reverse
RTO
started Reverse reduced
for handling

80
Runway
Average decel condition

70

0.1 g
60
Contaminated
Time to full stop (s)

50

40

0.2 g Wet
30

0.3 g
20
0.4 g Dry

0.45 g
10

0
0 20 40 60 80 100 120 140 160
True airspeed (knots)

Fig. 17.10 Reverse thrust modeling vs time.

 CHAPTER 17 Rejected Takeoff Performance 831

Off Off
C C
o o
n n
t t
r
r
o
1 2 o
l
l R R
l a P a l Power levers
o t o t o
c i w i c
k n e n k
g r g
On On
Flight
Flight idle gate
Beta range – propeller Idle
Blade angle controlled
by power lever position Disc
Propeller discing detent

Max
Rev Maximum reverse power

Fig. 17.11 Example of throttle quadrant with discing option. Source: http://
reports.aviation-safety.net/2014/20141003_DH8D_C-FSRN.pdf

On piston-prop and turboprop airplanes, the propeller pitch is most

likely controlled to a minimum flight value (say 20 deg) to prevent a high
drag condition and possible windmilling of the operating engine. In this con-
dition, the propeller/power setting combination can still lead to creating drag
instead of thrust, as discussed in Chapter 8. Once on the ground (weight on
wheels sensed), these airplanes can adopt a propeller discing condition (finer,
still positive, blade pitch) to create more drag (RTO and landing) by pilot
movement of the propeller/power setting into the beta range (propeller
pitch control) without going into the reverser thrust (negative propeller
blade pitch angles); see Fig. 17.11. That extra pilot input takes time and
must be modeled (verified by test). On electrically powered airplanes with
regeneration capability, the propeller position can be allowed to achieve
discing in flight and even allowed to windmill and “absorb” flight energy to
feed back to the batteries. (An example of this condition is the Pipistrel
Alpha Electro.) These airplanes would therefore have no additional delays
in the movement of the throttle to the drag-producing condition (simple
throttle movement to idle).
The guidance of FAA AC 23-8C specifies the following on this subject:
For propeller equipped aircraft, the critical engine’s propeller should be in
the position it would normally assume when an engine fails and the power
lever is closed. The highdrag position (not reverse) of the remaining
engines’ propellers may be utilized provided adequate directional control
832 Operational Aircraft Performance and Flight Test Practices

can be demonstrated on a wet runway. Simulating wet runway controllability

by disconnecting the nose wheel steering may be used. The use of the higher
propeller drag position, that is, ground fine, is conditional on the presence of
a throttle position that incorporates tactile feel that can consistently be
selected in service by a pilot with average skill. It should be determined
whether the throttle motions from takeoff power to this ground fine position
are one or two distinctive motions. If it is deemed to be two separate
motions, then accelerate-stop time delays should be determined accordingly
and applied to expansion of data.

Thus, the airplane’s handling characteristics under asymmetric thrust/drag

condition must be acceptable on a wet runway [conservatively simulated
with nose wheel steering (NWS) off] so as to be used (available) under all
conditions and credit for the deceleration force provided to be used to
define a RTO distance. If the handling is not acceptable, a higher (e.g.,
flight idle) minimum thrust condition must be used.
To show the complexity of the transition segment, we offer this example
(see Fig. 17.12). This is an OEI case. The RTO event starts with an engine
failure at time 1 s. The thrust decay for the failed engine is similar to the
model shown in Fig. 17.7. As the thrust decays, so does the acceleration of
the airplane, as seen in the top chart. The pilot reacts to the event by starting
brake application at time 1.6 s; full brake force is achieved by time 2.6 s
(modeled by a rolling friction force coefficient changing from 0.02 to 0.45).
At time 2.0 s, the pilot selects idle on the remaining engine, which will
take 4 s to spool down. The spoilers are selected next at time 3 s and take
1.3 s to deploy (as shown by the increasing drag coefficient). By time 4.3 s,
the aerodynamic configuration of the airplane is stable, and the brakes are
fully ramped up. One could select this as the end of the transition phase or
one can extend the transition phase to the point where the thrust on the
second engine is spooled down to idle, by which time, as seen on the top
chart of Fig. 17.12, the deceleration is stable. This author selects the point
where spoilers are fully deployed and brakes are providing maximum force
as the end of the transition phase and carries the thrust decay (lesser
impact, well-behaved model) into the next phase. Beyond the transition
phase, the airplane deceleration is in full braking mode, which we will
describe later in this chapter.
The events for an AEO RTO would have been similar to the ones of
Fig. 17.12 with the exception of no engine failure (both engines being
brought to idle at the same time).

Braking Force
The brakes provide by far the largest retarding force required to stop an
airplane on most runway conditions. This force is proportional to the
weight on wheels. At near-zero airspeed, with the lift force almost zero,
the retarding force generated by full brake application can be as much as
50% of the aircraft weight. The braking friction force coefficient typically
 V a
(KIAS) (g)
120 0.4
0.2
100
(KIAS) 0.0

(g)
80

a
V

–0.2
60 –0.4
40
Thrust R Thrust L
(lb) (lb)
20
Thrust L (lb) (103)

Engine
to idle
15
Thrust R (lb)

Engine
to idle
10
Engine
failure

5
0

CHAPTER 17
–5

deployed
Spoilers
D
CD

selected
(lb)

Spoilers
3000 0.070
2500 0.065
0.060
2000

CD
0.055
(lb)
D

Rejected Takeoff Performance

1500 0.050
1000 0.045
0.040
brake

f
Full

(lb) mu
60 0.6
Start of
brake
(lb) (103)

40 0.4

mu
f

20 0.2
0 0.0
0 1 2 3 4 5 6 7 8 9 10
Time (s)

Fig. 17.12 RTO transition phase example.

833
834 Operational Aircraft Performance and Flight Test Practices

varies from 0.35 to 0.5 (dry concrete and asphalt). Such braking coefficients
can provide decelerations of 11 to 16 ft/s2 . Rewriting Eq. (17.2), the braking
force is

fB ¼ mB ðW  LÞ  mB W ð17:3Þ

Expressed in this later form, mB can be viewed as a drag force divided by

the weight applied onto the tires. Of course, not all of the weight of the air-
plane is applied onto the tires providing the braking force. Brakes are typically
only on the MLG and the MLG does support a very large percentage of the
airplane weight. For example, Fig. 17.13 indicates that for the Boeing
727-100/100C, between 89% and 97% of the airplane weight will rest on
the MLG under static conditions. The airplane braking performance model
typically does not use weight on MLG as an input, but rather total airplane
weight. Just like we discussed for the stall speed, the braking performance
must be determined for the most adverse CG location—in this case, the
least weight on wheels condition (§25.109), which is forward CG for a tra-
ditional tricycle landing gear configuration.
Under nonstatic conditions, the location of the vertical CG and the
various thrust vectors changes the dynamics of the MLG vs the NLG
loading. Braking tends to increase the load on the NLG and decrease it on
the MLG (see Fig. 17.14). Thrust reversers on configurations with the
thrust line above the CG reduce NLG load and increase MLG load (improved
overall braking, but possible loss of nose wheel steering capability if the load
on the NLG is too low).
The braking force is created by applying a torque to the wheel that will
make it slip compared to the movement of the aircraft (i.e., wheel speed is
slower than aircraft ground speed). The braking coefficient will be
maximum for a given percentage (usually between 10% and 15%) of wheel
slip as measured by the ratio of wheel speed to aircraft ground speed (see
Fig. 17.15).
On small airplanes, the pilot can modulate the brakes to apply the proper
braking force and may reach maximum braking performance if enough
torque is applied to the wheels; however, this may cause the wheel to slip
more and lose braking power. On larger airplanes or airplanes with higher
ground speed, manual braking is not practical for normal use because it
requires the pilot to modulate the brakes to just below skid (locked
wheels) to achieve the performance, which risks damage to the tire or even
tire burst. One then either introduces an anti-skid system and/or provides
a system-limiting torque value at the wheels.
An antiskid system, first introduced to aviation by Gabriel Voisin in 1920
(initial concepts), will get the wheels to the maximum braking force the
tire-to-ground contact can provide and then reacts to the wheel starting to
skid by releasing brake pressure to allow the wheel to reaccelerate before
 Note: unshaded area represents operational limits
Percent mac
0 5 10 15 20 25 30 35 40 45
180 180

170 170

Maximum design taxi weight

170,000 lb (77,100 kg)
160 161,000 lb (73,000 kg) 160
153,000 lb (69,400 kg)

150 150
Load on main landing gear 1000 pounds

140 140
VNG = maximum vertical nosegear ground load at most forward CG

CHAPTER 17
1000 pounds
VMG = maximum vertical maingear ground load at most AFT CG
130 130 H = maximum horizontal ground load from braking
Note:
All loads calculated using airplane maximum design taxi weight
120 120

Rejected Takeoff Performance

110 110 H
VNG VMG
CG for ACN
calculations
VNG VMG per strut (4) H per strut (4)
100 100
Maximum
Model design taxi Static plus force Maximum load At steady braking At instantaneous
Static at most
weight due to braking at occurring at 10 ft/s2 braking (coeff of
forward CG
90 90 most forward CG static AFT CG deceleration friction = 0.8)
lb kg lb kg lb kg lb kg lb kg lb kg
–100 153,000 69,400 16,300 7400 24,000 10,900 73,200 33,200 23,800 10,800 58,600 26,600
80 80
80 85 90 95 100 –100C 161,000 73,000 17,100 7800 25,300 11,500 77,000 34,900 25,000 11,300 61,600 27,900
Percent of weight on main gear

Fig. 17.13 Load on main gear for a Boeing 727-100/100C. Source: https://www.boeing.com/resources/boeingdotcom/commercial/airports/acaps/727.pdf

835
836 Operational Aircraft Performance and Flight Test Practices

Static (no thrust) Dynamic (with thrust)

V=0 V>0
X2
X1 L Mf = yf f

D T yf
f
N2 W N1 N2 w N1

Fig. 17.14 Forces and moments impacting weight on wheels for MLG.

120
Wheel slip

100
Ground speed
Ground speed (kts)

80

60
Wheel speed
40

20

0
0 5 10 15 20 25 30 35 40
Time
0.8
Friction coefficient

Dry surface

Wet surface Icy surface

0
0.0 0.2 0.4 0.6 0.8 1.0
Antiskid control region Wheel slip
for maximum efficiency

Fig. 17.15 Wheel slip during an RTO.

 CHAPTER 17 Rejected Takeoff Performance 837

applying brake pressure again, repeating the cycle. There are three major
types of antiskid systems to consider:

• On-off type: The early design for this type of system consisted of full release
of the brake pressure when wheel skid was detected followed by
reapplication of full pressure. This type would be effective in preventing
wheels from locking and would minimize tire damage, but it is somewhat
less efficient in producing a good average deceleration force.
• Quasi-modulation: The next generation of antiskid system modulated the
brake pressure to a wheel speed deceleration threshold used to estimate
the start of a wheel skid. Once the wheel speed rate crossed the threshold,
the pressure would reduce to allow the wheel to spin up again, and then the
pressure is applied to just below the computed brake pressure for the skid.
This system is more efficient than the on-off type.
• Fully modulating: These systems use wheel speed monitoring to modulate
brake pressure reduction and application based on the rate the wheel is
going into or recovering from a skid. These have been shown to be the
most efficient antiskid brake systems.

An antiskid system on a new airplane will need tuning to provide

acceptable performance for a given airplane. This must be done at the
extremes of the weight and CG envelope (light and heavy weight forward
and aft CG). The performance engineer will therefore be able to get a first
look at the braking performance prior to going into a braking distance per-
formance campaign.
The braking system could also be designed to provide a reduced force by
controlling the maximum torque applied to the wheel for braking on a dry
runway (Fig. 17.16). This reduced torque value compared to the maximum
torque at wheel skid may be used to limit the force fed back into the structure
or may be a feature designed to provide smoother deceleration when the
runway length is long. This feature may still lead to wheel skid when the

␮B
Typically no wheel
skids in this region
␶max

Dry Wet Ice

Fig. 17.16 Torque-limited braking force.

838 Operational Aircraft Performance and Flight Test Practices

runway surface friction is degraded (wet or contaminated) and thus still

require the use of an antiskid system.

Braking Distance
To compute the ground run distance during the braking phase, we can
rewrite Eq. (15.7) from Chapter 15, now accounting for lower thrust and
higher drag in the following form (starting at the point the performance
engineer declares full braking and ending at a ground speed of 0):

W
ð XB ð0 ðVTAS  VW Þ dV
g
XB ¼ dX ¼  
0 VB T  mB W  12 rSL s VTAS
2 S CD;TO  mB CL;TO
ð17:4Þ

Note that the equation now includes the braking coefficient mB and that
the lift and drag coefficients should represent those of the airplane with
deceleration devices deployed. To solve Eq. (17.4), we will need to make
some assumptions. They are:
• The winds are constant, both in magnitude and direction, during
the deceleration.
• The weight of the plane is considered constant during the ground run.
• The braking coefficient can be considered constant.
• The model created must match the distance demonstrated in flight tests.
Sometimes the braking coefficient can be a function of ground speed or
even energy absorbed into the brakes.
With this, Eq. (17.4) can now be solved numerically. And just like we did for
the acceleration part, another way to write Eq. (17.4) is

ð0
ðVTAS  VW Þ dV
XB ¼ ð17:5Þ
VB a

where a is the airplane deceleration.

  
g T  mB W  12 rSL s VTAS
2 S CD;TO  mB CL;TO
a¼ ð17:6Þ
W

We can make one further simplification to the takeoff distance and accel-
eration equations. From Fig. 17.12, it can be observed that the retarding
forces in the braking section have achieved maximum values and do not
change much during the full braking phase. They are of the order of the
 CHAPTER 17 Rejected Takeoff Performance 839

braking force coefficient applied to the airplane’s weight. Therefore, if we

neglect the aerodynamic forces of drag and lift, Eq. (17.6) simplifies to
 
T
ag  mB ð17:7Þ
W

For zero winds, this leads to a simplified braking distance of (negative to

make distance positive):

V 2
XB   B  ð17:8Þ
T
2g  mB
W

Runway Surface Conditions

Maximum available braking friction coefficient on dry runway is based on
multiple variables that include:
• Surface texture
• We only address hard surfaces in this discussion (asphalt or concrete
runways).
• Tire compound
• Tire pressure (higher results in less friction)
• Ground speed
• Type of antiskid system (if installed) and any reduction due to torque limits
• Weight on the wheels with the brakes
• Impact of CG location
• Impact of wing lift
On dry hard surfaces, the maximum expected braking coefficient at the
tire-to-ground contact can be expected to be around 0.5 to 0.6 (50– 60% of
the weight on wheels) over the course of an RTO. If one considers that the
test is done at maximum forward CG (most adverse condition), then that
value typically reduces by about 10% (only 90% or so of the airplane weight
on wheels), so values of 0.45 to 0.54 would be the maximum expected. If
one then considers that the braking force with the vertical CG above the
runway surface pushes more load on the NLG, then those values can be
expected to fall between 0.4 and 0.5. Finally, if the airplane is not equipped
with a system that reduces the wing lift and thus reduces the weight on
wheels, the expected braking coefficient as used in the second part of Eq.
(17.3) could actually fall closer to 0.35. The actual value is determined/vali-
dated by flight testing.
On a wet runway, there will be a film of water between the runway and the
tire that needs to be displaced to provide a braking force, and some of the tire
will lose contact with the runway while displacing the film, reducing the
840 Operational Aircraft Performance and Flight Test Practices

0.7 Dry
Wet
0.6 Flooded
Braking friction coefficient
0.5

0.4

0.3

0.2

0.1

0
0 20 40 60 80 100 120
Ground speed (kts)

0.6 Dry runway

Wet runway
Effective braking friction coefficient

0.5 Flooded runway

0.4

0.3

0.2

0.1

0
0 20 40 60 80 100
Ground speed (kts)

Fig. 17.17 General brake coefficient trends vs ground speed: (a) fixed weight on wheels,
(b) Boeing 737– 100 [4].

braking force. At low ground speed, there will be little loss of friction, but as
the airplane travels faster, more tire-to-runway contact will be lost, and there
will be a further degradation of the braking performance.
A runway is considered flooded (more than wet, standing water) when
more than 1/8 in. of water is on the runway surface. The water film is con-
siderably larger, and significant loss of performance is expected (see
Fig. 17.17).
 CHAPTER 17 Rejected Takeoff Performance 841

The FAA worked with NASA to define some expected maximum braking
coefficients on wet runways that were later codified and introduced in
Amendment 92 (1998), the latest revision of §25.109. The requirement states:
25.109 (c) The wet runway braking coefficient of friction for a smooth wet
runway is defined as a curve of friction coefficient versus ground speed
and must be computed as follows:
(1) The maximum tire-to-ground wet runway braking coefficient of friction
is defined as:

Tire
Pressure Maximum braking coefficient
[psi] (tire-to-ground)
 3  2 
50 V V V
mt =gmax ¼ 0:0350 þ 0:306  0:851 þ 0:883
100 100 100
 3  2 
V V V
100 mt=gmax ¼ 0:0437 þ 0:320  0:805 þ 0:804
100 100 100
 3  2 
V V V
200 mt =gmax ¼ 0:0331 þ 0:252  0:658 þ 0:692
100 100 100
 3  2 
V V V
300 mt=gmax ¼ 0:0401 þ 0:263  0:611 þ 0:614
100 100 100

Where—
Tire Pressure ¼ maximum airplane operating tire pressure (psi);
mt =gmax ¼ maximum tire-to-ground braking coefficient;
V ¼ airplane true ground speed (knots); and
Linear interpolation may be used for tire pressures other than those
listed.
(2) The maximum tire-to-ground wet runway braking coefficient of friction
must be adjusted to take into account the efficiency of the anti-skid
system on a wet runway. Anti-skid system operation must be
demonstrated by flight testing on a smooth wet runway, and its efficiency
must be determined. Unless a specific anti-skid system efficiency is
determined from a quantitative analysis of the flight testing on a smooth
wet runway, the maximum tire-to-ground wet runway braking
coefficient of friction determined in paragraph (c)(1) of this section must
be multiplied by the efficiency value associated with the type of anti-skid
system installed on the airplane:

Type of anti-skid system Efficiency value

On-Off 0.30
Quasi-modulating 0.50
Fully modulating 0.80

Thus,
mTG ¼ efficiency  mTGmax
842 Operational Aircraft Performance and Flight Test Practices

0.9
Tire-to-ground max friction coefficient

0.8
50
0.7 p
10 si
0p
0.6 20 si
0p
30 si
0.5 0p
si
0.4

0.3

0.2

0.1

0
0 20 40 60 80 100 120 140 160 180
Ground speed (kt)

Fig. 17.18 Tire-to-ground tire friction coefficient of §25.109.

The data of §25.109, illustrated in Fig. 17.18, applies for smooth runways
(asphalt or concrete) where there is no special preparation of the runway
surface to help the water film move away from between the tire and the
surface. Again, based on actual weight on wheels and the braking coefficient
must be reduced proportional to the expected weight on the wheels with the
braking system. The value of the braking force is validated by testing.
We pause here to state how important information sharing is in the
industry. We have the case where the student becomes the teacher. This
author had a chance to meet Mr. Yager, one of NASA’s leads for brake
performance testing when he came to the Kansas University Operational
Aircraft Performance and Flight Test Practices class to learn about the
overall performance aspect, and we had a chance to share knowledge
about braking performance testing done by NASA. We always strive to
keep learning on the subject.

Grooved Runways
Airplanes with high takeoff and landing speeds face a major challenge on
wet runways in terms of braking distance. NASA initiated testing in the late
1960s to find ways to help remove the water film between the tires and the
runway by providing better drainage. They came up with the concept of pro-
viding transverse grooves on the runway of specific configurations that
greatly improved overall braking performance (see Fig. 17.19).
 CHAPTER 17 Rejected Takeoff Performance 843

Surface
A
B
.6 Wet with isolated puddles C
D
E
.5 Dry F
G
Braking friction coefficient
H
I
.4

.3
Grooved
.2

.1
Ungrooved

0 20 40 60 80 100 120 140

Ground speed, knots

Surface A Canvas-belt drag finished concrete, ungrooved

Surface B Canvas-belt drag finished concrete, grooved
Surface C Burlap drag finished concrete, grooved
Surface D Burlap drag finished concrete, ungrooved
Surface E Gripstop transition surface
Surface F Small-aggregate asphalt, ungrooved
Surface G Small-aggregate asphalt, grooved
Surface H Large-aggregate asphalt, grooved
Surface I Large-aggregate asphalt, ungrooved

Fig. 17.19 Impact of providing grooves on runway friction for wet runways [3].

The FAA advisory circular for grooved runways that provides guidance to
airport operators on the acceptable design for a grooved runway is FAA AC
150/5320-12D, Measurement, Construction, and Maintenance of Skid-
Resistant Airport Pavement Surfaces. The FAA standard groove configur-
ation is 14 in. (+1/16 in.) in depth by 14 in. (þ1/16 in., 20 in.) in width by
1.5 in. (–1/8 in., þ0 in.) center-to-center spacing. The grooves need to be
continuous for the entire runway length and transverse (perpendicular) to
the direction of airplane landing and takeoff operations.

Contaminated Runways
Runways with significant water (more than 1/8 in., flooded) are con-
sidered contaminated and have a more important impact on the ability of
the airplane to stop. We introduced several other contaminants in Chapter
15. EASA CS25 AMC 25.1591 introduces expected maximum average
844 Operational Aircraft Performance and Flight Test Practices

Table 17.1 Default Braking Coefficient vs Contaminants

Contaminant Default Friction Value m

 
Standing Water and Slush V 3 V 2
¼  0:0632 þ 0:2683 þ
100 100

V
 0:4321 þ 0:3485
100
where V is groundspeed in knots
Note: For V greater than the aquaplaning speed,
use m ¼ 0.05 constant
Wet Snow below 5mm depth 0.17
Wet Snow 0.17
Dry Snow below 10mm depth 0.17
Dry Snow 0.17
Compacted Snow 0.20
Ice 0.05
Note: Braking Force ¼ load on braked wheel  Default Friction Value m. Source: EASA AMC25.1591.

braking coefficient as a function of runway contaminant (see Table 17.1).

This friction coefficient is again based on weight on wheel.

Worn Brakes
The FAA introduced requirements to test an airplane with worn brakes
and provide performance distance based on fully worn brakes in Part 25,
Amendment 92. Per FAA AC 25.735:
On May 21, 1988, an American Airlines DC-10 experienced an 86 percent
maximum kinetic energy (KE) rejected takeoff (RTO) in a dispatch configur-
ation in which eight of the ten brakes were worn close to the maintenance
limits. The eight brakes failed in the early portion of the braking run and
the airplane overran the runway. As a result, the FAA reviewed the method-
ology used in the determination of allowable brake wear limits for transport
category airplanes. It was determined that brake wear limits should be estab-
lished during certification to ensure that fully worn brakes will function
properly during a maximum KE RTO. The FAA issued a series of airplane
specific airworthiness directives between 1989 and 1994 to establish brake
wear limits using the new criteria.

Changes in Amendment 92 that included the requirement are found in

§25.101 and §25.109. They also impact landing brake performance in
§25.125. The OEM must test at various wear limits, or simply as close as poss-
ible to full worn condition, to determine the braking capability of the braking
system. Worn brakes will be thinner (less mass) and not capable of absorbing
as much energy as new brakes. Some brakes may show signs of brake fading
(less braking force) as more and more energy is absorbed during a stop
(Fig. 17.20).
 CHAPTER 17 Rejected Takeoff Performance 845

0.55

0.5
Model

␮B 0.45

0.4

0.35
KE

Fig. 17.20 Impact of absorbed energy into the brake on braking coefficient (brake fading).

§25.101 General
(i) The accelerate-stop and landing distances prescribed in Secs. 25.109
and 25.125, respectively, must be determined with all the airplane
wheel brake assemblies at the fully worn limit of their allowable
wear range.

§25.109 Accelerate-stop distance

(i) A flight test demonstration of the maximum brake kinetic energy
accelerate-stop distance must be conducted with not more than 10
percent of the allowable brake wear range remaining on each of the
airplane wheel brakes.

We will address maximum brake energy in Chapter 18.

Stopway
FAA 14 CFR Part 1 defines a stopway as:
. . .an area beyond the takeoff runway, no less wide than the runway and
centered upon the extended centerline of the runway, able to support the
airplane during an aborted takeoff, without causing structural damage to
the airplane, and designated by the airport authorities for use in decelerating
the airplane during an aborted takeoff.

A pilot can therefore compute an accelerate-stop distance by summing the

available runway length and the stopway length (the accelerate-stop distance
available, ASDA). This calculation, along with other factors, is used by the
pilot to determine airplane loading and performance requirements. A
stopway can therefore be used as a clearway (see discussion in Chapter 15)
as well (see Fig. 17.21).
To be considered a stopway, the airport operator must commit to main-
taining the surface as such. The FAA recognizes that stopways may be of little
economic benefit as compared to having a full runway.
846 Operational Aircraft Performance and Flight Test Practices

CLRWY

LDA Clearway

500 ft ≈ 150 m
Stopway
Runway

EMAS
TORA STPWY

ASDA
Significant
TODA obstacle

Fig. 17.21 Clearway, stopway, RSA, and EMAS vs various takeoff distances.

Runway Safety Area (RSA)

A runway safety area (RSA) is typically 500 ft wide and extends 1000 ft
beyond each end of the runway. It provides a graded area in the event
that an airplane overruns, undershoots, or veers off the side of the
runway. The FAA requirements for an RSA are defined in FAA AC
150/5300-13a, which identifies that about 90% of overruns were less
than 1000 ft in length. An RSA cannot be used to compute takeoff
distances.
On 31 Dec. 2015, the FAA successfully completed a 15-yr effort to
improve RSAs at commercial service airports. More than 1000 runway
ends at 500 airports were improved to offer enhanced safety at the nation’s
airports [4].
As specified by the FAA, a stopway is an area beyond the runway with suf-
ficient strength to support a decelerating aircraft in all-weather conditions. It
is not an RSA and may not be used in lieu of an RSA. The RSA begins at the
end of the stopway, and the RSA may not be shortened to accommodate
a stopway.

Engineered Materials Arresting Systems (EMAS) for Airplane

Overruns
EMAS technology provides safety benefits in cases where land is not
available or it’s not possible to have the standard 1000-ft overrun. A standard
EMAS installation can stop an airplane from overrunning the runway at
approximately 80 mph. An EMAS arrestor bed can be installed to help
slow or stop an airplane that overruns the runway, even if less than a standard
RSA length is available. EMAS uses crushable material placed at the end of a
 CHAPTER 17 Rejected Takeoff Performance 847

runway to stop an airplane that overruns the runway. The tires of the airplane
sink into the lightweight material, and the airplane is decelerated as it rolls
through the material.
Specification for EMAS are defined by advisory circular 150-5220-22B,
Engineered Materials Arresting Systems for Aircraft Overruns.
Figure 17.21 shows the location of an EMAS with respect to the declared
runway distance and possible stopway. The ICAO also defines EMAS in
Annex 14, EMAS.

NTSB DCA10IA022, US Airways Express flight 2495, rejected takeoff,

Bombardier CL-600-2B19, N246PS, January 19, 2010: A PSA Airlines
Bombardier CRJ200 departing Charleston to Charlotte, NC, carried out a
high-speed rejected takeoff and overran the end of runway 23, stopping in
the Engineered Materials Arresting System (EMAS).
The investigation revealed that the correct flap setting (Flap 20) for the
speeds calculated (V1 ¼ 127 kt, Vr ¼ 128 kt) was called out during taxi,
but the crew selected Flap 8. The crew also selected FLEX 33 (reduced
thrust takeoff). According to the FDR, during the takeoff acceleration, 10 s
beyond the crew 80-kt callout, the flaps started moving from 8 to 20 (no
crew call on CVR); the airspeed was 120 kt. The first officer called “V one”
at about 127 kt, and 2 s later the master warning and flap configuration
aural alerts sounded. The captain initiated a rejected takeoff, and the airspeed
reached a peak of 143 kt prior to slowing down. The airplane exited the
runway at about 50 kt and entered the EMAS, where it stopped in about
3 s and 128 ft. The airplane suffered only minor damage to the gears and
flaps.
848 Operational Aircraft Performance and Flight Test Practices

References
[1] Federal Aviation Administration, FAA Training Guide, Section 2: Pilot Guide to
Takeoff Safety, https://www.faa.gov/other_visit/aviation_industry/airline_opera
tors/training/media/takeoff_safety.pdf
[2] Federal Aviation Administration, CFR Final Rule, Federal Register, Vol. 63, No. 32, 18
Feb. 1998, https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgFinalRule.nsf/
a09133bddc7f4fbb8525646000609712/b1759d97ff77f95086256936004d6e44! Open
Document [retrieved 11 Oct. 2020].
[3] Yager, T. J., “Comparative Braking Performance of Various Aircraft on Grooved and
Ungrooved Pavements at the Landing Research Runway,” NASA Wallops Station, 1
Jan. 1969.
[4] Federal Aviation Administration, “Fact Sheet’Runway Safety,” 5 July 2018, https://
www.faa.gov/news/fact_sheets/news_story.cfm?newsIdô¥14895#::textô¥Runway%20
Safety%20Areas%20(RSA)&textô¥The%20RSA%20is%20typically%20500,the%20si-
de%20of%20the%20runway [retrieved 11 Oct. 2020]..
Chapter 18 Rejected Takeoff
Performance
Testing and AFM
Takeoff Field
Length

Chapter Objective
This chapter completes the discussion on takeoff and rejected takeoff. We
cover the basic testing requirements, risks, and mitigations. We then
discuss maximum kinetic energy testing. Finally, we put it all together in
the discussion of the takeoff field length section of the airplane flight
manual (AFM).

Braking Performance and Testing RTO

B
raking capability is an essential part of the performance of an
airplane and its testing starting before the very first flight. Indeed,
the very first brake release is immediately followed (within a few
feet) by a full stop to verify the brakes. Eventually, the airplane goes on to
being taxied at low speed as other systems are being verified with engines
running. The maximum speed reached here depends on the manufacturer
but is typically near or just above 50 kt. (One of the systems to be verified
is the airspeed indicator.) This moves on to high-speed testing down a
runway at speeds up to near the expected rotation speed for the first flight
where the braking system is also progressively tested to the higher energy
(see Fig. 18.1). Finally, the first flight is completed.
The braking system, especially the antiskid, is then exercised and tuned to
provide good behavior at various weight and center of gravity (CG) con-
ditions, under light to heavy braking, from low speed to higher ones to
ensure it works as expected prior to performing rejected takeoff (RTO)
braking distance testing for the AFM.

849
850 Operational Aircraft Performance and Flight Test Practices

Antiskid tuning, rejected Test

takeoff procedure validation program
timeline

Low- High- First VMCG A/C config RTO distance AFM

speed speed flight testing frozen, testing
taxi taxi procedures and max KE
set, speeds testing
validated

Fig. 18.1 RTO development timeline.

Testing for Minimum Control Speed, Ground VMCG

One of the defining speeds for rejected takeoff is VMCG , which defines the
minimum V1 (V1 min or V1MCG ) for rejected takeoff distance purposes. As
required by §25.107, VEF cannot be less than VMCG , and V1 is the speed
gained from VEF to the point where the pilot reacts to the engine failure.
As part of the test preparation for VMCG , one needs to consider that the
critical engine will be shut down on purpose at conditions near minimum
control; thus, one needs to expect that the airplane will deviate laterally.
As a reminder (author’s underline):
§25.149(e) VMCG, the minimum control speed on the ground, is the cali-
brated airspeed during the takeoff run at which, when the critical engine is
suddenly made inoperative, it is possible to maintain control of the airplane
using the rudder control alone (without the use of nosewheel steering), as
limited by 150 pounds of force, and the lateral control to the extent of
keeping the wings level to enable the takeoff to be safely continued using
normal piloting skill. In the determination of VMCG, assuming that the
path of the airplane accelerating with all engines operating is along the cen-
terline of the runway, its path from the point at which the critical engine is
made inoperative to the point at which recovery to a direction parallel to the
centerline is completed may not deviate more than 30 feet laterally from the
centerline at any point.

The expected deviation vs engine cut speed can diverge very rapidly as VMCG
conditions are approached (see Fig. 18.7a later in this chapter). A runway of
sufficient width should be used for this test, and the crew should be ready to
abort testing if the deviation exceeds 30 ft (test failed); see Fig. 18.2. For the
example shown, the pilot aborts the test while the airplane’s deviation is only
18 ft (as seen by the thrust reduction on the second engine and performing an
RTO) when they clearly saw the heading was not coming back towards
center; a continued takeoff under these conditions is not recommended.
The airplane still deviated from center line by almost 50 ft. Testing on a
runway of 150-ft width provides minimum acceptable margins for the
purpose of VMCG test demonstration. The same figure also shows that an
engine failure 1 kt faster than VMCG can result in significantly less lateral
deviation than a failure at VMCG .
 KIAS (kt) Thrust L (% max) Thrust R (% max)
APR
140
99.17873 99.60644 108.278 96.75
120
Thrust R (% max)
Thrust L (% max)
100
80
KIAS (kt)

60
2.067589
40 87.67711
20
0
–20
Rudder (% max) Lateral deviation (ft)

CHAPTER 18
40 10

Lateral deviation (ft)

0
20 –18.00308 0
0 –10
(% max)

–41.59291
Rudder

–20 –20
–40 –30

Rejected Takeoff Performance Testing

–60 –92.88473 –40
–80 –50
–100 –60
True heading (deg) Ground track (deg)
–130
True heading (deg)
Ground track (deg)

–134.776
–132 –135.5294
–139.6542
–134
–136 –139.0802
–138
–140 e
lur rt
–142
fai 6 bo
10 15 20 e 9 25 est a 5015 30 35 40
gin 73 T .9
En 23.0 26

Time (s)

Fig. 18.2 Example of failed VMCG test condition with abort.

851
852 Operational Aircraft Performance and Flight Test Practices

Test requirements, as identified by §25.149(e), include:

• The airplane in each takeoff configuration or, at the option of the applicant,
in the most critical takeoff configuration
• The maximum available takeoff power or thrust on the operating engines
• The most unfavorable center of gravity
• The airplane trimmed for takeoff
• The most unfavorable weight in the range of takeoff weights
For the first item, the test team can test all takeoff configurations, trying
to see if there is a small difference in VMCG from one configuration to the
next, or simply test with the most adverse configuration; typically, this
means the configuration with the largest flap deflection where the resulting
weight on wheel is the lowest (larger flap deflection generating more lift at
the VMCG condition).
Weight on wheel does have an impact on the airplane’s track because
one should expect significant lateral skidding during the test. During the
maneuver, the rudder is deflected to counter the yawing moment from the
asymmetric thrust. This rudder deflection produces a significant side force
that is reacted by the main landing gear (MLG) wheels (see Fig. 18.3),
which limits the lateral deviation some. This skidding looks even more
impressive when viewed from the front. (During testing, manufacturers
often install cameras at the end of the runway to review the testing posttest.)
A search of the Internet will reveal a few examples of VMCG testing (e.g.,
https://www.youtube.com/watch?v ¼ RbPVIqgOpvM&list ¼ PLp1o4SmvP_
cmZzrqL8Gsn0X0_B-SC4tgs&index ¼ 4&t ¼ 0s).
The second item on the list is addressed by ensuring the test is performed
with the maximum expected in-service thrust. This is meant to represent
an engine having a thrust that falls on the high side of the tolerance (typically
2–3% greater than an average engine) while testing under atmospheric con-
ditions (temperature and pressure altitude) that will generate the maximum
thrust. This condition can be achieved by artificially increasing the thrust of
the test engine to support the test. If this artificial increase is not possible, the

Gear side
force Original track/heading
ding
Hea
30 ft
d track
Groun

Rudder side
force

Fig. 18.3 Skidding event during VMCG testing.

 CHAPTER 18 Rejected Takeoff Performance Testing 853

guidance material of FAA AC 25-7D allows a test to be conducted so that the

demonstrated thrust is not less than 95% of the expected maximum.
As well, VMCG should be accomplished with an engine failure that results
in the most severe asymmetric thrust and the most sudden loss (fastest
decay). On most airplanes, this is meant to be a fuel cut (instantaneous
fuel starvation). The guidance of FAA AC 25-7D stipulates that initial test
conditions can be done with an engine thrust reduction to idle (much
slower thrust reduction vs time) to help the crew get familiar with the air-
plane behavior, but the final VMCG should be validated with the most
adverse condition (fuel cut).
For the third item, the most adverse CG has a double impact. A forward
CG will result in lesser weight on wheels and may result in slightly more
lateral deviation, whereas an aft CG will result in a less effective rudder
that is used to stop the heading deviation. Most manufacturers test at aft
CG because they consider it the most adverse condition. The guidance
material of FAA AC 25-7D states:
VMCG testing should be conducted at aft CG and with the nose wheel free to
caster, to minimize the stabilizing effect of the nose gear. If the nose wheel
does not caster freely, the test may be conducted with enough nose up ele-
vator applied to lift the nose wheel off the runway.

For the fourth item, the airplane is trimmed for takeoff. This does not impact
the maneuver most of the time, but for airplanes that are limited by rudder
pedal forces (reversible flight controls), one should not use the trim to help
reduce the force. An airplane is typically trimmed at zero rudder trim deflec-
tion for takeoff.
For the final item, the weight selected is typically the highest weight
where VMCG impacts the takeoff speeds, V1 in particular.
VMCG testing can be conducted with a continued takeoff following the
engine failure or a rejected takeoff once the lateral deviation has been
clearly arrested (no differential braking to help stop deviation). If the
VMCG test is performed with continued takeoff, one should establish what
the expected takeoff speeds with the final VMCG would be so that the
follow-on maneuver is representative (proper VR and V2 ). This author had
a few debates on this approach because some crew/engineers stipulate that
the airplane should have completely stopped the deviation prior to rotation;
however, this is not the intent. (We will expand on this later in the chapter.)
The FAA position on this was spelled out in the final rule when VMCG
was introduced in Amendment 42 of Part 25 (underline added by author):
One commentator recommended that the second sentence of proposed Sec.
25.149(e) be revised to read "During this demonstration, the permissible
lateral deviation of the path of the airplane would be limited to 30 feet."
He said that the revision would eliminate the possibility of misinterpretation.
The FAA believes that the language of the proposal is clear; however, it may be
too restrictive in requiring the ground track to be parallel to or converging
854 Operational Aircraft Performance and Flight Test Practices

toward the centerline of the runway when the airplane is rotated for takeoff,
and thereby unnecessarily delay rotation in determining takeoff performance
under Secs. 25.107(e) and 25.111. Sec. 25.149(e) is therefore revised to state
that the airplane’s path, from the point at which the critical engine is made
inoperative to the point at which recovery to a direction parallel to the
runway centerline is completed, may not deviate more than 30 feet laterally
from the centerline. The adopted rule would allow the airplane to be rotated
for takeoff before recovery to a direction parallel to the runway centerline is
completed; however, it should be noted that it requires that VMCG must be
determined to enable the takeoff to be safely continued using normal
piloting skill.

We provide two examples of VMCG testing performed with a continued

takeoff (see Figs. 18.4 and 18.5). In the first case, Fig. 18.4, the engine
failure occurs at time 44.2 s, and one can note the difference in heading
(where the nose is pointing, third strip from the top) compared to the
ground track (where the CG is going) in the initial few seconds, similar
to what is shown in Fig. 18.3. As the ground track recovers to a value
similar to that prior to the engine failure, the lateral deviation stops. The
pilot then continues the takeoff, but one notes the “relaxed” rudder
(lower deflection) in the continued takeoff, meaning the airplane is fully
controllable. As the airplane takes off, some roll will occur, and there will
be an increase in sideslip angle. The airplane started deviating again
during the initial rotation and climb, but we also note that the pilot is
not using maximum rudder.
In the second example, Fig. 18.5, the deviation was also arrested prior to
liftoff, and no data are presented beyond that point. This test also meets the
intent of the VMCG requirements: the airplane’s path, from the point at which
the critical engine is made inoperative to the point at which recovery to a
direction parallel to the runway centerline is completed, may not deviate
more than 30 ft laterally from the centerline. For a continued takeoff VMCG
test, one should not expect that the pilot can maintain the lateral deviation
similar to that on the ground if the pilot concentrates on maintaining the
runway heading and lifting off the ground (loss of visual cues with the
ground). The only reasonable thing is an expectation of heading control.
For Fig. 18.4, the lateral deviation was actually arrested at about 16 ft devi-
ation; the rudder was then relaxed as the pilot continued the takeoff, allowing
the displacement to drift towards 30 ft. The VMCG speed could then
be reduced.
For a VMCG test followed by an RTO, discussion now focuses on pure
lateral deviation with control via rudder only. The maneuver is executed,
and once the pilot has clearly arrested the lateral deviation of the airplane,
the remaining engine(s) can be brought to idle and differential braking
used [and possibly reconnection of the nose wheel steering (NWS)] to
regain full control of the airplane. Of course, this does require a long(er)
runway, and the tester must also consider other airplane limits like brake
energy and maximum tire speeds. (The airplane will reach much higher
 KIAS (kt) RADALT (ft)
150 150

RADALT (ft)
80.90009 96.27756
KIAS (kt)

100 100
107.7449 114.9515
50 1.144482 1.249215 5.25 10.43801 50
0 0
displacement (ft) round track (deg) Prop speed R (RPM)
True heading (deg) Prop speed L (RPM)

Prop speed L (RPM) Prop speed R (RPM) Roll (deg)

10
1000 1006.551 1019.475 1019.805 1020.146

Roll (deg)
281.5206 2.062604 5
500 –0.686406
–0.8298096 –0.9127628 0 0
0

CHAPTER 18
0
True heading (deg) Track (deg) AoSS (deg)
10

AoSS (deg)
–130 –1.204494 –3.78143 –133.8526–130.1717 –134.4466
5
–134.4271 –134.0073
0
–135 –5

Rejected Takeoff Performance Testing

–134.4562 –134.5499 –5.997502
–10
Lateral displacement (ft) Wheel spd L (kt)

Wheel spd L (kt)

50 73.37209 150
88.99097 92.58811
Lateral

0 100
–16
0 99.29387 –21 –29 50
0
–50 –50
Rudder (% max)
(% max)

0
Rudder

–68.15466 –68.26078 –52.74709

2.583674 e d
–100 ilur ere off on
cov ati
30 40 e fa 07 Re .9996
9 Lift 6499 evi 70
gi n
En 4.214 . 6 ax d 2507
4 49 56 M 0.7
6
Time (s)

Fig. 18.4 VMCG testing with continued takeoff—prop airplane.

855
856 Operational Aircraft Performance and Flight Test Practices

35 120

30 105

25 90
Lateral deviation (ft)

Windmiling @ 33.3 s
CAS_ADC
20 75

Lift-off @ 38.9 s
Engine failure @ 31.7 s

CAS_ADC
Rotation @ 37.0 s
Brake release @ 18.7 s

15 60
Max lateral
deviation
10 26.6 ft 45

5 30

0 15

–5 0
15 20 25 30 35 40
Elapsed time (s)

Fig. 18.5 VMCG with continued takeoff—jet airplane.

speeds on the ground during this type of testing.) As well, the airplane may
start large skids in the recovery and possibly deviate more in the opposite
direction. Note the large heading swings and skids (difference in angles
between the heading and the ground track) in Fig. 18.6.
In the end, VMCG is a regulation margin to define a minimum V1 . The
testing is executed on a dry, smooth runway with NWS off, with zero cross-
winds (if not possible, crosswind in the adverse direction, one that tends to
weathervane the airplane away from centerline), with runways of minimum
crown. When VMCG was introduced in Amendment 42 (formally codified),
one OEM said it would be too conservative to assume zero friction from
the NWS (castering); the FAA did not agree. From the final rule [1]:
Several commentators recommended that the proposal be revised to allow
the use of nose wheel steering in the determination of VMCG under Sec.
25.149(e), if control is through the rudder pedals and the demonstration is
made on a wet runway. The FAA does not agree. The effectiveness of nose
wheel steering depends to a large degree on runway friction characteristics
and the load on the nose wheel. Certification tests on a wet runway would
not cover the more extreme slippery runway conditions or all possible vari-
ations in takeoff conditions and techniques likely to occur in service. The
FAA therefore believes that VMCG should be determined without the use
of nose wheel steering, as stated in proposed Sec. 25.149(e).

As a margin (VMCG defining VEF ), considering the minimum time increment

from VEF to V1 plus a maneuver achieved with NWS off, the expected devi-
ation for a continued takeoff will be significantly less than 30 ft operationally,
even if the critical engine really fails at critical VEF (see Fig. 18.7).
 KIAS (kt) Thrust L (% max) Thrust R (% max)
140
99.41418 99.31392 105.6875 108.7315
120
Thrust R (% max)
Thrust L (% max) 100
KIAS (kt)

80 92.1875
60
40 4.04531
20
0
–20

CHAPTER 18
Rudder (% max) Lateral deviation (ft)
100 40

Lateral deviation (ft)

Rudder (% max)

30
50 0 20
–11.95767
–19.65786 10
0 –21 0

Rejected Takeoff Performance Testing

–10
–50 –20
–30
–100 –40
True heading (deg) Ground track (deg)
–130
True heading (deg)
Ground track (deg)

–131.044
–132 –134.5587
–135.4077 –135.2471
–134
–136
–138
re e
–140 ailu let
ef 9 mp
125 130 135 n gin .184 140 es t co .8458 150 155 160
E 39 T 45
1 Time (s) 1

Fig. 18.6 VMCG test with RTO.

857
858 Operational Aircraft Performance and Flight Test Practices

a) b)
Speed gain after 1 s
Typical reaction time
near VMCG conditions
0.2 – 0.6 s 5.7 kt

30 ft 3.8 kt
VMCG test
results
Lateral
deviation 1.9 kt
trend

VMCG +1 kt +2 kt +3 kt 0.1 g 0.2 g 0.3 g

VEFmin 1s V1MCG Typical OEI Typical AEO

minimum near VMCG near VMCG
Rejected Continued
takeoff takeoff

Fig. 18.7 Lateral deviation from centerline for a continued takeoff vs VMCG .

As discussed previously, part of the side force generated by the rudder is

countered by the cornering force of the rolling tires. Testing on wet and
flooded runways has shown that the cornering force (the resistance to side
force) produced by tires is degraded (see Fig. 18.8). The certification regu-
lations do recognize this impact, and it is even stated in EASA AMC 25.1591:
Minimum V1: For the purpose of take-off distance determination, it has
been accepted that the minimum V1 speed may be established using the
VMCG value established in accordance with CS 25.149(g). As implied in
paragraph 8.1.3, this may not ensure that the lateral deviation after engine
failure will not exceed 30 ft on a contaminated runway

The requirements do not require the VMCG demonstration to be performed

either way (continued takeoff or RTO); that remains an applicant’s decision.
In recent years, the trend has been going toward remaining on the ground for
twin-engine airplanes (test safety) where one engine is killed on purpose
while the remaining engine is still at risk of failing (bird ingestion always
being a concern). For airplanes with three or more engines with an assured

.6
␮s , max

.4

.2

0 4 8 12 0 4 8 12 0 4 8 12
Yaw angle, deg Yaw angle, deg Yaw angle, deg
Ground speed ≈ 100 knots.

Fig. 18.8 Tire cornering force coefficient on dry, damp, and flooded runways [2].
 CHAPTER 18 Rejected Takeoff Performance Testing 859

climb capability following the loss of the test engine plus one more, a continued
takeoff remains an acceptable and safe (though still risky) test.
To illustrate the risk associated with continued takeoff VMCG testing,
consider the accident box describing the testing on Lockheed Martin’s
HTTB airplane (NTSB A-94-101). An engine failure is not the only risk;
the airplane is simply at increased risk of another failure resulting in the
loss of control of the airplane during the test.

NTSB Safety Recommendation A-94-101: Lockheed C-130 High Technol-

ogy Test Bed (HTTB) VMCG test accident, 3 Feb. 1993.
The accident occurred while the test crew was performing a VMCG evalu-
ation that required them to accelerate from a stop and intentionally shut down
the No. 1 engine as the indicated airspeed reached 83 kt. Handling qualities
were to be evaluated as the crew attempted to restore the airplane’s track
on the runway centerline while continuing to accelerate with the remaining
three engines producing power. Although the planned test did not involve
flight, the airplane became airborne and crashed just north of the airport.
The evidence indicates that when the HTTB’s No. 1 engine was shut down,
a large right rudder pedal input was made as the crew attempted to restore the
track of the airplane on the runway centerline. Several seconds later, control of
the rudder was lost, and the airplane veered off to the left of the runway. The
pilot then elected to take off, and the airplane crashed shortly after
becoming airborne.
The Safety Board determined that the probable cause of the Lockheed
HTTB accident was the disengagement of the rudder flight control system
because of inadequate design criteria by the manufacturer of the integrated
actuator package (IAP), which allowed a total loss of rudder control capability,
as well as insufficient system safety review by the airplane manufacturer of the
consequences of the known design feature to all flight regimes.
860 Operational Aircraft Performance and Flight Test Practices

RTO Testing
An RTO performed with the intent of taking credit for the distance
model must be performed using the intended AFM procedure to be certi-
fied. RTO test points do not necessarily need to be performed at the
intended AFM V1 for a given weight; rather, combinations of weight,
kinetic energy, ground speeds, and all engines operating/one engine inop-
erable (AEO/OEI) should be considered by the applicant to build the AFM
RTO model.
Per the takeoff performance testing, the winds should be calm (preferably
lower than 5 kt, in any case no more than 10 kt along runway and 5 kt across
it). FAA AC 25-7D recommends that at least six accelerate–stop flight tests
should be conducted for each takeoff configuration.
If the same braking friction coefficient is claimed for multiple aerody-
namic configurations, a minimum of two RTOs per configuration should
be performed to show that the data from one configuration fall into the
scatter of the data of the baseline configuration (with six RTOs). Some air-
planes are designed to use a single takeoff configuration (one flap setting)
whereas others may offer four or five configurations that provide the best
mix of takeoff distance and climb gradient capability. One should account
for this while planning the scope of testing for RTO.
Testing should be done on a smooth (ungrooved) runway with some tests
on a grooved runway as required/desired.

Instrumentation
The aim of RTO testing is to build a model that will be representative of
the airplane’s deceleration capability when used in service. Therefore, the
minimum instrumentation required to support this testing would be:
• Known deceleration with speed and distance, thus, accelerometers,
differential global positioning system (DGPS), inertial reference unit/
inertial navigation system (IRU/INS), and so on.
• Known airplane configuration (can be via test log).
• Weight and CG for worst case; generally forward CG will be the worst case
because of less weight on the main gear (where the brakes are).
• Known spoiler position (dictates drag increase and lift reduction).
• Able to record pilot’s instrument (position error on airspeed) because this
will be the reference for the engine failure speed.
• Wind, temperature, humidity, and pressure.
• Wind must be corrected to airplane’s wing level (AC 25-7D, AC 23-8C).
• Wheel speed monitoring to verify that the antiskid system works (no
locked wheels).
• Brake hydraulic pressure to ensure proper working order for brakes (time
delay for pressure ramp-up), spoilers, and so forth.
 CHAPTER 18 Rejected Takeoff Performance Testing 861

• Engine parameters:
• Pressure and temperature throughout the engine plus N1 to verify thrust
output as expected. (Remember, this is not a thrust calibration exercise;
this should have been done earlier.)
• Fuel flow.
• Throttle position (to verify time delays for pilot input).

Risks and Mitigations for RTO Testing

Rejected takeoff testing is inherently risky, especially when performed for
takeoff distance credit where the braking will often be performed at higher
reject speeds. Some of the risks to expect are:
• Tire speeds limit exceedance (tire failure).
• Brakes energy limit exceedance (brake fade, fire, fuse plug release).
• Runway length not sufficient (runway overrun, braking performance model
may not be finalized).
• Runway conditions impact expected braking performance (wet runway
testing).
• Loss of directional control (steering failure, brake lock).
To reduce RTO test risks, the following mitigations have proven useful
while the applicant learns about the behavior of the airplane under RTO
conditions:
• Build up weight and speed with progressively higher weights and reject
speeds (buildup in energy).
• Monitor wheel and tire temperature.
• Prepare risk sheets with the best available data in terms of limit energy and
estimated braking distance from a given reject speed.
• Define RTO procedure (i.e., throttle chop, brake application, etc.).
• For nonmaximum kinetic energy (KE) testing, limit the energy absorption
to approximately 80% of maximum KE. (The actual value should be below
the expected fuse plug release; see Chapter 20 for a discussion of this.)
• That number should be such that there is no expectation of fuse plug
release during and after the test. This may involve coordination to ensure
timely brake cooldown soon after the test.
• For successive testing, to ensure energy absorption capability exists, cool
wheels (fans on ground, fly with gear down) in between test points.
The available brake energy of the system is managed via known brake
temperature and energy margins.

Analysis: Braking Coefficient

We discussed the transition phase in Chapter 17, Fig. 17.12, showing the
variation of the various forces with time. The minimum six RTO test points
862 Operational Aircraft Performance and Flight Test Practices

L
T
V
␪
f D
W

Fig. 18.9 Forces acting on an airplane during a ground run (with and without braking).

will allow the performance engineer to come up with an initial model of the
evolution of the forces with time while following the prescribed RTO pro-
cedure. If the scatter is small, further testing may not be required.
For the braking phase with a stable aerodynamic configuration, brakes
applying maximum force and all gears on the ground, we turn to Eq. (17.1)
and Fig. 18.9. During the RTO testing, we will measure the airplane decelera-
tion (dVGS /dt). The thrust comes from a model provided by the engine man-
ufacturer that is adjusted as required. The runway slope variation with
position during the test must be known (runway survey or DGPS). The air-
craft weight prior to brake application must be known; it can be assumed
to be constant during the RTO. The lift and drag coefficient, with and
without ground spoilers, can be derived from wind tunnel testing for the
ground angle of attack (pitch) and corrected for ground effect.
We are thus left with one unknown force, the ground friction force f.
From Eq. (17.1),
 
1 dVGS
f ¼ T  D  W sin (u) þ (18:1)
g dt

We must now determine the braking coefficient model mB for the takeoff
configuration being analyzed. The braking coefficient for a given interval of
the integration is
f
mB ¼ (18:2)
(W  L)
Figure 18.10 shows an example of an RTO analysis using these equations.
In this example, the brake application and throttle movement to idle happen at
about the same time (time 43 s), and the spoilers are selected soon after, being
fully deployed by time 44.17 s. The friction force f is computed throughout
the maneuver (the first part being proportional to the rolling friction force).
The friction force coefficient m is computed throughout the acceleration
and deceleration, provided in the bottom chart of Fig. 18.10 and in Fig. 18.11.
Note that the value is not a constant but includes the expected test scatter
coming from the various variables involved in the testing, such as possible
unsteady winds, vibration on the airplane that may influence some of the par-
ameters used, and uneven braking force naturally occurring with possible
 Ground speed (kt) Wheel speed L (kt)

Wheel speed L (kt)

Ground speed (kt)
100
82.85959 61.375
80 47.20491
60 86.25
40 80.67087
56.23534 43.63373
20 78.14708
0
L (lb) D (lb)
25 2500
20735.69
20 2000
L (lb) (103)

1029.702
15 606.7403 1500

D (lb)
1959.339
10 1000

CHAPTER 18
3425.749
5 1301.484 11779.59 500
5824.131
0 0
Thrust L (lb) f (lb)
Thrust L (lb) (103)

20 14856.54 26111.33 40
12083.93 18547.01 30
15

f (lb) (103)

Rejected Takeoff Performance Testing

6001.332 20
10
17616.31 10
5 1080.103 781.25
0
0 –10
Nx (g) Mu_B
0.4 0.2242911 0.192049 0.4
0.2 0.2
Nx (g)

Mu_B
0.0 0.1994716 0.2772638 0.0
–0.2 0.07571285 0.04665675 –0.2
–0.4 –0.2497799 –0.1755375 –0.4
30 40 dle t 50 g1 ing
2 60 70
le i in
h rott ers ou - b rak 59 -brak 21
T il Mid 2.754 Mid 6.807
spo 239
43 4.17 5 5
4
Time (s)

Fig. 18.10 Moderate braking test analysis.

863
864 Operational Aircraft Performance and Flight Test Practices

0.4

ent
Braking segm
0.3
Mu_B

0.2
Brake
application

0.1
Brake
n
release Rolling frictio

0.0
0 10 20 30 40 50 60 70 80 90 100
Ground speed (kt)

Fig. 18.11 Model friction force coefficient.

skidding of the wheels during the braking segment. The example here is for a
moderate braking test, so pilot brake modulation may also have
been involved.
Out of this testing, one can extract a single value that provides an equival-
ent average braking coefficient (by matching test distance with model dis-
tance). The testing is then repeated at several conditions (minimum of six
RTOs) of varying brake energy to see how the model behaves. For some
brakes, the braking friction coefficient may decrease with cumulative kinetic
energy absorbed. The chart in Fig. 18.12 shows a typical brake coefficient
model that could be expected. (The data points are average brake friction coef-
ficient for a single run.) More than six RTOs may be required to come up with
a good model (represented by the solid line on the chart of Fig. 18.12).
During each test point, one can quickly get a first estimate of the braking
force coefficient by making the following assumptions: all of the braking force
is provided by the brakes only and the runway is flat (no drag, no lift, no
runway slope weight component, all weight on wheels); see Fig. 18.13. Also
note how the braking coefficient shown on the last trace of Fig. 18.10 with
the longitudinal acceleration Nx is recorded.
fB ¼ (W  L)  mB W (18:3)
1 dVGS
mB  (18:4)
g dt
In the example provide in Fig. 18.13, one notes that the recorded longi-
tudinal acceleration can be noisy. Part of the noise in the longitudinal
 CHAPTER 18 Rejected Takeoff Performance Testing 865

0.55

0.5
Model
␮B 0.45

0.4

0.35
KE

Fig. 18.12 Braking coefficient model example.

0.2

0.1

0
Longitudinal Accel (g)

–0.1
Avg decel ≈ 0.35 g
–0.2

–0.3

–0.4

–0.5

–0.6
120

100
Ground speed (KTGS)

80 Av
gd
ec
e
≈0 l=1
60 .34 0.93
g ft/
s2
40

20

0
10 12 14 16 18 20 22 24 26 28 30

Fig. 18.13 Signal noise during brake performance testing.

866 Operational Aircraft Performance and Flight Test Practices

acceleration is airframe vibration (high frequency), airframe rocking (lower

frequency), variation in deceleration rate due to wheel skids, and so forth.
We also note that the change of ground speed with time can give us the
same value. (It averages out the noise.)

Model Validation
No matter what the braking model ends up looking like, it must be vali-
dated against measured deceleration distance (the parameter of importance).
For each test point done, the braking data (from V1 to full stop) should be
expanded with the braking coefficient model inserted into the equation,
and the resulting braking distance should then be compared to the actual
test distance (see Fig. 18.14).

5000

4500

4000

3500

3000
Model distance (ft)

2500

2000

1500

1000
+/– 200 ft
tolerance
500

0
0 1000 2000 3000 4000 5000
Test distance (ft)

Fig. 18.14 Comparing model distance to test distance for RTO.

 CHAPTER 18 Rejected Takeoff Performance Testing 867

The resulting stopping distance should fall within reasonable differences

(scatter) from the actual test results, similar to the continued takeoff distance
test. Variability between the model and test distance could include test winds
not constant, wheel skids during the deceleration, and so on.

Expanding the Data

The applicant should review the results of the transition phase and ensure
the time delays for each pilot action and automatic system are well captured.
Then, for expansion, each action requiring a pilot action should be the time
required to perform the action as demonstrated during testing but not less
than one second (see Fig. 17.6 in Chapter 17). To expand the data in the tran-
sition phase, one should numerically integrate the phase on a time basis.
During this phase, several actions occur that are modeled on a time basis
such as engine spool down, spoiler deployment, and pilot reaction time.
For a given time interval Dt, speed-dependent forces (such as drag and lift)
are computed at the beginning of the time interval, and time-dependent
forces (such as thrust decay and spoilers) are computed at the midpoint of
the time interval (see Fig. 18.15). Then,

DVGSi ¼ ai Dt (18:5)

1
DXi ¼ VGSi Dt þ ai (Dt)2 (18:6)
2
 
T f D
ai ¼ g  sin (u) (18:7)
W i

When the airplane configuration stops changing and the full braking
phase starts, it is recommended to continue the integration on a speed
basis (Fig. 18.16). There should be no more time-dependent forces except
maybe the near idle thrust decay. Then the forces should be computed at
the root mean square airspeed of the speed interval (at VRMS;i ), where
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi r
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1  2 2 2 þ V2

VRMSi ¼ VGSi þ VGSiþ1  0:707 VGS GSiþ1 (18:8)
2 i

Then,

DVGS
Dti ¼ (18:9)
ai

1
DXi ¼ VGSi Dti þ ai (Dti )2 (18:10)
2
The acceleration being similar to Eq. (18.7) but at the RMS speed.
868 Operational Aircraft Performance and Flight Test Practices

Time based
120

110
⌬t
KTAS

100

90

80
120

110
Wheel (GS)

100

90

80
30

20
Spoilers

10

–10
7248 7250 7252 7254 7256 7258 7260

Fig. 18.15 Time interval integration.

Once the RTO distance is determined per the model, one must also add
the certification margin of a distance equivalent to 2 s at V1 .

Maximum Kinetic Energy Testing (Max KE)

The authorities allow the applicant to use the design maximum energy
capability of the wheels and brakes in the preparation of the AFM, provided
that a successful demonstration of the capability is done. The braking
requirement of §25.109 is that the demonstration be done with fully worn
brakes. Brake manufacturers will typically perform the max KE (design
limit) test on fully worn brakes for the compliance of the system. This
must be followed by testing on the airplane to cater for possible uneven
 CHAPTER 18 Rejected Takeoff Performance Testing 869

Speed based
120

110
⌬t
KTAS

100

90

80
120

110
Wheel (GS)

100

90

80
30

20
Spoilers

10

–10
7248 7250 7252 7254 7256 7258 7260

Fig. 18.16 Speed interval integration.

braking distribution and thus uneven energy input into the various brakes on
the airplane. For flight test safety reasons, the authorities allow brakes to be
not fully worn for the execution of the test as long as one does not perform
the test with more than 10% of the usable brake material remaining. As
brakes wear down, their mass reduces. The temperature rise of a brake
system is proportional to its mass (see Fig. 18.17), and depending on the
brake material, they can only reach a given maximum temperature before
they risk failing.
The maximum value the applicant will be allowed to use will be the
lowest of the dynamometer test (for a single brake) or of the flight test
demonstrated energy (total test energy divided by the number of brakes to
compare to the dynamometer test).
870 Operational Aircraft Performance and Flight Test Practices

rn
es wo
ak ly
Extra capacity br ial
w rt ing
Brake Ne Pa ain Brake
energy rem energy
%
10
Max KE
BE
⌬T ≈
rn mC
ly wo

temperature
Ful

Design max
Brake mass Brake material
specific heat

Cold Brake
brake temperature

Fig. 18.17 Relationship between temperature rise and brake wear.

The max KE flight test demonstration should be conducted at high

weight (representative of operational conditions). FAA AC 25-7D specifies
that max KE testing should be performed with not less than maximum
takeoff weight. The max KE demonstration must be preceded by a 3-mile
taxi, including three full stops using normal braking and all engines operating
(brake warmup and energy buildup, representative of an in-service scenario).
FAA AC 25-7D also specifies that the following be done in order to satisfac-
torily demonstrate maximum energy RTO:
• Max KE testing should be conducted with not more than 10% of the
allowable brake wear range remaining on each of the airplane wheel brakes
[Part 25.109(i), Amendment 92]. The 10% allowance on the brake wear
state is intended to ease test logistics and increase test safety, not to allow
the accelerate –stop distance to be determined with less than fully worn
brakes. Following the maximum brake energy stop, it will not be necessary
to demonstrate the airplane’s ability to taxi.
• The maximum airplane brake energy allowed for dispatch should not
exceed the value for which a satisfactory after-stop condition exists, or the
value documented under the applicable technical standard order (TSO) (or
an acceptable equivalent), whichever value is less.
• Landings are not an acceptable means for demonstrating the maximum
rejected takeoff brake energy, one reason being the brakes would not have
the preheating of the long taxi event before the RTO.
Performing dedicated RTO testing always involves a degree of risk; the
risk level increases proportionally to the kinetic energy (speed/weight com-
bination) that must be absorbed. AC 25-7D specifies that a satisfactory after-
stop condition is defined as one in which fires are confined to tires, wheels,
and brakes, such that progressive engulfment of the rest of the airplane
would not occur during the time of passenger and crew evacuation. The
 CHAPTER 18 Rejected Takeoff Performance Testing 871

0.55

0.5
Model

␮B 0.45

0.4
Max KE Demonstration
0.35
KE

Fig. 18.18 Max KE test condition anchors the brake model.

application of firefighting means or artificial coolants should not be required

for a period of 5 min following the stop.
In practice, this test is meant to be done one time only (if it passes), and
the resulting demonstrated braking coefficient for this test should be used as
an anchor point for the brake model (see Fig. 18.18). Depending on the
assumptions used to size the brakes, the applicant will be left with more or
less material in the brakes at the time they are declared fully worn. One
should expect that a fire may start, but as described by the guidance material,
shall remained contained to the wheels. Test risks are still present, as demon-
strated by an accident for the Airbus team when they performed the first max
KE test on the Airbus A340-600 (see accident box).

On 12 Feb. 2002, a test airplane of the Airbus A340-600 fleet suffered severe
poststop brakes/wheels failure following an attempt to demonstrate max KE
value with worn brakes. Several of the wheels suffered structural failure when
the tire pressure rose too high due to the heat from the brakes. This also led to
major fires, and part of the airplane was damaged. The fire was shown to be
beyond acceptable levels before the end of the 5-min time limit. Airbus per-
formed some modification of the wheels and successfully passed the next
max KE demonstration.

https://www.youtube.com/ https://www.youtube.com/
watch?v=kw-QFvGjWfo watch?v=irTizOVM-3U
872 Operational Aircraft Performance and Flight Test Practices

When successfully demonstrated, the operator has little chance of being

limited in service. This limit is typically achieved at maximum ground speed
conditions with downsloping runways, tailwinds, and so forth.
High-energy stops could lead to tire and wheel bursting under the high
pressure resulting from the increase in temperature of the wheel assembly.
To alleviate this condition, wheels are fitted with plugs (called fuse plugs)
that will melt when the assembly reaches a given temperature and release
the pressure in the tire.
The applicant must demonstrate that the plugs will not release under
normal conditions. To do this, the applicant must show that the plugs can
absorb the maximum landing energy without melting, but that they will
release after an RTO of energy superior to that of maximum landing
energy (see Chapter 20).
Let’s introduce a new speed for our RTO model. V1MBE is the airspeed,
for a combination of runway slopes, winds, pressure altitude, air temperature,
and airplane weight, that will result in the maximum certified kinetic energy
being absorbed in the event of a rejected takeoff at V1 .
The kinetic energy absorbed for each available brake can also be com-
puted by numerical integration as
( f DX)
BKE ¼ (18:11)
Nbrakes
Computing the energy absorbed per brake from the total kinetic energy
being absorbed (f DX ) allows the applicant to account for cases of failed
brakes in the AFM model and distribute the kinetic energy of a given man-
euver into the remaining available brakes. Note that not all brakes may
absorb the energy equally; the applicant’s model will need to account for this.

Maximum Wheel Speed

Tires are usually rated in terms of maximum ground speed (true airspeed
combined with winds). The maximum nose wheel can be more limiting, and
the maximum VR should not exceed this limit. In practice, because of
expected deviation in service on the VR , one could consider the limit to be
the maximum ground speed at VLOF under AEO conditions.
The MLG tire speed should be such that it is nonlimiting under the
maximum ground speed at VLOF under AEO conditions. One should con-
sider that a slow rotation may occur, reported winds may be higher than
reported, or the pilot may just rotate late when selecting tires to support a
proposed takeoff envelope.

Airplane Flight Manual: Takeoff Field Length (TOFL)

Flight manuals for Part 25 and Part 23 certified airplanes are designed to
provide only the longest takeoff distance that meets the requirements for a
 CHAPTER 18 Rejected Takeoff Performance Testing 873

Selected by
Applicant
25.107(a)(1)

VEFmin VEF Critical engine

failure speed
“Decision Speed”
ⱖVEF + Vpilot
25.107(a)(2)
V1min V1 ⱕV1MBE Highest speed for start
of breaking - RTO
25.107(e)(1)(i)
ⱖV1 Rotation speed
VRmin VR
ⱖ1.05 VMC 25.107(e)(1)(ii)

ⱖ1.10 VMU, AEO ⱕVMtire

25.107(e)(1)(iv) VLOFmin VLOF
ⱖ1.05 VMU, OEI Liftoff speed
25.107(b)(1) Takeoff
ⱖ1.13 VSR V2min V2 Safety
25.107(b)(3)
ⱖ1.10 VMCA Speed
25.107(c)
25.143(h) Maneuvering capability
25.107(c)(3) ⱕ35 ft AGL 25.107(c)(2) 25.121(b)
Climb gradient

Fig. 18.19 Part 25 example for takeoff speed limitations.

given set of conditions. The takeoff field length (TOFL) is the longest dis-
tance of the AEO takeoff (§25.113), OEI accel –go (§25.111), and OEI
accel–stop (§25.109).
The performance engineer preparing the AFM should focus on:
• Set up airplane correctly for takeoff (configuration, thrust setting, propeller
setting) for atmospheric conditions (pressure altitude, temperature,
winds).
• Select speeds (see Fig. 18.19) as a reminder of the many certification limits
for Part 25.
• Compute distance.
• Verify limits (weights, climb, brake energy, obstacle clearance).
The applicant must provide the TOFL in the AFM for a given combi-
nation of takeoff configuration weight, altitude, and temperature (underline
added by author):
§25.113 Takeoff distance and takeoff run
(a) Takeoff distance on a dry runway is the greater of—
(1) The horizontal distance along the takeoff path from the start of the
takeoff to the point at which the airplane is 35 feet above the takeoff
surface, determined under §25.111 for a dry runway; or [This the OEI
takeoff distance.]
(2) 115 percent of the horizontal distance along the takeoff path, with all
engines operating, from the start of the takeoff to the point at which
the airplane is 35 feet above the takeoff surface, as determined by a
procedure consistent with §25.111.
874 Operational Aircraft Performance and Flight Test Practices

(b) Takeoff distance on a wet runway is the greater of—

(1) The takeoff distance on a dry runway determined in accordance with
paragraph (a) of this section; or
(2) The horizontal distance along the takeoff path from the start
of the takeoff to the point at which the airplane is 15 feet
above the takeoff surface, achieved in a manner consistent with the
achievement of V2 before reaching 35 feet above the takeoff
surface, determined under §25.111 for a wet runway.

Winds for AFM

For safety purposes, and to account for the nonregular reporting of the
wind speed and direction, certification regulations add margins in computing
the distance for takeoff field length in the AFM. Specifically:
§25.105 Takeoff
(d) The takeoff data must include, within the established operational limits
of the airplane, the following operational correction factors:
(1) Not more than 50 percent of nominal wind components along the
takeoff path opposite to the direction of takeoff, and not less than
150 percent of nominal wind components along the takeoff path in
the direction of takeoff.

This margin is included in the distance information and is “invisible” to

the crew.
The AFM performance section should contain a chart that will allow the
operator to compute the component of wind along the runway that will then
be used to determine the takeoff distance (see Fig. 18.20). This is also
required to meet the requirements of §25.1587.
§25.1587 Performance information
(b) Each Airplane Flight Manual must contain the performance information
computed under the applicable provisions of this part (including
§§25.115, 25.123, and 25.125 for the weights, altitudes, temperatures,
wind components, and runway gradients, as applicable) within the
operational limits of the airplane, and must contain the following:
(1) In each case, the conditions of power, configuration, and speeds, and
the procedures for handling the airplane and any system having a
significant effect on the performance information.

This chart is then also used by the crew to verify the airplane with respect to
various wind limits (crosswinds, tailwinds, thrust application procedure vs
winds, etc.).

Minimum Control Speed

The minimum control speeds (VMC , VMCA , and VMCG ) impact the
takeoff speeds and general safety of the airplane, so they must be provided
in the AFM. At the discretion of the manufacturer, the value for these
speeds can be a single value (one for VMCA and one for VMCG ) that
 CHAPTER 18 Rejected Takeoff Performance Testing 875

Wind direction from runway

50
10° 20° 30° 40° 45°

Wind component parallel to runway – knots

W loci ts
ve kno 0
50

ind ty
Win 1
d ve

6
kno locity 50° 40
Tailwind or headwind – knots

55
ts

10
40 45

20

50
30
40

45
30

40
40
35
30 60°

35

50
3 30

30
2 20 60

25
25

20
20 70
20 70° 10

15
80

10
15

5
10 10 80° 0
90
Angle between
5 14 wind direction
0 and runway –

135
120 degrees
0 –10
0 10 20 30 40 50 0 10 20 30 40 50
Crosswind – knots Wind component perpendicular to
runway – knots

Fig. 18.20 Wind component charts for AFM.

represents the most conservative (highest) value expected in service, or the

manufacturer may elect to program the minimum control speed for takeoff
as a function of density altitude (expected thrust reduction with increasing
density altitude); see Fig. 18.21.

VMCG (kt)
100 110 120 130
100 Pressure altitude (FT)
OA

90
T(

15 96
⬚C
)

80 20
Altitude pression (× 100 ft)

25 94
70 30 92
60 35 90
Sea 2000
l lev

40 88
50
VMCG (KCAS)

el
400 000

45 86
40
0
6

84
30 50
800

82
0

20 55
10,0

80
00

10
12

78
,00

14
0

,0
0 76 00

–10 74
–60 –50 –40 –30 –20 –10 0 10 20 30 40 50 60
90 95 100 105 110 Ambient temperature (°C)
VMCA 3 engines (kt)

125 130 135 140 145 150 155

VMCA 2 engines (kt)

Fig. 18.21 Minimum control speeds as a function of density altitude.

876 Operational Aircraft Performance and Flight Test Practices

If the VMC is included in the published V1 /VR /V2 speeds presented to

the crew in the AFM, then the manufacturer did the work and verified the
results, so there is little risk of crew reading errors and the reduced VMC
may provide performance benefits under hot and high conditions. On the
other hand, if the operator is free to select the V1 /VR ratio (unbalance the
takeoff) as well as VMC (not a single number), then a clear procedure
should be provided to help minimize reading errors. We should note that
VMC is critical only in the case of an engine failure near the critical speed;
there are very few accidents related to VMC per se on takeoff under Part
25 and Part 23 (the DC-8 accident discussed in Chapter 9 being one of
them), which indicates that this speed margin has proven to be safe
operationally.

Thrust/Power Setting
The airplane’s acceleration is dictated by the thrust available and the air-
plane’s weight/configuration. The AFM must contain the proper engine
thrust setting to be used for the takeoff. Even if the thrust is set by moving
the throttles to a detent with a computer taking over the setting of the
engine thrust/power, the pilot should still be capable of confirming that
the engine output will actually provide him or her the correct thrust that
will lead to the published takeoff distance the pilot has determined using
the AFM takeoff distance charts.
The engine setting for turbojets/turbofans can be provided in terms of
fan speed (% N1 ) or engine pressure ratio (EPR). For propeller-driven air-
planes, a combination of engine power/torque and propeller RPM/pitch
setting may be required. Any power extraction from the engine (bleeds, elec-
trical power, hydraulics pumps setting, etc.) needs to be clearly identified on
the chart (see Fig. 18.22).
Before we move on to discuss takeoff speeds, we review various takeoff/
RTO distance scenarios. Note that the approach used by the crew can be
easy (the manufacturer offers only one option, thus will fix speeds) or a
little more complicated (allows for performance optimization given the
takeoff/RTO scenarios, but requires more attention to detail when selecting
speeds).

RTO Expansion
When performing the RTO expansion, the applicant’s model must
account for any residual acceleration that occurs after V1 , while the
airplane and its systems become stabilized in the braking configuration,
in the expansion of accelerate–stop performance data for presentation in
the AFM. The applicant must also determine which of the OEI accel –
stop and AEO accel –stop is the longest and use this distance for AFM
expansion.
 CHAPTER 18 Rejected Takeoff Performance Testing 877

(a)
Takeoff thrust setting
to (all engines)
bleeds off
Thrust setting –%N1
SAT Altitude-feet
°C –1000 0 1000 2000 3000 4000 5000 6000 7000
50 83.9 83.7
48 84.5 84.3 84.1
46 85.1 84.9 84.7 84.4
44 85.6 85.4 85.2 85.0 84.8
42 86.2 86.0 85.8 85.6 85.3 85.1
40 86.8 86.5 86.3 86.2 85.9 85.7 85.4
38 87.4 87.1 86.9 86.7 86.5 86.2 86.0 85.6
36 87.9 87.6 87.4 87.3 87.0 86.8 86.6 86.2 85.8
34 87.9 88.1 87.8 87.7 87.5 87.2 87.0 86.8 86.4
32 87.6 88.4 88.3 88.1 87.9 87.6 87.4 87.3 87.0
30 87.3 88.6 88.7 88.5 88.3 88.1 87.8 87.7 87.6
28 87.0 88.3 89.1 88.9 88.7 88.6 88.3 88.1 88.0
26 86.7 88.0 89.2 89.4 89.2 89.1 88.9 88.7 88.5
24 86.4 87.7 88.9 89.7 89.7 89.5 89.4 89.2 89.0
22 86.1 87.3 88.6 89.9 90.0 90.0 89.9 89.7 89.6
20 85.8 87.0 88.3 89.6 90.3 90.3 90.3 90.2 90.0
18 85.5 86.7 88.0 89.3 90.5 90.6 90.6 90.6 90.4
16 85.2 86.4 87.7 88.9 90.2 90.9 90.9 90.9 90.8
14 84.9 86.1 87.4 88.6 89.9 91.1 91.2 91.1 91.1
12 84.6 85.8 87.1 88.3 89.5 90.8 91.4 91.4 91.4
10 84.2 85.5 86.7 88.0 89.2 90.4 91.6 91.6 91.6
8 83.9 85.2 86.4 87.7 88.9 90.1 91.3 91.9 91.9
6 83.6 84.9 86.1 87.3 88.6 89.8 91.0 92.1 92.1
4 83.3 84.6 85.8 87.0 88.2 89.4 90.7 92.0 92.3
2 83.0 84.2 85.5 86.7 87.9 89.1 90.3 91.6 92.6
0 82.7 83.9 85.2 86.4 87.6 88.8 90.0 91.3 92.6
–2 82.4 83.6 84.8 86.1 87.3 88.4 89.7 91.0 92.3
–4 82.1 83.3 84.5 85.8 86.9 88.1 89.3 90.6 91.9
–6 81.8 83.0 84.2 85.4 86.6 87.8 89.0 90.3 91.6
–8 81.5 82.7 83.9 85.1 86.3 87.5 88.7 89.9 91.3
–10 81.2 82.4 83.6 84.8 86.0 87.1 88.3 89.6 90.9
–12 80.9 82.1 83.3 84.5 85.7 86.8 88.0 89.3 90.6
–14 80.6 81.8 82.9 84.1 85.3 86.5 87.7 88.9 90.2
–16 80.3 81.4 82.6 83.8 85.0 86.2 87.3 88.6 89.9
–18 80.0 81.1 82.3 83.5 84.7 85.8 87.0 88.3 89.6
–20 79.6 80.8 82.0 83.2 84.3 85.5 86.6 87.9 89.2
–22 79.3 80.5 81.7 82.8 84.0 85.2 86.3 87.6 88.9
–24 78.9 80.1 81.3 82.5 83.7 84.8 86.0 87.2 88.5
–26 78.6 79.8 81.0 82.2 83.3 84.5 85.6 86.9 88.2
–28 78.2 79.4 80.6 81.8 83.0 84.2 85.3 86.6 87.8
(b)
Take off torque computed for VC = 50. KT
SAT (°C) Propeller speed 100 %
Air Normal High Pressure altitude (ft)
cond air cond air cond
OFF ON ON –1000. 0 1000. 2000. 3000. 4000. 5000. 6000. 7000. 8000. 8500.
–40 – 63. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0
–10 – 27. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 89.7
–8. – 24. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 88.7
–6. –22. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 89.7 87.8
–4. –19. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 88.7 86.8
–2. –17. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 87.7 85.8
0. –14. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 86.7 84.9
2. –12. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 89.3 85.7 83.9
4. –10. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 88.3 84.7 82.9
6. – 7. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 87.2 83.6 81.9
8. – 5. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 89.9 86.2 82.6 80.9
10. – 2. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 88.8 85.2 81.7 79.9
12. 0. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 87.7 84.1 80.7 79.0
14. 3. 90.0 90.0 90.0 90.0 90.0 90.0 90.0 86.5 83.0 79.5 77.9
16. 5. 90.0 90.0 90.0 90.0 90.0 90.0 88.9 85.2 81.7 78.4 76.7
18. 8. 90.0 90.0 90.0 90.0 90.0 90.0 87.5 83.9 80.5 77.1 75.5
20. 10. 90.0 90.0 90.0 90.0 90.0 89.6 86.0 82.5 79.1 75.8 74 2
22. 13. 90.0 90.0 90.0 90.0 90.0 88.1 84.5 81.0 77.7 74.5 72.9
24. 15. 90.0 90.0 90.0 90.0 90.0 86.5 83.0 79.6 76.3 73.2 71 .7
26. 18. 90.0 90.0 90.0 90.0 88.5 85.0 81.5 78.2 75.0 71.9 70.4
28. 20. 90.0 90.0 90.0 90.0 86.9 83.4 80.0 76.7 73.6 70.5 69.1
30. 23. 90.0 90.0 90.0 88.8 85.2 81.8 78.5 75.3 72.2 69.2 67.7
32. 25. 90.0 90.0 90.0 87.1 83.6 80.2 77.0 73.8 70.8 67.9 66.4
34. 28. 90.0 90.0 88.9 85.4 81.9 78.6 75.4 72.4 69.4 66.5 65.1
36. 30. 90.0 90.0 87.1 83.7 80.3 77.0 73.9 70.9 68.0 65.2 63.8
38. 33. 90.0 88.9 85.4 82.0 78.7 75.5 72.4 69.5 66.6 63.8 62.5
40. 36. 90.0 87.1 83.6 80.3 77.1 73.9 70.9 68.0 65.2
42. 38. 88.8 85.3 81.9 78.6 75.4 72.4 69.4 66.6
44. 41. 86.9 83.5 80.1 76.9 73.8 70.8 68.0
46. 43. 85.0 81.6 78.4 75.3 72.2 69.3
48. 46. 83.1 79.8 76.6 73.6 70.6
50. 48. 81.2 78.0 74.9 71.9
52. 51. 79.3 76.2 73.2
54. 53. 77.5 74.4
55 54. 76.5 73.5

Fig. 18.22 Examples of (a) thrust setting and (b) power setting for AFM.
878 Operational Aircraft Performance and Flight Test Practices

FAA AC 25-7D provides the following guidance:

For the all-engines-operating accelerate-stop distance prescribed by
§ 25.109(a)(2), apply the demonstrated time intervals, and associated
delays, of paragraphs 11c(3)(iv) through (vi) after the airplane has accelerated
to V1. The minimum time delays for the AFM shall be the longest of the
demonstrated time between each action (brake application, thrust to idle,
etc) or 1 second. The procedures used to determine the accelerate-stop
distance must be clearly described in the performance section of the AFM
to ensure repeatability in the RTO distance.

14 CFR 25.109(a)(1)(iv) and (a)(2)(iii) (Amendment 92) require the OEI and
AEO accelerate–stop distances, respectively, to include a distance incre-
ment equivalent to 2 s at V1 . The 2-s time period is only provided as a
method to calculate the required distance increment (certification
margin); it is not considered to be a part of the accelerate–stop braking
transition sequence (see Fig. 18.23). The chart in Fig. 18.23 is provided as
part of FAA guidance on the RTO distance. No credit for a shorter distance
can be used for the 2 s at V1 .

AFM model must accurately

represent demonstrated energy state
of the airplane after V1
3rd Action
2nd Action Distance for 2 secs. at V1
V1 (1st Action)

VEF
Speed

Distance
Source: AC25-7A

Fig. 18.23 FAA guidance on RTO expansion.

 CHAPTER 18 Rejected Takeoff Performance Testing 879

Accel-go

BFL

Accel-stop

V1

Fig. 18.24 Example of BFL.

Balanced Field Length (BFL)

The applicant must select the VEF with the minimum set by
§25.107(a)(1). The minimum VEF will yield the minimum V1
[§25.107(a)(2)], but the applicant may elect to have a slightly higher V1
than the minimum. V1 is the maximum speed at which braking must have
been initiated for an RTO to meet the distance published. Obviously, the
lower the V1 , the shorter the accel –stop distance. But lowering V1 also
means that the computed OEI –go distance from this same speed will
increase. One can build a diagram similar to Fig. 18.24 by varying the V1
speed and computing the resulting OEI–go and accel –stop (longest of
OEI or AEO) distances. An engine failure that leads to a V1 near zero air-
speed will have a very short stopping distance, but a continued takeoff
from that speed will require much more runway. As V1 is increased, the
difference between the two reduces.
By inspection, one can see that there will exist a V1 speed that will provide
an equal accel –stop and accel– go distance (see Fig. 18.24). When both dis-
tances are equal, the resulting distance is called a balanced field length (BFL).
The BFL is the shortest distance for an OEI scenario (considering both
accel–go and accel–stop and the AEO accel –stop is included to meet the
requirements).
Figure 18.25 shows the relationship between airspeed selected and the
resulting distance for a BFL case dictated by the OEI –stop condition.
880 Operational Aircraft Performance and Flight Test Practices

(For this figure, we assumed the AEO accel –stop was shorter.) Up to VEF ,
both AEO and OEI cases are essentially the same. From VEF to V1 , there
will be a small difference in the distance between AEO and OEI due to
the loss of thrust. For the go distance, the same VR is used for both AEO
and OEI [§25.107(e)(2)]. Due to the higher acceleration of the AEO, the air-
speed at 35 ft AGL resulting from a single VR speed will be higher; there-
fore, the notation is V2 for OEI and V35 for AEO. The applicant must
provide the delta speed for the AEO case (e.g., V2 þ 10) that aligns with
the takeoff procedure used by the crew and that meets the published
AFM distances.
RTO for brake initiation at V1 is shown for this BFL case, including the
required safety margin (2 s at V1 ). Finally, the AEO takeoff safety margin
of þ15% is also included. The published takeoff field length is the longest
of the distances computed; here, the BFL case is that distance.

Unbalancing the Takeoff

It may not always be advantageous or possible to balance the takeoff. An
airplane with good stopping capability may have, in theory, a V1 for the
accel– stop case greater than the VR required for the OEI accel–go case
for the current runway length (equivalent to a higher weight capability for
RTO than for the OEI takeoff); see Fig. 18.26. The requirements of

AEO to 35 ft
V +15%
XAEO OEI to 35 ft
V35
V2

VLOF
Accelerate-go
VR

V1
VEF
Accelerate-stop
AFM Margin, 2 s at V1

Fig. 18.25 BFL, an airspeed perspective.

 CHAPTER 18 Rejected Takeoff Performance Testing 881

V1MCG V1 MBE

VR V2 min
Shorter RTO Longer RTO
Longer OEI-go V1 Shorter OEI-go

Applicant
selected

Fig. 18.26 Unbalancing the takeoff.

§25.107(e)(i)(1) state that VR  V1 ; therefore, the takeoff weight will be

limited by the OEI takeoff, and the distance will be unbalanced.
For high T/W cases, it may be possible, while trying to balance the
takeoff, that the calculated VEF will be smaller than VMCG . In this con-
dition, VEF must be increased to at least VMCG [§25.107(a)(1)], with the
resulting increase in V1 (V1  V1MCG ). Then the field will again be unba-
lanced because the RTO distance will be longer than the OEI takeoff
distance.
If the computed V1 and VR for the field to be balanced result in a V2 that
is smaller than 110% of VMCA , then the rotation speed must be increased, and
the OEI takeoff distance becomes longer than the RTO distance.
For a very long runway, it may be possible to have a very large takeoff
weight based on OEI takeoff distance, but the maximum V1 may actually
be limited by the brake energy limits (V1  V1MBE ). Therefore, the field
will again be unbalanced due to the limit on V1 (and the associated weight).
The availability of a clearway (see discussion in Chapter 15) may allow for
longer OEI takeoff than RTO distances. A reduction in the runway friction
(wet runway, contaminated runway) will result in a longer braking distance,
which can be countered by reducing V1 to try to balance the takeoff distance.
But again, V1 may not be less than V1MCG . The presence of a stopway (see
Chapter 17) may allow the operator to adjust the V1 speed due to the
longer accelerate–stop distance available (ASDA).

AFM Takeoff Speeds

With the previously discussed impact of the speed selection on distance,
the applicant must decide how much flexibility they will offer to the operators
in selecting takeoff speeds (V1 , VR , V2 ). They can all be prescribed (i.e., for a
given flap, altitude, temperature, and runway condition there is only one set
of V1 , VR , V2 ) or the applicant can offer a sliding V1 . For the computed
takeoff distance to materialize, the correct takeoff speeds must be used.
The applicant must provide, in either chart or table format, the appropriate
V1 , VR , and V2 speeds for the range of temperature, altitudes, and takeoff
882 Operational Aircraft Performance and Flight Test Practices

VMCG limited VMU limited

VMCA limited V2

V2min VR
V1
Airspeed

VSR

Takeoff weight

Fig. 18.27 Possible limits on takeoff speeds.

weights certified (see Fig. 18.27). Should the applicant elect to correct the
speeds for temperature, runway slope, winds, antiskid nonoperational, and
so forth, the corrections must be clearly identified and easy to determine
(for the overall safety of the operator). Figure 18.28 shows some examples
of defined takeoff speed information for an AFM.
If the applicant allows the operator to select its own V1 speed, then some
flexibility in the selection of the speed vs available runway is required. One
format used in the industry is shown in Fig. 18.29.
With the introduction of a sliding V1 , the operator is left to ensure that
the V1 selected does not exceed any limits (less than V1MCG , more than
V1MBE ). The applicant must then provide additional charts to the operator
to find V1MCG and V1MBE to ensure the selected V1 is adequate.
Figure 18.30 is one such example.
Note that the chart in Figure 18.30 provides V1MCG vs airplane weight as
well as a function of altitude and temperature. This author does not rec-
ommend correcting V1MCG with weight.

AFM Takeoff Distance

With the thrust set and speed selected, the applicant then provides
charts or tables that allow the operator to find the right takeoff distance.
We offer examples of such information in Figs. 18.31 (table format) and
18.32 (chart format). Note the information is often listed as uncorrected
distances; corrections then need to be provided. Figure 18.33 is one such
example. Another format with part of the corrections included is shown
in Fig. 18.34.
 CHAPTER 18
Rejected Takeoff Performance Testing
Fig. 18.28 Examples of charts and tables providing reference speeds.

883
 884
Operational Aircraft Performance and Flight Test Practices
(used to obtain one-engine-inoperative weight
allowed from all-engine operating weight chart)

(hidden scale)

(Equiv. all-eng.-). oper. distance

1B

corrected runway length

1 /V
Corrected takeoff distance

corrected runway length

ne
o,

i
Runway length available

f. l
ati

re
its

corrected runway length

dr
max. clearway lim

ee
Detailed view of sliding V1

sp

(one engine loop)

ure
ail
ef

0
2
4
6
8
0

6
V1/VR

0.8
0.8
0.8
0.8
0.8
0.9

0.9

0.9

0.9
gin
en

8
0.9
3.7
(hidden scale) 0

0
3.6

1.0
Clearway (downhill) (uphill) t.w. h.w. 0
Corrected acc-stop distance 3.5
length Slope wind 3.4 0
Anti-skid Inoperative
3. 0
3.2 30
3.1 0
Anti-skid on 3.0 0
2.9 0
2 0
h.w. 2 .80
2. .70
2.5 60
2 0
2 .40
2 .30
2.1 .20 e A
wind 2 0 nc
1 .00 re
t.w. 1.8 .90 R efe
1 0
1 .70
1 .60
(uphill) 1.4 .50
1 0
Slope 1 .30
1.1 .20
(downhill) 0
Accelerate-stop distance available

Fig. 18.29 Takeoff distance chart with sliding V1 .

 CHAPTER 18 Rejected Takeoff Performance Testing 885

V1 (MCG)

Airp
60 ort
pre
ssur
e al
–1 titu
50 de –
0 1 100
0 ft
5 4 2 .
40 6 3
7
8
30 –1
9
0
Airport OAT – degrees C

20
10
0
–10
–20
–30
–40
–50
–60

30 Ref line
weight – 1000 kg.
Takeoff gross

40

50
60
70

84 88 92 96 100 104 108

Minimum V1 for control on the ground V1(MGG) – knots IAS

Fig. 18.30 A V1MCG chart.

Maximum Takeoff Weight

The maximum takeoff weight for a given altitude, temperature, and air-
plane configuration will be the lowest of the following:
• Field length limited (discussed in the “AFM Takeoff Distance” section of
this chapter)
• Weight, altitude, and temperature (WAT) limited (discussed in Chapter 13)
• Obstacle clearance limited (discussed in Chapter 13)
• Maximum certified structural weight
• Maximum brake energy
886 Operational Aircraft Performance and Flight Test Practices

Fig. 18.31 Table formats for AFM takeoff distance.

 Limitation
field length required for
Balanced field length takeoff gross weight of 80,000 lb
15

(×1000 kg)
35 dry runway at 20° flaps

(×1000 Ib)
17 SF2 altitude:oft 60.0
40 0
19
45 2 Airport
21 50.0
Take-off weight 4 pressure altitude
50 23 Am 5 –1000 FT
bi
55 25 en
tt 40.0 6
em
60 27 pe 8
ra 10
29 55 tu 30.0
65 45
50 re
4 (°C

CHAPTER 18
31 3
35 0 )
70 20
15 0
Weight limitations 12
33 40 0
20.0

Ambient temperature –°C

75 see 1–100–05
35
(×1000 m)
0.5 1 1.5 2 2.5 3 3.5 4 4.5 10.0 14
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
(×1000 ft) 0

Rejected Takeoff Performance Testing

REF
anti-ice

off
Engine

on –10.0

+2
slope (%)

+1
Runway

REF –20.0
0
–1
–2
Headwind
–30.0 15
+50
+40
Wind (kt)

+30
+20 –40.0
+10 REF
0
–10 Tailwind –50.0
(×1000 m)
0.5 1 1.5 2 2.5 3 3.5 4 4.5 4 5 6 7 8 9 10 11 12 13 14 15 16
–1000 ft
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Field length (×1000 ft)
14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48
–100 M
Effective runway length

Fig. 18.32 Chart formats for takeoff distance.

887
888 Operational Aircraft Performance and Flight Test Practices

Dry runway takeoff field length correction

chart flaps 10⬚
TOFLOOR_DRY_F10_EO100_COLOUR

45 15 15 45

Reference line
43 14 14 43
41 41
13 13
39 39
37 12 12 37
Uncorrected takeoff field length – 100 meters

35 35
Uncorrected takeoff field length – 1000 feet

Corrected takeoff field length – 100 meters

Corrected takeoff field length – 1000 feet
33 11 11 33
31 10 10 31
29 29
27 9 9 27
25 8 8 25
23 4 23
7 1 7
21 21
2 3
19 19
6 6
17 17
15 5 5 15
13 13
4 4
11 11
9 3 3 9
il ad UP
7 Ta DOWN 7
2 He 2
5 5
1 1
0 10 20 30 ᎐2 ᎐1 0 1 2
Wind–knots Runway gradient–%

Fig. 18.33 Field length correction chart.

The applicant should consider providing in the AFM a clear methodology

for the operator to follow (guide them through the AFM charts) to determine
the correct (and safe) maximum weight that can be used for a given combi-
nation of airport pressure altitude, temperature, winds, and runway slope.
Commercial airplane manufacturers usually provide to airline operators
computer programs [in the form of a Standard Computerized Aircraft Per-
formance (SCAP; https://www.iata.org/en/publications/store/standard-
computerized-airplane-performance--scap-/), electronic AFM, etc.] that
 CHAPTER 18 Rejected Takeoff Performance Testing 889

V1 (KIAS) VR (KIAS)
V2 (KIAS) TOFL (FT)
RWY TAKEOFF WEIGHT, LBS (KG)
TEMP GRAD 9000 10,000 11,000 11,500 12,000 12,500
(°C) (%) (4082) (4536) (4990) (5216) (5443) (5670)
88 89 88 88 87 88 88 88 87 88 87 88
88 89 88 88 87 88 88 91 88 94 87 97
᎐2.0
OAT 103 3098 101 3007 99 2957 101 2967 103 2957 105 2951
10
0.0 88 89 88 88 88 90 88 93 87 96 90 99
101 2970 99 2904 98 2900 101 2893 103 2892 105 3091
ISA
1.0 88 89 88 88 88 91 87 94 89 97 92 100
–5 100 2915 98 2861 98 2867 101 2864 103 2995 105 3220
1.5 88 89 87 88 88 92 87 95 90 97 95 100
N1(%) 100 2890 97 2842 98 2852 101 2852 103 3039 105 3368
101.2 2.0 88 89 87 88 88 92 88 95 92 98 99 100
99 2866 96 2824 98 2838 101 2881 103 3175 105 3693
Wind correction:
Decrease the uncorrected TOFL by 5% for every 10 knots of headwind;
Increase the uncorrected TOFL by 3% for every 1 knot of tailwind.

Fig. 18.34 Takeoff speed and distance with runway slope correction included.

will determine the maximum takeoff weight for these listed parameters by
considering all limiting factors.
The guidance material for the certification of computerized AFM
(CAFM) is provided by FAA AC 120-76C, Guidelines for the Certification,
Airworthiness, and Operational Use of Portable Electronic Flight Bags.
FAA Order 8900.1 Volume 4, Chapter 15, Section 1, Electronic Flight Bag
Operational Authorization Process, defines three classes of electronic flight
bags (EFBs):
• Class 1: These EFBs are portable, commercial off-the-shelf (COTS) devices
that are part of a pilot’s/crewmember’s flight kit. Class 1 EFBs are not
mounted to the aircraft, connected to the aircraft systems for data, or
connected to a dedicated aircraft power supply (FAA Order 8900.1,
4-1643A).
• Class 2: These EFBs are portable, COTS devices that are part of a pilot’s/
crewmember’s flight kit. Class 2 EFBs are typically mounted to a
permanently installed mounting device and may be connected to a data
source (wired or wireless), a hardwired power source, or an installed
antenna (FAA Order 8900.1, 4-1643B).
• Class 3: These hardware devices are installed with design approval (in
accordance with applicable airworthiness regulations) (FAA Order 8900.1,
4-1643C).
Using an EFB does not mean the operator will get error-free data. Several
incidents and accidents have resulted from operators entering the wrong
890 Operational Aircraft Performance and Flight Test Practices

information (often wrong weight) into the EFB and thus getting (correctly
calculated) wrong information for the upcoming takeoff.

TOFL: Military Approach

In order to maximize payload capability from a given airfield, some mili-
tary operational procedures will allow the use of multiple critical speeds
instead of the transport category single V1 speed (see Fig. 18.35).
• Planning for minimum go speed (decision speed):
• Minimum speed that will allow the airplane to become airborne (ground
run, not distance to 35 ft AGL) within the available runway
• Planning for maximum refusal speed (max brake speed):
• Max speed for AEO stop on available runway
• Critical engine failure speed (speed that results in an equal distance
between a rejected takeoff or a continued takeoff to 50 ft AGL
after an engine failure); the resulting distance is the critical field
length (CFL).
There is a clear advantage in using this approach to maximize the
payload-carrying capability of an airplane out of a given airfield compared
to the transport category approach. There is also an increased risk of
getting a speed, and thus distance, wrong. Military airplanes typically
operate with a given crew (always the same people) with dedicated training;
this helps with coordination compared to transport category airplanes, which
often get a new combination of pilot flying and copilot on every flight.

Special Test Airplanes

When dealing with nonsymmetric airplanes like the multitude of engine
flying test beds (FTBs) in use today (see Fig. 18.36), one must carefully review
the assumptions used for computing the TOFL because the critical engine for
the accelerate– stop condition may not be the same as for the accelerate –go
distance. Each configuration (additional engine, one engine changed com-
pared to baseline airplane) will have its own new minimum control speed

V V
Liftoff
Vgo Vstop
VEF
Available

Available
runway

runway

X X

Fig. 18.35 Military operational takeoff distances.

 CHAPTER 18 Rejected Takeoff Performance Testing 891

T2 T5 T3
T1

DW

Fig. 18.36 Engine flying test beds.

and may impact the overall airplane’s aerodynamics; VSR may also be
affected. The question becomes, under asymmetric thrust, which one is the
critical engine for takeoff?

References
[1] Federal Aviation Administration, CFR Final Rule, Federal Register, Vol. 43, No. 10,
16 Jan. 1978, https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgFinalRule.nsf/
a09133bddc7f4fbb8525646000609712/a95af4c6be88807e862568240073cdcd!OpenDo
cument [accessed 12 Oct. 2020].
[2] Yager, T. J., and McCarty, J. L., “Friction Characteristics of Three 30 x 11.5-14.5, Type
VIII, Aircraft Tires with Various Tread Groove Patterns and Rubber Compound,”
NASA Technical Paper 1080, Dec. 1977.
Chapter 19 Landing
Performance

Chapter Objective
The landing is a very small part of any flight, typically lasting less than
1 minute, yet it is an extremely busy time for the flight crew and a very impor-
tant performance parameter. This chapter will provide tools to build up a
landing performance module and will discuss possible variables impacting
such distance. It will also briefly discuss some operational considerations
for landing.

Basic Modeling

L
anding an airplane involves many variables that must be controlled by
the crew (or software for unmanned airplanes) over a short period.
How much control the crew has during landing will have a major
impact on the resulting landing distance. For modeling the landing distance,
and seeing the impact of the many variables, we will break down the landing
into three segments (see Fig. 19.1):
1. Air segment that extends from a given height above ground with the
airplane at a steady approach angle and airspeed to the touchdown point.
This segment can be further broken down into two subsegments:
a) Stable approach angle.
b) A crew input is required to slow down the descent rate prior to
touchdown.
2. Transition segment from touchdown to a stable full braking segment.

Air segment Transition Full braking

VREF

Approach angle
Screen height

Touchdown Full decel Full stop

Fig. 19.1 Landing model segments.

893
894 Operational Aircraft Performance and Flight Test Practices

Pilot FOV

MLG flight path

Fig. 19.2 Pilot FOV vs MLG flight path.

3. Full braking segment where all gears are on the ground and deceleration
devices are extended.
FAA 14 CFR 25.125, Landing, states:
(a) The horizontal distance necessary to land and to come to a complete
stop (or to a speed of approximately 3 knots for water landings) from a
point 50 feet above the landing surface must be determined (for standard
temperatures, at each weight, altitude, and wind within the operational
limits established by the applicant for the airplane).

The nominal approach angle for transport category airplanes (certificated

under FAA 14 CFR Part 25) is 3 deg. Should this angle be maintained
from 50 ft to touchdown, then the resulting air distance Xa would be

50 ft
Xa ¼ ¼ 955 ft (19:1)
sin(3 deg)

So the baseline regulation would put the airplane down approximately

955 ft beyond the start of the runway. Why this far? The easy answer is to
cater for normally expected operational deviation. On most airplanes, the
pilot is not presented with a flight path indication (vector where the air-
plane is going) in their field of view. As the airplane becomes aligned
with the runway and the pilot gauges where the airplane will land, a touch-
down point is chosen from the airplane’s field of view (FOV). Looking
through the windshield in the final moments of the approach, the pilot
will see the entire runway in a slanted way. From this and the perceived
flight path, the pilot needs to gauge where the main landing gear (MLG)
will touch (see Fig. 19.2). That view will be influenced by the geometry
and size of the airplane (the relative position of the pilot vs the location
of the MLG).

NTSB AAR-14/01 On July 6, 2013, about 1128 Pacific daylight time, a Boeing
777-200ER, Korean registration HL7742, operating as Asiana Airlines flight
214, was on approach to runway 28L when it struck a seawall at
San Francisco International Airport (SFO), San Francisco, California. . . .
The flight was vectored for a visual approach to runway 28L and intercepted
 CHAPTER 19 Landing Performance 895

the final approach course about 14 nautical miles (nm) from the threshold at
an altitude slightly above the desired 38 glidepath. . . . the flight crew misman-
aged the airplane’s descent, which resulted in the airplane being well above the
desired 38 glidepath when it reached the 5 nm point. . . . The pilot flying (PF)
selected an autopilot (A/P) mode . . . that . . . resulted in the autoflight system
initiating a climb because the airplane was below the selected altitude. The PF
disconnected the A/P and moved the thrust levers to idle, which caused the
autothrottle (A/T) to change to the HOLD mode, a mode in which the
A/T does not control airspeed. The PF then pitched the airplane down and
increased the descent rate. . . . As the airplane reached 500 ft above airport
elevation, . . . the airplane was slightly above the desired glidepath . . . and
the airspeed [was] decreasing rapidly. The thrust levers were still at idle,
and the descent rate was about 1,200 ft per minute, well above the descent
rate of about 700 fpm needed to maintain the desired glidepath. . . . The
decreasing trend in airspeed continued, and about 200 ft, the flight crew
became aware of the low airspeed and low path conditions but did not initiate
a go-around until the airplane was below 100 ft, at which point the airplane
did not have the performance capability to accomplish a go-around.

In this next accident (Canada TSB A07A0134), the flight crew misman-
aged the eye-to-main-landing-gear position at an airport with a relatively
short runway while operating a new, larger airplane type.

Canada TSB A07A0134: On 11 Nov. 2007, a Bombardier Global 5000

touched down 7 ft short of the runway at Fox Harbour aerodrome in Nova
Scotia. The main landing gear was damaged when it struck the edge of the
runway, and directional control was lost when the right main landing gear col-
lapsed. The aircraft departed the right side of the runway and came to a stop
1000 ft from the initial touchdown point.
The crew had selected a touchdown point 500 ft from the beginning of the
runway to maximum roll out distance and had performed such a landing at
that airport several times in the previous year but on smaller planes.
Ten seconds before impact, and while the aircraft was approximately 0.5 n
miles from the threshold, the pilot flying, preparing for the landing, transi-
tioned completely to the ground and runway environment for visual refer-
ences used to judge height and timing of the landing flare. Four seconds
before impact, 0.1 n miles from the threshold, the auto callout system
announced “50”. However, because of the low terrain before the runway,
the aircraft was less than 50 ft above runway elevation. One second after the
“50” auto callout, the autothrottle speed mode changed from SPEED to
RETARD mode, and the thrust levers automatically moved to the IDLE pos-
ition. From this point on, the airspeed continued to decrease until it reached
its minimum recorded value of 102 kt, 16 kt below the corrected Vref [author
896 Operational Aircraft Performance and Flight Test Practices

comment: meaning, at a higher angle of attack]. The airplane main landing

gear struck the ground prior to the runway.
The Transportation Safety Board concluded the flight crew members flew
the approach profile as they had done in the past on the smaller Bombardier
Challenger 604 (CL604), with no consideration for the Global 5000 greater
aircraft eye-to-wheel height (EWH), resulting in a reduced threshold crossing
height (TCH). The crew also allowed the airplane to reach a low energy state
and a pitch angle up to 10.6 deg without an associated thrust increase, and so
could not correct the flight profile, resulting in the impact with the sloped
surface before the runway threshold.

Pilot eye
height

Challenger 604 attitude

ude
000 attit
Global 5

Runway 52 ft. asl

Ground impact
Road 45 ft. asl

With this in mind, let’s look at the details of a landing.

Air Segment
The air segment, for the purpose of computing landing distance, will start
at 50 ft above ground and end when the airplane touches the ground (part of
the weight being carried by the landing gear). The aircraft should ideally cross
the 50-ft height (screen height) at the reference approach speed VREF , at a
flight path angle of –3 deg (nominal), and exactly at the beginning of
the runway.
To ensure a stable approach at 50 ft, testing requires the airplane to be
stable at higher altitude and approach maintained down through 50 ft, typi-
cally no later than 500 ft above ground level (AGL). For a stable approach, the
sink rate (SR, þ down) at the screen height is
SR ¼ VREF sin(g) (19:2)
With a nominal approach angle of –3 deg, one should expect sink rates of
the order of that shown in Fig. 19.3. In the same figure, a second line at
 CHAPTER 19 Landing Performance 897

1200

1000

800
Sink rate (ft/min)

c h
p roa
600 deg ap
3.5 ch
pp roa
e ga
400 3d

200

0
50 70 90 110 130 150 170
Approach speed, VREF (KTAS)

Fig. 19.3 Sink rate vs true airspeed and approach angle.

–3.5 deg was added to show the sensibility of the sink rate to the deviation
from the nominal angle. A given airplane, with its normal landing configur-
ation, over the landing weight range would have a nominal expected sink rate
that the pilot would come to expect; however, that cue may not always be suf-
ficient, as was seen in the example of the Asiana Airlines accident (see acci-
dent box).
To maintain that steady approach, there must be a thrust deficit where the
drag of the airplane is larger than the thrust available during the descent. As
seen in Chapter 12 of this book, to maintain the steady descent angle, the fol-
lowing equation applies:
 
T D
g ¼ sin1 (19:3)
W

For a nominal 3-deg approach angle, a 100,000-lb airplane would

require approximately 5235 lb more drag than what the engine produces
in thrust. This extra drag is usually a combination of landing gear down
and a draggy landing configuration plus a low idle thrust setting.
Figure 19.4 provides a quick understanding of the thrust deficit for
various airplanes based on weight and a nominal 3-deg approach angle
[sin(3 deg) ¼ 0.0523].
As a rough rule of thumb, the flight idle thrust of modern jet engines is
equal to about 5% of the takeoff thrust. We saw in Chapter 12 that
898 Operational Aircraft Performance and Flight Test Practices

60,000

50,000
Airbus A380 MLW

40,000
Thrust deficit (lb)

30,000 Boeing 777-300ER MLW

Boeing 787-9 MLW

20,000

10,000
CS100

0
0 200,000 400,000 600,000 800,000 1,000,000
Weight (lb)

Fig. 19.4 Required thrust deficit for nominal 3-deg approach.

modern jetliners have about 30 –35% thrust to maximum takeoff weight

ratio, so it can be expected that the flight idle thrust would represent approxi-
mately 1.5% of the maximum takeoff weight and about 2% of the maximum
landing weight of typical jetliners. Rewriting Eq. (19.3) for a 3-deg approach
and a thrust value equal to 2% of the landing weight gives the following
minimum drag value required for the landing configuration:
D T
¼  sin (3 deg) ¼ 0:02 þ 0:0523 ¼ 0:0723
W W
This represents an aerodynamic efficiency (L/D) of no greater than
L 1
¼   ¼ 13:83
D D
W
At that value of L/D on approach, the airplane’s engines would need to be
at flight idle thrust detent for the duration of the approach, offering no capa-
bility for the pilot to modulate thrust to maintain speed on a steady descent
when faced with various weather conditions or if off the approach path (too
high). Therefore, some margin (extra drag relative to available minimum
thrust) is required. Figure 19.5 shows an approach with an airplane that
had a good thrust deficit margin between the approach thrust and idle
Airspeed KCAS Airspeed KCAS Approach speed 132 KCAS Wind gust
135 VREF127 KCAS
130
125 Turbulence
120

Altitude ft
1000
Altitude ft

50 ft AGL
500

0
FPA deg
0
–1
FPA deg

–2 Nominal 3 deg

CHAPTER 19
–3
–4
–5
TLA % full N1 % max
TLA % full

N1 % max
30 80
20 60

Landing Performance
10 40
20
0 10 20 30 40 50 60 70 80 90

Time (s)

Fig. 19.5 Throttle movement on nominal 3-deg approach.

899
900 Operational Aircraft Performance and Flight Test Practices

Start of
VREF flare

3 deg (nominal)
Thrust cut

50 ft Possible ground effects

screen height
hflare

Xa1

Fig. 19.6 Thrust cut height and ground effects.

thrust; the crew could execute small throttle adjustments with good engine
response (as seen by the responsiveness of the engine fan speed, N1).
Normal speed excursion due to turbulence and even a wind gust closer to
the ground were easily controlled here. As well, the pilot was also able to
reduce the speed from approach (132 KCAS) to VREF (127 KCAS) prior to
reaching 50 ft above ground level (AGL) with the use of throttle (thrust
control) alone.
At one point near the ground, the crew will reduce thrust in preparation
to land. Also, the ground proximity will start impacting the airplane’s aero-
dynamics. One will tend to increase the sink rate whereas the other will
tend to reduce it (see Fig. 19.6). We will address both in the following
subsections.

Thrust Cut Height Impact

The pilot is approaching the ground to land while maintaining a
steady flight path. To maintain that steady path, the thrust remains essentially
constant, with only small adjustments made to account for small deviations
from the nominal flight path due to turbulence or fine adjustment of the
flight path. Figure 19.7 provides an example of such a steady approach
with small airspeed variation (turbulence) from a nominal value, pilot
thrust adjustment (throttle input), and engine reaction to that adjustment
(engine fan speed, N1); note the time delay in engine response to pilot input.
While on a steady flight path and steady airspeed, the airplane will be at a
constant pitch attitude u, the pitch angle being equal to the flight path angle
plus the angle of attack a [see Eq. (12.1) in Chapter 12]. As the airplane gets
closer to the ground, typically at or slightly higher than 100 ft above ground,
the pilot will pick a target touchdown point on the runway. From this point
on, the pilot will manage the airplane path to this touchdown point.
 Airspeed (KTAS) RADALT (ft)
130
250
Airspeed (KTAS)

125

RADALT (ft)
120 200
115 150
110 50 ft 100
TD
105 50
100 0
Pitch (deg) FPA (deg)
Pitch (deg) FPA (deg)

10
5
0
−5

CHAPTER 19
−10
Throttle (% max) N1 (% max)
15 70
Throttle (% max)

60

N1 (% max)
10 50
5 40

Landing Performance
30
0 20
−5 10
10 15 20 25 30
Time (s)

Fig. 19.7 Steady approach with engine cut height.

901
902 Operational Aircraft Performance and Flight Test Practices

At approximately the screen height (50 ft AGL), the pilot will reduce thrust in
preparation for touchdown. Regulation for the change in thrust is as follows:
FAA 14 CFR §25.101 (f) Unless otherwise prescribed, in determining the . . .
landing distances, changes in the airplane’s configuration, speed, power, and
thrust, must be made in accordance with procedures established by the appli-
cant for operation in service.
FAA 14 CFR §25.125 (b)(3) Changes in configuration, power or thrust, and
speed, must be made in accordance with the established procedures for
service operation.

As we look at the impact of the thrust reduction in flight on the airplane

landing distance, we will put this change in configuration in perspective to
the typical time scale of the maneuver. Equation (19.1) indicated that the
nominal 3-deg approach would result in a 955-ft landing distance if the
flight path is maintained at exactly 3 deg to touchdown. The air time from
50 ft to touchdown, again if speed and flight path were maintained, would be

50 ft
t¼ (19:4)
SR
An airplane with an approach speed of 100 KTAS on a 3-deg approach
(sink rate SR of 8.83 ft/s) would have a nominal air time of 5.66 s; in contrast,
an approach speed of 140 KTAS on a 3-deg approach (SR ¼ 12.37 ft/s)
would have an air time of 4.04 s. Both are relatively short times for the
pilot to manage thrust, flight path, and touchdown sink rate. We will use
this reference time frame (4 to 6 s) to look at the impact of the thrust cut
height on the overall landing air distance.
The first impact of the thrust cut, other than a possible pitching moment
effect if the thrust line does not go through the center of gravity, is the
reduction in the flight path angle; see Eq. (19.3). The change in thrust is, of
course, not a step input as seen in both Fig. 19.5 and Fig. 19.7; rather, the
pilot reduces the thrust setting (throttle) rapidly, and the engine reacts (as
seen by the fan speed, N1) to the thrust reduction command in a finite
amount of time, typically 2 to 6 s.

Example 19.1
An airplane configured for landing (landing flaps and landing gear down) has
a parasitic drag coefficient of 0.1000, a wing aspect ratio of 9, and an Oswald
coefficient of 0.85 for the wing. The airplane’s weight is 75,000 lb, and it has a
reference wing area of 1200 ft2 . At that weight, it has a reference landing speed
VREF of 140 KTAS. Plot the steady level flight drag of this airplane over a
small speed range on either side of VREF if the airplane is out of ground
effect (OGE). What is the thrust for a –3-deg approach OGE, and what is

(Continued)
 CHAPTER 19 Landing Performance 903

Example 19.1 (Continued)

the airplane sink rate? If the thrust is reduced to either 6000 lb (Option 1)
or 4000 lb (Option 2), what is the impact on flight path angle and sink rate?
Solution: Using the information provided, we can create the chart shown in
Fig. 19.8. The computation also indicates that it would require 10,902 lb of
thrust to maintain level flight. The thrust required for a – 30-deg approach
angle would be
T ¼ D þ W sin (g) ¼ 6977 lb

16,000
AR = 9

14,000
VREF

12,000
Thrust for level
Drage (lb)

flight
10,000

8000
Thrust for –3 deg

Idle thrust
6000 option 1
Idle thrust
option 2
4000
100 110 120 130 140 150 160 170 180
KTAS

Fig. 19.8 Options 1 and 2 idle thrust vs airplane drag and thrust for –3-deg approach.

The sink rate for this approach angle would be

SR ¼ VREF sin (g) ¼ 12:37 ft=s
If the thrust is reduced instantly to the first idle thrust option, the new
flight path angle and sink rate would then be
Option 1: g1 ¼ 3:75 deg SR1 ¼ 15:44 ft=s
This would be a noticeable change to the pilot. If we now look at option 2,
the numbers would be
Option 2: g2 ¼ 5:28 deg SR2 ¼ 21:74 ft=s

(Continued)
904 Operational Aircraft Performance and Flight Test Practices

Example 19.1 (Continued)

This very large change would create a sudden feeling of sinking to the crew
that may lead the pilot to delay reducing the thrust or do it in a slower
manner.
If instead of allowing the airplane to sink, the pilot maintained the flight
path as the thrust reduced, this would result in a reduction in airspeed due
to the thrust deficit. That slowdown can be estimated by the following
equation [derived from Eq. (12.2)]:

Tidle  D 1 dV
 sin (gapp ) ¼ (19:5)
W g dt
where gapp is the initial approach angle (nominal –3 deg). For the previous
example, with option 2 idle thrust, if the nominal – 3 deg is maintained at
the time of the instantaneous thrust reduction, the initial deceleration
would be
dV =dt
sin (5:28 deg)  sin (3 deg) ¼
g
dV ft
¼ 0:04 g ¼ 1:28 2 ¼ 0:76 kt=s
dt s

It is most likely that a combination of both an increasing flight path and an

airplane slowdown will occur as the thrust is cut. Both, in isolation or in com-
bination, lead to reduced air distance.
On most modern transport category airplanes, the engines will take any-
where from 2 to 7 s to slow down from approach thrust to flight idle thrust.
This is of the same order of time as the air distance from 50 ft to touchdown;
therefore, a thrust cut will be done by the crew usually below 50 ft above
ground. [Remember that the thrust cut is done using an approved crew pro-
cedure per §25.101(f) and §25.125(b)(3) for which the AFM published air dis-
tance is determined.]
In summary, the main elements that will influence the air distance when it
comes to thrust cut height are:
• Engine spool-down time
• Thrust cut height
• Difference in thrust between approach thrust and flight idle thrust
We did not cover the following in the previous discussion:
• The impact of the thrust line vertical position vs the center of gravity
where a change in thrust would have an impact on the pitching moment.
• The impact of the thrust on the lift of the plane; for example, for a
propeller-driven airplane, the propeller flow that produces thrust also flows
around the wing and influences the generation of lift.
 CHAPTER 19 Landing Performance 905

Ground Effects
To add to the complexity of the thrust cut impact on the flight path and
airspeed, one must consider that as the airplane approaches the ground, the
airflow around it will be altered by the ground proximity. The height at
which the airplane starts being impacted by the ground proximity is
equal to about one wing span. This topic was covered extensively in [1],
and an equation for the effects of the ground proximity was developed;
it will be adopted for this book. The ground effect factor F will take the
following form:
"   #
2 p b 2
F ¼ 1  2 ln 1 þ (19:6)
p 8 h

where h/b is the ground effect height parameter. It represents the height of
the wing over its wing span (see Fig. 19.9).
The ground effect height parameter typically varies from 0.055 (high
aspect ratio, low wing) to 0.20 (low aspect ratio, high wing) when gear is
on the ground. The limit of Eq. (19.5) is about an h/b of 0.033, where the
value goes from a positive one to a negative one. Figure 19.5 compares the
ground effect factor of Eq. (19.5) to another commonly used equation and
to flight test data collected by NASA in Figure 19.10.
The impact of the ground proximity on the airplane’s lift curve slope
translates approximately as follows:
C‘a
CLa  (19:7)
C‘a
1þF
p AR
where:
C‘a ¼ Two-dimensional lift curve slope, 2p
AR ¼ Wing aspect ratio, b 2 /S (wing span squared over wing area)

The combined effect of wing aspect ratio and ground proximity is illus-
trated in Fig. 19.11. As expected, the higher aspect ratio wings tend to

Fig. 19.9 Ground effect impact on airflow and definition of ground effect height parameter.
906 Operational Aircraft Performance and Flight Test Practices

1
1.0

0.9 .8 NASA TM 85920

.6
Relative
0.8 ground .4
effect
Relative ground effect (1 – )

.2
0.7 .0
−.2
0.6 0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0
Wheel height
Span
0.5
2
16 h b
0.4 = 2
1+ 16 h b
0.3 2
2 ln 1  b
 = 1– +
0.2 2 8h

0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Height factor (h/b)

Fig. 19.10 Relative ground effect.

behave more like a two-dimensional wing (Fig. 19.11a). Typical transport cat-
egory airplanes have wing aspect ratios varying from about 7 to 12. What can
be seen in Fig. 19.11b is that the ground proximity has a larger impact on
smaller aspect ratio wings than larger ones based on the same height over
wing span.
In essence, decreasing the height parameter h/b has a similar impact as
increasing the wing aspect ratio. This can be expressed as an in-ground-effect

(a) (b)

Fig. 19.11 Ground effects’ impact on wing lift curve slope.

 CHAPTER 19 Landing Performance 907

(IGE) equivalent aspect ratio

AR
ARIGE ¼ (19:8)
F
Not shown in this quick analysis is the loss of stall angle of attack (astall )
and often of maximum lift coefficient (CLmax ) while IGE; this was covered
during our discussion on VMU in the takeoff Chapters 15 and 16. This
reduction in astall is typically not a dominant factor during a landing.
The change in lift curve slope will have a cushioning effect as the airplane
gets closer to the ground. This will feel like a small increase in the load factor
for the plane approaching the ground. The magnitude of this ground
effect load factor will be proportional to the ratio of the IGE lift coefficient
to that required for the nominal approach angle out-of-ground-effect
(OGE), or
CLIGE
NZIGE ¼ (19:9)
CLOGE

Example 19.2
An airplane approaches the ground for landing at a nominal –3-deg flight
path. It has a wing span of 100 ft with an aspect ratio of 9. The pilot maintains
a constant thrust and, as ground effects begin, the pilot elects to maintain a
constant angle of attack all the way to near touchdown height; that height is
h/b ¼ 0.075 (MAC height). What would be the impact on the load factor?
The lift coefficient in the approach OGE is 1.18, and the angle of attack for
zero lift is –6.5 deg in the landing configuration.
Solution: We first compute the OGE lift curve slope using the approximation
given by Eq. (19.7) using a ground effect factor of 1.0; that slope is 5.14 rad21
(about 0.09 per degree). Then we compute the angle of attack required to fly at
the specified lift coefficient of 1.18; we get
CL ¼ (a  a0 ) CLa ¼ (a þ 6:5 deg) CLa
a ¼ 6:68 deg
The pitch attitude on approach for a nominal flight path angle of –3 deg
would be
u ¼ a þ g ¼ 3:68 deg
For the next step, we make an assumption that the zero-lift angle of attack
remains unchanged—a reasonable assumption. With this assumption, we can
now create lift curve slopes from Eq. (19.7) to get the chart shown in Fig. 19.12.

(Continued)
908 Operational Aircraft Performance and Flight Test Practices

Example 19.2 (Continued)

1.6

h/b = 0.1
1.4
h/b = 0.5
OGE
1.2

0.8
CL

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8 9
Angle of attack (deg)

Fig. 19.12 CL increase at constant a for Example 19.2.

Now, if the airplane comes in at a constant angle of attack per the

problem statement, the lift coefficient resulting from ground effects will
not change much until 50 ft AGL (h/b ¼ 0.5), going from 1.18 to about
1.20. From Eq. (19.9), this would yield a load factor of 1.017, typically
within the noise of a typical approach. Below 50 ft AGL, the lift coefficient
will start increasing faster as h/b reduces; by a height of 10 ft AGL (h/b ¼
0.10), the lift coefficient would be 1.32 for a load factor of 1.114 if the airplane
angle of attack was maintained constant. Using this, one can generate the
load factor curve for a constant angle of attack approach to touchdown
(see Fig. 19.13).
Note how the load factor rapidly increases as the airplane nears the
ground. This will have a cushioning effect in that an increasing load factor
will introduce a curved flight path that will tend to reduce the flight path
at touchdown, helping with the flare.
The previous example did not consider the impact of the ground proxi-
mity on the airplane’s drag; it simply stated that the thrust was maintained to
a constant level, the OGE of which would result in a steady approach angle.
 CHAPTER 19 Landing Performance 909

1.2

AR = 9
1.15 b = 100 ft
h/bmin = 0.075

1.1
Nz

1.05

1
Touchdown

0.95

0.9
0 20 40 60 80 100 120 140 160 180 200
Wing MAC altitude AGL (ft)

Fig. 19.13 Load factor increase in ground effect for Example 19.2.

But, in the same way the ground proximity impacts the generation of lift, it
will also impact the drag due to lift. The analogy of increasing effective
wing aspect ratio while approaching the ground, as expressed by Eq. (19.8),
can be used to quantify the impact.

1
KIGE ¼ ¼FK (19:10)
p ARIGE e

The induced drag would then be

CDi ¼ F K CL2 (19:11)

The airplane’s drag would thus tend to reduce as the airplane approaches
the ground, but only the induced drag part of the total drag would reduce.
The total drag reduction is thus a combination of airplane initial lift coeffi-
cient (combination of weight and airspeed), wing aspect ratio, and ground
proximity. This is a complex mix of parameters that is best explored
through an example. The typical aspect ratio for transport category airplanes
lies between 7 and 12, so we will use a reference of 9 for the discussion per
Example 19.2.
910 Operational Aircraft Performance and Flight Test Practices

Example 19.3
For the airplane used in Example 19.1, at the reference landing speed VREF
of 140 KTAS, plot the steady level flight drag of this airplane over a small
speed range on either side of VREF if the airplane is OGE, and for height
parameters h/b ¼ 0.5, 0.3, 0.1. Assume no impact to the generation of lift at
this time.
Solution: The chart shown in Fig. 19.14 is created.

16,000
AR = 9

14,000
VREF

12,000
Thrust for level
Drag (lb)

flight
OGE, h/b > 1
10,000
h/b = 0.5
8000 h/b = 0.3
Thrust for –3 deg

h/b = 0.1
6000

4000
100 110 120 130 140 150 160 170 180
KTAS

Fig. 19.14 Drag variation with airspeed and height over ground, Example 19.3.

Observations from Fig. 19.14:

• Total drag will decrease significantly (about 20% for this example) from
OGE to a height of h/b ¼ 0.1.
• The total drag decrease would be greater at lower speed.
• This assumes the lift coefficient remains unchanged as the airplane
approaches the ground and the pilot adjusting a to maintain CL ;
therefore, NZ does not increase.
• If the thrust is not changed from that required for OGE (horizontal line
at 140 KTAS, thrust for steady approach at nominal – 3 deg) while the
altitude decreases, the airplane will tend to accelerate if the flight path is
unchanged.

(Continued)
 CHAPTER 19 Landing Performance 911

Example 19.3 (Continued)

• A higher speed, for a given angle of attack, would produce higher lift that
would then increase the load factor, which would tend to also curve the
flight path and reduce the sink rate.
• This acceleration would possibly lead the airplane to want to “float” (stay
above ground) unless the pilot makes a direct input to put the airplane on
the ground (lower the nose, which reduces the angle of attack and lift
coefficient, and/or reduces the thrust).
• The minimum drag speed reduces as the airplane gets closer to the ground.

Flare
As the airplane nears the ground, the pilot will often need to reduce the
sink rate prior to touchdown—the flare height hflare . On a conventional air-
plane, this is done through an elevator input that increases the airplane load
factor. An increase in the approach load factor will lead to a curved flight path
from the initial stable flight path approach.
The pilot must increase the load factor at a flare height that will allow suf-
ficient time to increase the load factor to a target value (usually between 1.05 g
and 1.2 g) (Fig. 19.15). For the purpose of this initial analysis, we will assume
the airspeed remains constant (VREF ) during this maneuver, and that the
load factor n is increased instantaneously to the target value and maintained
until touchdown. The flare height required to bring the airplane’s sink rate
to exactly zero at main landing gear touchdown (see Fig. 19.16) would be
hflare ¼ r ½1  cos (g) (19:12)
where r is the flare radius. As seen in Chapter 12, the radius would be equal to
2
VREF
r¼ (19:13)
g ½n  cos (g)

VREF Start of pull Peak g

3 deg
50 ft
screen height

Touchdown
Flare

Fig. 19.15 Flare segment of the approach to land.

912 Operational Aircraft Performance and Flight Test Practices

r
r  VREF


hflare
Xflare

Fig. 19.16 Flare parameters.

Note that this is an approximation because the term in the brackets

would be [n – 1] at the bottom of the maneuver (flight path equal to
zero). The resulting air distance during the flare would be
 
r
Xflare ¼ r sin (g)  g VW (19:14)
VREF

The first part of Eq. (19.14) is simply the horizontal component of the
flare arc (flight path angle reduced from the approach angle g to zero
flight path). It is the zero-wind air distance in the flare. The second com-
ponent is the correction for the wind speed along the runway; a headwind
is positive here, resulting in a reduced distance compared to the ground.

Example 19.4
An airplane is approaching at 100 KTAS on a 3-deg path. The pilot initiates a
flare at 1.05 g so as to reach exactly zero sink rate at touchdown. What is the
flare height and flare distance? What would be the flare height and distance if
the load factor used was 1.2 g?
Solution: Assuming a constant airspeed (100 KTAS ¼ 168.8 ft/s) in the flare,
the radius of the flare for a 1.05-g load factor would be
2
VREF ð168:8 ft=sÞ2
r¼ ¼  ¼ 17,282 ft
g ½n  cos (g) ft
32:17 2 ½1:05  0:99863
s
For that radius, the flare height would be
hflare ¼ r ½1  cos (g) ¼ 17,282 ft ½1  0:99863 ¼ 23:6 ft
and the flare distance, under no wind conditions, would be
Xflare ¼ r sin (g) ¼ 904:5 ft

(Continued)
 CHAPTER 19 Landing Performance 913

Example 19.4 (Continued)

Had the same maneuver been executed with a load factor of 1.2 g, the
numbers would have been
r ¼ 4398 ft
hflare ¼ 6 ft
Xflare ¼ 230:2 ft

It can be seen that increasing the load factor in the flare will reduce
the flare height and flare distance. Although not shown in this example, redu-
cing the approach speed will have a similar effect. The maximum load factor
for a normal approach speed of a transport category airplane is nominally
1.3 g because the minimum VREF is defined by 14 CFR 25.125(b)(2) as
follows:
A stabilized approach, with a calibrated airspeed of not less than VREF, must
be maintained down to the 50-foot height.
(i) In non-icing conditions, VREF may not be less than:
(A) 1.23 VSR0 ;
(B) VMCL established under §25.149(f); and
(C) A speed that provides the maneuvering capability specified in
§25.143(h).

And by 14 CFR 25.143(h):

The maneuvering capabilities in a constant speed coordinated turn at
forward center of gravity, as specified in the following table, must be free
of stall warning or other characteristics that might interfere with normal
maneuvering:

Configuration Speed Maneuvering Thrust/power

bank
Angle
Landing VREF 408 Symmetric for -38
flight path

The 40-deg bank angle provides a maneuvering of 1.305 g (see Chapter 14).
A maximum of 1.2 to 1.25 g can reliably be achieved by the pilot without
risking triggering the stall warning.
Not discussed is the finite time it takes to achieve the target load factor.
With a sharp elevator input from the pilot, the airplane can possibly produce
as much as 0.5 g/s (typical transport category airplane). So going from a
steady load factor of about 1.0 on a 3-deg approach to a 1.2-g peak in the
flare would require 0.4 s. If the airspeed on the approach is 130 KTAS,
giving a sink rate of 11.48 ft/s, then the airplane would travel a vertical
914 Operational Aircraft Performance and Flight Test Practices

distance of 4.6 ft while the load factor is increased to reach the target
value of 1.2 g.
It should be noted that the peak g is at the pilot’s discretion (not man-
dated in the flight manual), and no specific value should be targeted;
the pilot should control the landing. The numbers provided here are for
the purpose of decomposing the events during the landing. It should also
be noted that once the flare is initiated, the pilot can control the flight
path and push the airplane down if he or she feels it is floating (flared too
soon or too much); that is, the pilot should land the plane.
Figure 19.17 shows a typical variation of the load factor in a landing. In
this case, the peak load factor is approximately 1.05 to 1.07 g. We included
in the figure the path to touch down with no flare (touchdown a little over
time 5 s). Note how the airplane sink rate stops as the airplane gets near
2 ft above ground; here the pilot reduced the load factor to slightly less
than 1.0 to allow the airplane to touch down. The touchdown in this case
is seen as a spike in the load factor (gear contact load). For this example,
the flare started at about 25 ft above ground, and the peak load factor was
reached in about 0.3 s or a rate near 0.15 g/s.
One subject not covered in this flare analysis discussion is the impact of
the center of gravity (CG) on the elevator required to achieve the desired load
factor. On traditionally configured transport category airplanes where the
pitch control is through an elevator near the tail of the airplane, the critical

60 1.15

50 ft screen height
50 1.1
Load factor
Start flare
RA

40 1.05
DA
LT

Load factor (g)

RADALT (ft)

30 1
Overflared, push nose down

20 0.95
Touchdown

10 0.9

No flare air time

0 0.85
0 1 2 3 4 5 6 7 8
Time (s)

Fig. 19.17 Flare path and load factor.

 CHAPTER 19 Landing Performance 915

100

90

80
% Max elevator

70

60

50

40

30
All test points 7.5 deg approach Linear (All test points)
20
0 5 10 15 20 25
CG (% MAC)

Fig. 19.18 Elevator required to flare vs airplane CG [2].

CG location is the forward limit for a given weight. Figure 19.18 (from [2])
shows such an example.
Because the ability to flare is dependent on the longitudinal control, the
VREF speed used will need to be validated by testing. Although we will cover
this in the next chapter, suffice to say that testing usually uses a normally
expected variation of the approach speed (VREF – 5), where the pilot must
show that by using normal landing techniques, with a resulting touchdown
speed about 5 kt slower than normal, the resulting sink rate should not be
greater than 6 ft/s.

Touchdown
For most touchdowns, the sink rate is not completely reduced, but rather
a small value remains that is absorbed by the main landing gear shock struts.
Typical touchdown sink rates vary from less than 1 ft/s (very smooth) to
approximately 4 ft/s (see Fig. 19.19); anything higher is usually considered
uncomfortable for passengers. The combination of airspeed and sink rate
at touchdown can be converted to a flight path at MLG contact with the
ground [Eq. (19.15) and Fig. 19.19]:
 
1 SRTD
gTD ¼ sin (19:15)
V
So, for a touchdown sink rate of less than 4 ft/s from a nominal approach
of 3 deg, the airplane’s flight path would need to be reduced by about 2 deg
916 Operational Aircraft Performance and Flight Test Practices

4.5

3.5
Flight path at touchdown (deg)

Nominal Sink rate

3 approach at touchdown
(ft/s)
2.5 10

2 8

1.5 6

1 4

2
0.5
1
0
80 90 100 110 120 130 140 150
Airspeed at touchdown (KTAS)

Fig. 19.19 Flight path angle at touchdown, per Eq. (19.15).

for airspeeds of more than 100 KTAS. For all practical purposes, this is also a
2-deg change in pitch attitude.
Of course, the altitude above ground where the lower sink rate is
adopted will have a significant impact on the landing distance, so the
pilot’s ability to achieve the desired touchdown flight path (both how fast
and how precisely) will impact the altitude at which the flare maneuver
will be initiated.

Example 19.5
A pilot is preparing to land. The airspeed is 200 ft/s (about 118.5 KTAS), and
the airplane is on a nominal approach of 3 deg. The pilot initiates a rapid flare
and reaches a steady state sink rate by 10 ft above ground. What is the total
landing distance if the steady sink rate is 4 ft/s? What is it if the steady sink
rate is 2 ft/s?
Solution: As a simplified analysis, assume that the air distance from 50 ft AGL
to 10 ft AGL is at a constant flight path angle of 3 deg; the air distance covered
during that time would be 764 ft. For the last 10 ft altitude, two scenarios are
considered—one where the sink rate is 4 ft/s and one where it is 2 ft/s. For the

(Continued)
 CHAPTER 19 Landing Performance 917

Example 19.5 (Continued)

first one, the air time would be
10 ft
t¼ ¼ 2:5 s
4 ft=s
During that time, the air distance covered would be
X  200 ft=s  2:5 s ¼ 500 ft
Combined with the 764-ft distance for the initial 40-ft altitude, this gives a
total air distance at touchdown of 1264 ft. Repeating this with the 2 ft/s case
would yield an air time for the last 10 ft of altitude of 5 s and an air distance of
1000 ft, for a total air distance from 50 ft of 1764 ft, a much-increased distance
(see Fig. 19.20).

VREF

3 deg
50 ft
Screen height
10 ft
191 ft 500 ft 4 fps TD 2 fps TD

1000 ft
955 ft

Fig. 19.20 Landing distances for Example 19.5.

It becomes clear from this example that adopting a lower sink rate prior to
touchdown at the higher altitudes will significantly increase the air distance;
Fig. 19.21 shows the order of magnitude increase for the example. Using the
same simplified analysis as the example, one could deduce that the airplane
could have had the same total air distance with a 2 ft/s touchdown as for
the 4 ft/s case if the flare had started at 3.8 ft above ground rather than
10 ft; however, in this case, the pilot would have had to reach that sink rate
less than 0.4 s prior to touchdown, which does not leave much room for error.
If, instead of adopting a low sink rate from a higher altitude, the pilot
allowed the airplane to touch down at the reduced sink rate by timing the
flare, the overall landing distance would be reduced (see Fig. 19.22).
hflare ¼ r ½cos(gTD )  cos (ga ) (19:16)
 
r
Xflare ¼ r[ sin (ga )  sin(gTD )]  (ga  gTD ) VW (19:17)
VREF

(Continued)
918 Operational Aircraft Performance and Flight Test Practices

Example 19.5 (Continued)

2000

1800
Air distance from 10 ft AGL to touchdown (ft)

1600

1400

1200 Average airspeed

200 ft/s (118.5 KTAS)
1000

800

600

400

200

0
0 1 2 3 4 5 6 7 8 9
Sink rate from 10 ft AGL to touchdown (ft/s)

Fig. 19.21 Impact of sink rate at touchdown on distance for Example 19.5.

r
VREF
r
a
TD
hflare

Xflare

Fig. 19.22 Flare to a nonzero touchdown sink rate.

Redoing Example 19.4 but assuming a sink rate at touchdown of 4 ft/s instead
of the zero sink rate value used in the example, keeping the airspeed constant
in the maneuver, would give the following numbers for the 1.05-g flare:
gTD ¼ 1:358 deg

(Continued)
 CHAPTER 19 Landing Performance 919

Example 19.5 (Continued)

hflare ¼ 18:8 ft
Xflare ¼ 495 ft
This is not a significant change in flare height, but a 45% reduction in
landing flare distance. For the 1.2-g flare of the same example, the numbers
would have been
gTD ¼ 1:358 deg
hflare ¼ 4:8 ft
Xflare ¼ 126 ft
There is clearly a distance advantage of flaring as late as possible to
the highest peak g available, but as the flare height reduces, so does the
pilot available time to reduce the sink rate prior to touchdown. In the case
of the example, the sink rate prior to flare was (100 KTAS  sin(3 deg) ¼ 8.83
ft/s), so a flare at 4.8 ft as shown would mean the pilot must have reached the
1.2-g peak load factor just over 0.54 s prior to reaching the ground, which
again leaves not much room for error.

Managing Eye-to-MLG Sink Rate

As discussed in the Global 5000’s accident at the start of the chapter, the
pilot must manage the MLG sink rate for touchdown based on the perceived
sink rate at the cockpit. That sink rate is a function of the airplane’s sink rate
SR at the CG, the pitch rate u̇, and the relative distances of the cockpit and
MLG to the CG. On conventional transport category airplanes, the MLG
resides behind the CG and the cockpit is forward of the CG, so the apparent
sink rate for the pilot will be less than that for the actual sink rate for the
MLG (see Fig. 19.23). The cockpit sink rate and the MLG sink rate can be

10 ft
100 ft

Pitch
rate

Sink rate
at CG

Fig. 19.23 Pilot eye vs MLG positions’ impact on relative sink rate.
920 Operational Aircraft Performance and Flight Test Practices

related as follows:
SRcockpit ¼ SR  u̇ xcockpit (19:18)

SRMLG ¼ SR þ u̇ xMLG (19:19)

Example 19.6
A pilot initiated a flare prior to touchdown. The pilot is located 100 ft ahead of
the CG and the MLG is located 10 ft behind the CG. As the airplane sink rate
(at CG) reduces to 8 ft/s, the airplane reaches a peak pitch rate of 2 deg/s.
What is the pilot perceived sink rate vs the MLG sink rate at this same
moment?
Solution: The sink rate as perceived by the pilot at the moment of reaching
2 deg/s (0.035 rad/s) would be
ft 0:035 rad
SRcockpit ¼ SR  u̇ xcockpit ¼ 8  100 ft ¼ 4:5 ft=s
s s
This would be perceived as an acceptable sink rate for touchdown. But at
the same moment, the MLG would have a sink rate of
ft 0:035 rad
SRMLG ¼ SR þ u̇ xMLG ¼ 8 þ 10 ft ¼ 8:35 ft=s
s s
Should the airplane touch down at this same moment, the resulting load
factor through the MLG absorbing the high sink rate would definitely be per-
ceived as a very heavy landing.
Steady (zero pitch rate) pitch obviously gives best indication of sink rate to
the pilot, but adopting lower sink rate at a higher altitude means longer
landing air distance. The balance between runway distance and when to
flare is always a delicate one.
Another item affecting the management of eye-to-MLG coordination is
the approach speed management. Most approaches are done at a reference
airspeed VREF that is a constant stall speed ratio (i.e., 1.23 VSR ); therefore,
there is a constant angle of attack (AoA). A constant AoA and a typical
approach flight path of –3 deg lead to a repeatable pitch attitude on approach
with its associated field of view. If the airplane deviates significantly from
VREF , the AoA may change enough to noticeably affect the pilot’s field of
view, which may change the expected pilot response in the flare.
A different approach path than the nominal –3 deg, strong winds, or event
runway width will also impact the pilot field of view on approach.

Tendency to Float
An airplane can experience a tendency to float when, as it gets very
close to the ground, it may feel like it will refuse to touch down. Some of
 CHAPTER 19 Landing Performance 921

the causes of the tendency to float were described in the section on

ground effects and the observations made for Fig. 19.14. In short, the fol-
lowing will make it harder for the airplane to touch down in a short
distance
• Increased airplane lift that tends to shallow the flight path
• Ground effects affecting the lift curve slope
• Reduced drag that may make the airplane accelerate
• Low thrust deficit and engine response that does not easily allow flight
path control through thrust modulation
The following helps reduce the ground effects and tendency to float:
(1) Having sufficient thrust deficit with a rapid engine response (rapid
spool-down)

Total Air Distance

The total air distance from 50 ft above ground (AGL) to touchdown
can be modeled as a single event or broken down into segments. We will
expand on acceptable means of compliance in the next chapter. Mathemat-
ically, we will break it down into two segments, prior to and after the start of
the flare.
The air distance from 50 ft AGL to just prior to flare can be
approximated as
(50 ft  hflare ) (50 ft  hflare )
Xa1 ¼  VW (19:20)
tan(g) SR
where:
hflare ¼ Flare height
SR ¼ Average sink rate from 50 ft AGL to hflare
VW ¼ Average wind speed along runway (in this case, the headwind
value is positive; tailwind is negative)
g ¼ Nominal flight path

The flare distance was covered extensively in the flare section of this
chapter, with Eq. (19.17) representing the second segment (Xa2 ) of the
total air distance (Xa ).
Xa ¼ Xa1 þ Xa2 (19:21)

Use of Thrust Reversers in Flight

The air distance is established by the use of prescribed crew procedures
that result in repeatable air distance. For most transport category airplanes,
this is limited to thrust reduction and flare prior to touchdown.
922 Operational Aircraft Performance and Flight Test Practices

25.101 General (f) Unless otherwise prescribed, in determining the

accelerate-stop distances, takeoff flight paths, takeoff distances, and
landing distances, changes in the airplane’s configuration, speed, power,
and thrust, must be made in accordance with procedures established by
the applicant for operation in service.

Most Western manufacturers do not allow the use of thrust reversers in flight
for the purpose of reducing the approach distance in order to meet the
requirements of §25.101(h).

25.101 General
(h) The procedures established under paragraphs (f) and (g) of this section
must—
(1) Be able to be consistently executed in service by crews of average
skill;
(2) Use methods or devices that are safe and reliable; and
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.

This is not to say that thrust reversers cannot be selected prior to touchdown,
but that their deployment will generally be prevented until the airplane
reaches a condition of weight on wheels (WoW) if their use on the ground
is shown to be safe.
Some Russian-built jetliners and some military airplanes do use thrust
reversers in-flight to shorten the air distance. Transport category airplane
certification requirements do not preclude the use of thrust reversers in
flight provided that they meet the requirements of §25.933.

25.933 Reversing Systems

(a) For turbojet reversing systems—

(1) Each system intended for ground operation only must be designed so
that during any reversal in flight the engine will produce no more
than flight idle thrust. In addition, it must be shown by analysis or
test, or both, that—
(i) Each operable reverser can be restored to the forward thrust
position; and
(ii) The airplane is capable of continued safe flight and landing under
any possible position of the thrust reverser.
(2) Each system intended for inflight use must be designed so that no
unsafe condition will result during normal operation of the system, or
from any failure (or reasonably likely combination of failures) of the
reversing system, under any anticipated condition of operation of the
airplane including ground operation. Failure of structural elements
need not be considered if the probability of this kind of failure is
extremely remote.
(3) Each system must have means to prevent the engine from producing
more than idle thrust when the reversing system malfunctions,
except that it may produce any greater forward thrust that is shown
to allow directional control to be maintained, with aerodynamic
means alone, under the most critical reversing condition expected
in operation.
 CHAPTER 19 Landing Performance 923

Thrust reversers can be used in flight to steepen the approach angle (increas-
ing the thrust deficit) while maintaining a steady approach speed, or they can
be used just prior to touchdown to increase the deceleration rate of the plane.
Most commercial airplanes (transport category) have enough thrust deficit to
meet the nominal –3-deg approach, so the use of thrust reversers is not
deemed required for flight path control. For the second case, just prior to
touchdown, a good coordination is required by the crew because the deploy-
ment of the thrust reversers will result in a sudden increase in sink rate that
must be compensated for by a nose-up command to prevent a hard landing.
This may make it hard to meet §25.101(h)(1).

Thailand CAB On 26 May 1991, a Lauda Air Boeing 767-300, en route from
Bangkok to Vienna, suffered a catastrophic event during the climb to cruise alti-
tude when the left thrust reverser deployed while the engines were at climb
thrust. At the moment of the deployment, the airplane was climbing through
FL247 at Mach 0.78. Although the deployment of the thrust reverser caused
the engine to start spooling down towards idle, the event was so rapid that
loss of control occurred within 10 s with the airplane losing both directional
control and roll control, followed by a breakup in flight and subsequent crash.

Need for Steady Approach

Several parameters affect the air distance; most were discussed in the pre-
ceding material. To achieve repeatable landing distance with reduced risk of a
hard landing, it is highly desirable to achieve a stable approach. A stable
approach can be defined by:

• Adopting the final landing configuration before reaching a certain height

above ground (typically 1500 ft to 500 ft). This ensures a steady
airplane drag.
• Stabilizing on approach speed (either Vapproach or VREF ) early enough, say
above 500 ft AGL, to stabilize AoA and flight path.
• A stable airspeed would typically be within þ5 kt to þ0 kt of the desired
speed (either Vapproach or VREF ).
• Some approach procedures will say to maintain Vapproach (VREF plus a
small delta) until landing is assured, slowing down to VREF by 50 ft AGL.
The small delta speed provides increased pitch control. To do this,
sufficient thrust deficit must exist as the throttle is reduced to rapidly go
from Vapproach to VREF .
• Stabilizing the approach path to the nominal value to land with a small
tolerance on deviation (typically +0.5 deg).
• Larger deviations could result in a very large increase in sink rate near
the ground, which may lead to a hard landing.
924 Operational Aircraft Performance and Flight Test Practices

• The nominal sink rate of transport category airplanes on a nominal

–3-deg approach varies between 600 fpm and 800 fpm.
• Figure 19.19 showed the impact of flight path on the airplane sink rate.
It is recognized that an approach is a dynamic condition with varying
winds, airspeeds, pitch attitudes, and so forth. Using defined stabilization
criteria helps reduce crew workload and may help the crew to decide if a
go-around is warranted instead of a continued approach to land.
We recopied Fig. 19.5 here as Fig. 19.24 to review the previous discussion
about the need to decide whether the approach can be continued. It shows a
stable initial part of the approach from above 1000 ft AGL to about 250 ft
AGL, where the previously described criteria were met (although the airspeed
tended to be a little below the approach speed of 132 KCAS). At about 250 ft
AGL, frontal wind gusts occurred, as seen through the rapid increase in the
airspeed; at the same time, the flight path increased rapidly by about 1 deg.
In this event, the pilot had enough thrust control [thrust lever angle (TLA)
movement followed by rapid thrust change (N1)] to both adjust the flight
path and recover to the desired approach speed, then go down to VREF by
50 ft AGL.

Transition Segment
The airplane transition segment extends from touchdown, as defined in
the previous section, to the point where the airplane adopts a stable configur-
ation and pitch attitude. For transport category airplanes, this would typically
include (see Fig. 19.25):
• MLG touchdown to NLG touchdown derotation phase
• Application of brakes
• Deployment of speed brakes and/or ground spoilers
• Thrust reversers deployment and spool-up to maximum reverse thrust
The only devices that an applicant can take credit for during the AFM
landing distance expansion are defined by §25.101(h) and §25.125(b):
25.101 General
(h) The procedures established under paragraphs (f) and (g) of this section
must—
(1) Be able to be consistently executed in service by crews of average
skill;
(2) Use methods or devices that are safe and reliable; and
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.

25.125 Landing
(b) In determining the distance in paragraph (a) of this section:
(1) The airplane must be in the landing configuration.
(2) A stabilized approach, with a calibrated airspeed of not less than
VREF, must be maintained down to the 50-foot height.
Airspeed KCAS Airspeed KCAS Approach speed 132 KCAS Wind gust
135 VREF127 KCAS
130
125 Turbulence
120

Altitude ft
1000
Altitude ft

50 ft AGL
500

0
FPA deg
0
–1
FPA deg

–2 Nominal 3 deg

CHAPTER 19
–3
–4
–5
TLA % full N1 % max
TLA % full

N1 % max
30 80
20 60

Landing Performance
10 40
20
0 10 20 30 40 50 60 70 80 90

Time (s)

Fig. 19.24 Parameters of interest for stable approach.

925
926 Operational Aircraft Performance and Flight Test Practices

MLG NLG Decel Decel

touchdown touchdown device 1 device 2

t1 t2 t3

Fig. 19.25 Transition segment.

(3) Changes in configuration, power or thrust, and speed, must be made

in accordance with the established procedures for service operation.
(4) The landing must be made without excessive vertical acceleration,
tendency to bounce, nose over, ground loop, porpoise, or water loop.
(5) The landings may not require exceptional piloting skill or alertness.
(c) For landplanes and amphibians, the landing distance on land must be
determined on a level, smooth, dry, hard-surfaced runway. In addition—
(1) The pressures on the wheel braking systems may not exceed those
specified by the brake manufacturer;
(2) The brakes may not be used so as to cause excessive wear of brakes
or tires; and
(3) Means other than wheel brakes may be used if that means—
(i) Is safe and reliable;
(ii) Is used so that consistent results can be expected in service; and
(iii) Is such that exceptional skill is not required to control the
airplane.

Just like for the rejected takeoff distance, the landing distance on a dry
runway cannot take any credit for the use of thrust reversers for the
purpose of computing the minimum landing distance for the AFM. This is
not to say that the crew cannot use them for normal landings, just that the
published landing distance on a dry runway will not account for their use.
Credit for the use of thrust reversers can be used for wet or contaminated
runways:
25.125(g) If any device is used that depends on the operation of any engine,
and if the landing distance would be noticeably increased when a landing is
made with that engine inoperative, the landing distance must be determined
with that engine inoperative unless the use of compensating means will result
in a landing distance not more than that with each engine operating.

The derotation time from MLG touchdown to NLG touchdown has a big
impact on the total transition time, especially for large airplanes or airplanes
with a high pitch attitude on touchdown. That time is proportional to the
difference from pitch attitude at touchdown to nose gear on ground and
the average nose-down pitch rate. Just like the main landing gear has a
design maximum sink rate at touchdown, so does the nose landing gear.
FAA AC 25-7D specifies that the nose gear touchdown rate must not
exceed 8 ft/s for certification testing. Exceeding the maximum nose gear
sink rate could lead to structural failure or damage.
 CHAPTER 19 Landing Performance 927

Applying brakes prior to main landing gear touchdown (i.e., landing with
locked brakes) is also not allowed for certification testing due to the risk of
generating an excessive nose-down pitch rate. Several manufacturers
also preclude the use of brakes prior to the nose gear being on the ground
for the same reason—preventing excessive nose-down pitch rate.

Transition Segment Distance

The ground distance covered during this phase will depend on the time
required to derotate the aircraft, followed by the time to deploy each
additional deceleration device
DtT ¼ Dtdr þ DtB þ DtSB þ . . . (19:22)
where:
DtT ¼ Transition time
Dtdr ¼ Derotation time
DtB ¼ Time to brake application
DtSB ¼ Time to speed brake/ground spoilers full deployment

As per the rejected takeoff testing, some of these events are sequential
and some occur in parallel. Testing defines the minimum total time; for
the purpose of the AFM expansion, a minimum of 1 s per pilot event is
required. Figure 19.26 shows a typical transition segment with various
events of interest; for reference, major vertical lines are spaced 1 s apart.
Each event in this landing is represented by a vertical dashed line on
Fig. 19.26. The touchdown is followed by an immediate derotation (pilot
pushing the nose down); as seen on the pitch strip, this event takes approxi-
mately 2 s for a derotation of 10 deg. As the airplane derotates, the pilot
selects speed brakes, which take approximately 1.5 s to deploy (vertical line
at spoilers out). Soon after the nose landing gear is on the ground, the
pilot starts braking (Brakes ON vertical line). Finally, after a certain
amount of time on the ground, the engines automatically spool down to
ground idle (a feature on some engines), as seen by the engine fan
speed N1 spooling down. The total transition time demonstrated DtT in
this case is approximately 4.5 s. This event included several pilot actions
(derotation, speed brake selection, brakes), so at the time of data expansion
for the AFM, these will need to be a minimum of 1 s in duration or as demon-
strated, whichever is longer. We will cover this in more detail in the next
chapter.
The transition time usually occurs over a short period, typically 2 to 6 s
total , and is often modeled through an average demonstrated deceleration
over the transition time. For no brake application during the derotation,
one could expect a deceleration aT of the order of –0.02 to – 0.1 g [negative
value for Eq. (19.23), the speed reduces with time], higher if brakes are used
928 Operational Aircraft Performance and Flight Test Practices

120 VREF
110
Airspeed

100
90
80
100

Brakes ON
RADALT

50

0
100
Whl Spd

50
Touchdown

NLG down
0
15
10
Pitch

5
0
−5
30
Spd Brk

20
Spoilers out
10
Ground idle
Pilot
0 action
60
50
N1

40
30
20
Time tT

Fig. 19.26 Transition segment events.

during that time.

VB ¼ VTD þ aT DtT (19:23)

The resulting ground distance is

 
(VTD þ VB )
XT ¼  VW DtT (19:24)
2
 CHAPTER 19 Landing Performance 929

Ground Run Segment

In this part of the landing, the airplane has all gears on the ground, all
deceleration devices extended, and engine at the minimum ground thrust.
The airplane pitch attitude is essentially constant in a condition that is essen-
tially similar to the rejected takeoff condition analyzed in previous chapters.
The ground force equation along the runway takes on the following form for
a level runway:
 
1 2 1 2
T  rSL s VTAS S CD,L  mB W  rSL s VTAS S CL,L
2 2
W dV
¼ (VTAS  VW ) (19:25)
g dX
where CL;L and CD;L are the lift and drag coefficient, respectively, in the
landing configuration with all deceleration devices and ground spoilers
deployed. The braking coefficient mB typically varies between 0.3 and 0.5
on a dry concrete/asphalt runway. The distance can then be calculated by
numerical integration as
VðW
ðW =gÞ(VTAS  VW ) dV
XGR ¼ (19:26)
1 2
VB T  mB W  rSL s VTAS S (CD,L  mB CL,L )
2
To solve this equation, we will make a few assumptions, as we did for the
rejected takeoff equations. They include:
• Constant weight, which is very reasonable given the low engine thrust (very
low fuel burn)
• Constant braking coefficient (or has a known variation vs ground speed)
If we also assume that the aerodynamic forces are relatively small compared
to the deceleration force provided by the brakes, Eq. (19.26) reduces to

(VB  VW )2
XGR ¼    (19:27)
T
2g  mB
W

Total Landing Distance

The airplane’s total landing distance is the sum of the air distance, tran-
sition distance, and ground run distance. It is also referred to as the actual
landing distance (ALD).
ALD ¼ Xa þ XT þ XGR (19:28)
Through testing, the manufacturer will show the minimum value that the
airplane can accomplish by design. As we have discussed earlier, the first two
930 Operational Aircraft Performance and Flight Test Practices

segments are highly pilot technique dependent; good crew procedures are
required to get a consistent value. The later segment is less dependent on
pilot technique but highly dependent on braking system performance and
runway surface conditions.
The ALD can be published in the airplane flight manual as being the best
(shortest) landing distance capability of the plane. Some transport category
airplanes may not be allowed to use this value operationally.

Operational Requirements
We have covered the basic certification requirements for transport
category airplanes in the preceding text. Those requirements dictate the
minimum landing distance capability of the airplane that can be used,
the ALD. That ALD can be achieved only if a precise approach is used—
the pilot does minimum flare, executes a maximum rate derotation, and
applies maximum braking until full stop.
This type of landing would not be easy to achieve consistently, of course,
especially if atmospheric conditions (winds, turbulence, etc.) were less than
ideal. So operationally, the FAA provides additional requirements that
must be applied to the ALD for use on commercial flights, for Part 121,
Part 135, and Part 91 Sub-part K. Of interest will be the definition of the
landing field length (LFL).
Part 121, Operations (Air Carriers), provides the following operational
requirements for landing performance (underline added by author):
FAA 14 CFR §121.195 Airplanes: Turbine engine powered: Landing limit-
ations: Destination airports [Amdt. 121-9, 30 FR 8572, July 7, 1965]
(a) No person operating a turbine engine powered airplane may take off that
airplane at such a weight that (allowing for normal consumption of fuel
and oil in flight to the destination or alternate airport) the weight of the
airplane on arrival would exceed the landing weight set forth in the
Airplane Flight Manual for the elevation of the destination or alternate
airport and the ambient temperature anticipated at the time of landing.

This paragraph means an operator must determine the expected landing

weight of the airplane prior to dispatch and verify that the airplane is
capable of being landed at the planned airport or the selected alternate air-
ports for the planned operation. This includes meeting the minimum
climb capability for go-around.
(b) Except as provided in paragraph (c), (d), or (e) of this section, no person
operating a turbine engine powered airplane may take off that airplane
unless its weight on arrival, allowing for normal consumption of fuel and
oil in flight (in accordance with the landing distance set forth in the
Airplane Flight Manual for the elevation of the destination airport and
the wind conditions anticipated there at the time of landing), would
allow a full stop landing at the intended destination airport within 60
percent of the effective length of each runway described below from a
point 50 feet above the intersection of the obstruction clearance plane
 CHAPTER 19 Landing Performance 931

and the runway. For the purpose of determining the allowable landing
weight at the destination airport the following is assumed:
(1) The airplane is landed on the most favorable runway and in the most
favorable direction, in still air.
(2) The airplane is landed on the most suitable runway considering the
probable wind velocity and direction and the ground handling
characteristics of the airplane, and considering other conditions such
as landing aids and terrain.

Paragraph b stipulates that the weight of the airplane must allow for a full
stop within 60% of the available runway length using the AFM procedure
that was validated by flight testing (the ALD). This has been shown to
provide sufficient margin for expected operational deviations.

ALD
LFL ¼  1:667 ALD (19:29)
0:6
(c) A turbopropeller powered airplane that would be prohibited from being
taken off because it could not meet the requirements of paragraph (b)(2)
of this section, may be taken off if an alternate airport is specified that
meets all the requirements of this section except that the airplane can
accomplish a full stop landing within 70 percent of the effective length of
the runway.
(d) Unless, based on a showing of actual operating landing techniques on
wet runways, a shorter landing distance (but never less than that
required by paragraph (b) of this section) has been approved for a
specific type and model airplane and included in the Airplane Flight
Manual, no person may takeoff a turbojet powered airplane when the
appropriate weather reports and forecasts, or a combination thereof,
indicate that the runways at the destination airport may be wet or
slippery at the estimated time of arrival unless the effective runway
length at the destination airport is at least 115 percent of the runway
length required under paragraph (b) of this section.

Paragraph d adds another 15% margin on top of the requirements of para-

graph b to account for less braking capability on wet runways.

LFLwet ¼ 1:15 LFLdry (19:30)

The exception, as allowed by paragraph d, is that a lesser value than the

one defined by Eq. (19.30) can be used if the airplane manufacturer suc-
cessfully demonstrated it through testing. That lesser value cannot be
less than the value determined for dry runway and shown in Eq. (19.29).
One type of runway that may have some braking benefit when wet is a
grooved runway.
Part 135 Commuter and on demand operations
135.385 Large transport category airplanes: Turbine engine powered:
Landing limitations: Destination airports. [Amdt. 135-91, 68 FR 54588,
Sept. 17, 2003] Paragraphs (a) through (d) read very much like the
requirements of 121.195 shown above. It also includes paragraph (f) that
states:
932 Operational Aircraft Performance and Flight Test Practices

( f) An eligible on-demand operator may take off a turbine engine powered

large transport category airplane on an on-demand flight if all of the
following conditions exist:
(1) The operation is permitted by an approved Destination Airport
Analysis in that person’s operations manual.
(2) The airplane’s weight on arrival, allowing for normal consumption of
fuel and oil in flight (in accordance with the landing distance in the
Airplane Flight Manual for the elevation of the destination airport and
the wind conditions expected there at the time of landing), would
allow a full stop landing at the intended destination airport within 80
percent of the effective length of each runway described below from a
point 50 feet above the intersection of the obstruction clearance plane
and the runway. For the purpose of determining the allowable landing
weight at the destination airport, the following is assumed:
(i) The airplane is landed on the most favorable runway and in the
most favorable direction, in still air.
(ii) The airplane is landed on the most suitable runway considering
the probable wind velocity and direction and the ground
handling characteristics of the airplane, and considering other
conditions such as landing aids and terrain.
(3) The operation is authorized by operations specifications.

These regulations provide guidance for planning the landing performance

prior to the dispatch of the airplane. Following a Southwest Airlines accident
at Chicago Midway airport in 2005, the FAA issued an Advisory Circular, AC
25-32 [3], to provide operators and airplane manufacturers with a standar-
dized approach for computing time-of-arrival landing data based on a stan-
dardized runway condition assessment matrix (RCAM). We will discuss
AC 25-32 in the next chapter.

NTSB AAR-07/06 On 8 Dec. 2005, about 1914 central standard time, South-
west Airlines (SWA) flight 1248, a Boeing 737-7H4, N471WN, ran off the
departure end of runway 31C after landing at Chicago Midway International
Airport, Chicago, Illinois. The airplane rolled through a blast fence and an
airport perimeter fence and onto an adjacent roadway, where it struck an
automobile before coming to a stop. There were 11-kt tailwinds at the time
of the accident.
The National Transportation Safety Board determined that the probable
cause of this accident was the pilots’ failure to use available reverse thrust
in a timely manner to safely slow or stop the airplane after landing, which
resulted in a runway overrun.
Contributing to the accident were Southwest Airlines’ (1) failure to
provide its pilots with clear and consistent guidance and training regarding
company policies and procedures related to arrival landing distance calcu-
lations; (2) programming and design of its on-board performance computer,
which did not present inherent assumptions in the program critical to pilot
decision making; (3) plan to implement new autobrake procedures without
a familiarization period; and (4) failure to include a margin of safety in the
arrival assessment to account for operational uncertainties.
 CHAPTER 19 Landing Performance 933

Exercises
1. An airliner is on a nominal 3-deg ILS beam (approach angle with
respect to the ground) approach under calm wind conditions; the
approach speed is 125 KTAS. The resulting flight path angle is an
average of 2.8 deg (Fig. 19.27). What is the horizontal wind
component? The airplane weight is 186,000 lb, and the wing area is
1600 ft2 . What is the airplane thrust deficit? What is the lift-to-drag
(L/D) ratio if thrust is 2% of weight?
2. An airplane is coming in to land at 124 KTAS on a nominal 3-deg
approach. The pilot initiates a strong/sudden pull at 12 ft AGL, which
results in a 2-ft/s touchdown without float. What was the average load
factor during this flare?
3. An airplane comes in to land with an approach speed of 120 KTAS. You
extract the data from the flight recorder and see the data shown in
Fig. 19.28. What is the average load factor in the flare? What is the
resulting flight path radius?

0 400

−0.5 350

−1 300

RADALT
−1.5 250
Angles (deg)

RADALT (ft)

−2 200

−2.5 150
Average flight path angle
−3 100

−3.5 50

−4 0
0 5 10 15 20 25 30 35
Time (s)

Fig. 19.27 Exercise 1 data.

934 Operational Aircraft Performance and Flight Test Practices

60 1.1

50 1.08
RADALT

40 1.06
Load factor

Load factor (g)

RADALT (ft)

30 1.04

20 1.02

10 1

0 0.98
33 34 35 36 37 38 39 40 41
Time (s)

Fig. 19.28 Exercise 3 data.

References
[1] Asselin, M., An Introduction to Aircraft Performance, AIAA Education Series, AIAA,
Reston, VA, 1997.
[2] Asselin, M., Hinson, M., and Storrer, B., “Learjet Model 45 Steep Approach Certifica-
tion,” SAE Paper 2004-01-1805, 2004.
[3] Federal Aviation Administration, “Landing Performance Data for Time-of-Arrival
Landing Performance Assessments,” Advisory Circular AC 25-32, 22 Dec. 2015.
Chapter 20 Landing
Performance
Testing and AFM

Chapter Objective
After the discussion of landing performance theory in Chapter 19, we discuss
landing performance testing here in order to establish a framework to capture
landing performance data and develop the aircraft flight manual (AFM)
landing procedure and model. We will perform data reduction to establish
the performance coefficients and understand the level of testing required to
develop models based on guidance material, showing along the way some of
the scatter we can encounter. Then we discuss expansion of the data for the
AFM and provide some examples of charts for the flight crew. We conclude
with a brief discussion on the special requirements for steep approach.

Preparing for a Landing Distance Test Campaign

J
ust like it was done for takeoff performance testing, landing performance
testing will start early in a flight test program by doing nondedicated
landings. Then, the landing performance technique can be evaluated
and adjusted as required prior to performing dedicated landing distance
testing (see Fig. 20.1). The most important aspect of the landing performance
is the approach airspeed, so the manufacturer must proceed to stall speed
validation and minimum control speed testing prior to being able to properly
evaluate the airplane behavior near the ground at the proper airspeed.
The landing distance extends from 50 ft above ground to full stop, so the
total test time for this event is easily in the order of 30 s, with the air time
(50 ft to touchdown) being anywhere from 3 to 7 s. Over such short dur-
ations, a small deviation in the execution of the maneuver can have signifi-
cant impact on the scatter of the results. For example, Fig. 20.2 shows a
nondedicated test landing where the crew extended the flare and then per-
formed a normal derotation followed by moderate braking and idle reverse
thrust until the speed was down to taxi speed; this was not the shortest
landing distance for this airplane. Testing for landing distance needs a con-
trolled environment (low winds, no turbulence) and precise execution of the
maneuver per an approved procedure [§25.101(f) and (h)].

935
936 Operational Aircraft Performance and Flight Test Practices

Stall speed Landing procedure validation

testing Validation of ground decel and Test
validating VSR transition spread model program
timeline

First Reverse thrust VMCL A/C config Landing distance AFM

flight testing– testing frozen, testing
directional control procedures including abuse
set, speeds testing
validated

Fig. 20.1 Timeline for landing distance testing.

§25.101 General
(f) Unless otherwise prescribed, in determining the accelerate-stop
distances, takeoff flight paths, takeoff distances, and landing distances,
changes in the airplane’s configuration, speed, power, and thrust, must be
made in accordance with procedures established by the applicant for
operation in service.
(h) The procedures established under paragraphs (f) and (g) of this section
must—
(1) Be able to be consistently executed in service by crews of average
skill;
(2) Use methods or devices that are safe and reliable; and
(3) Include allowance for any time delays, in the execution of the
procedures, that may reasonably be expected in service.
(i) The accelerate-stop and landing distances prescribed in §§25.109
and 25.125, respectively, must be determined with all the
airplane wheel brake assemblies at the fully worn limit of their
allowable wear range.

Testing for the Air Segment

The air segment is the part of the landing with perhaps the most scatter in
the data. One must control the flight path angle down to 50 ft [§25.125(b)(2)]
while faced with possible wind gradients and turbulence.
§25.125 Landing
(b) In determining the distance in paragraph (a) of this section:
(1) The airplane must be in the landing configuration.
(2) A stabilized approach, with a calibrated airspeed of not less than
VREF, must be maintained down to the 50-foot height.

Then, how the pilot executes the landing (flare height, thrust cut height, how
much load factor used to stop the descent rate, etc.) will greatly impact the air
time, as discussed in Chapter 19.

Landing Technique Development

An airplane manufacturer will get several chances to review the airplane’s
behavior in landing in the initial part of the program while performing non-
performance landings. One should then take this opportunity to start
 KIAS (kt) Wheel spd (kt) Ground speed (kt) RADALT (ft)
160 200
127.5581 122.8551 116.9675
140
Ground speed (kt)
Wheel spd (kt)
120 150
KIAS (kt)

RADALT (ft)
100
130.5549 124.5 Moderate braking
80 50.22223 100
125.125 117.1846
60 9.508851
Taxi speed off runway 50
40 0.7530437 0.073567
0.21253 0.21253
20
0 0

CHAPTER 20
Pitch (deg) FPA (deg)
6
4.187347 4.618099
4
0.8100581 –0.4233868
–0.3518302
2
Pitch (deg)
FPA (deg)

–1.363998
0

Landing Performance Testing and AFM

–2
–4 –1.039001
–2.773925
–6
TLA (deg) N1 (%)
70
61.21596
60
50 29.5834 29.15625
TLA (deg)

23.125 Idle reverse thrust

N1 (%)

40 55.11128
30
20 0.3125 0.3125
10 25.83499
0
12 L

17 p

1 n

40 50 80 90 100 110 120 130

.7 AG

78 W
.6 ho

33 w
Id 2

85

.5 o
.1 do
2
61 le c

72 G W
85
56 0 ft

NL 3
68 h
5

uc

Time (s)
To

Fig. 20.2 Normal landing event.

937
938 Operational Aircraft Performance and Flight Test Practices

collecting data and adjusting the landing technique. This is considered the
landing technique development phase of the program.
Initial nondedicated test landings will typically be done at faster than
expected final VREF before stall speeds are validated (often 10 to 15 kt
higher). While carrying this extra energy, as the airplane gets into ground
effects, the performance engineer can start to evaluate if the airplane has a
tendency to float, requiring the pilots to push the nose down to touchdown.
As seen in Chapter 19, Fig. 19.14, the drag will reduce as the airplane gets
closer to the ground, and things may be amplified if the airplane’s approach
speed is at or below minimum drag for the landing configuration. (The air-
plane’s airspeed may actually start increasing for a fixed thrust.) The per-
formance engineer, working with the test pilot, must evaluate different
thrust cut heights to see how the pilot perceives the deceleration and the
“feeling of sink.” This can be done at the lower altitudes initially (less
speed loss, possibly even speed increase due to ground effects) and then
higher from the ground, but typically below 50 ft. The test team verifies
the combined effects of engine spool-down rate (thrust reduction) vs the air-
plane sink rate as the ground is approached.
On some airplanes, the power setting to maintain the approach path may
have a strong impact on the field of view (see Fig. 20.3). (This is especially true
for multiengine propeller-driven airplanes.) The test team needs to investi-
gate this impact at various weights and center of gravity (CG) positions in
the course of the landing technique development phase and adjust the
landing procedure as required. For these same airplanes, the feeling of sink
will be amplified with power reduction in the flare.

Flare Technique
An airplane that is approaching at the nominal approach angle (typically
23 deg) will have a given sink rate [Eq. (19.2)]. This sink rate must be

CT1
CL CT3 > CT2 > CT1 V
CT3
CT2
CT1 CT2
V

CT3
V

Fig. 20.3 Thrust impact on pilot field of view on approach.

 CHAPTER 20 Landing Performance Testing and AFM 939

35 4.5

Flight path angle at touchdown (deg)

30 4.0
ch
Sink rate at 50 ft AGL (ft/s)
approa 3.5
7.5 deg
25 proach Nominal approach angle
6.5 deg ap 3.0 Sink rate
20 proach at touchdown
5.5 deg ap 2.5 (ft/s)

4.5 deg approa

ch 10
15 2.0
3.5 deg approach
8
h 1.5
3.0 deg approac 6
10 2.5 deg approac
h
1.0 4
5
Example 0.5 2
1
0 0.0
100 105 110 115 120 125 130 135 140 145 150 80 90 100 110 120 130 140 150
Approach speed, VREF (KTAS) Touchdown airspeed (KTAS)

Fig. 20.4 Flare flight path angle change.

reduced to an acceptable value at touchdown, which is typically less than 4 ft/

s (which results in about 1-deg flight path angle). The flare height is selected
by systematically investigating a combination of flare height from as high as
50 ft above ground (reduce the flight path angle to the desired value at touch-
down; see Fig. 20.4) to as low as touchdown with various engine thrust cut
heights. Flaring too high will lead to longer air distance; flaring too low
will lead to a possible hard landing. For a practical example, we look at
Fig. 20.4 and use the following data: VREF ¼ 125 KTAS, nominal approach
angle of 23 deg, desired touchdown sink rate of 4 ft/s, and expected speed
decay in the flare of 3 kt. Note that the required change in the flight path
angle is of the order of 2 deg. This is not much change in flight path angle,
and given the essentially constant speed (small decay so essentially constant
angle of attack), this is the same as an approximately 2-deg change in pitch
attitude. Figure 20.4 also shows that the airplane was descending at about
12 ft/s through 50 ft, leaving the pilot with less than 4.2 s before reaching
the ground (if no flare)—a short period of time. Ground effects may help
soften the descent, as was discussed in Chapter 19. When one combines
this with the thrust cut height (which may increase the sink rate and
change the pitch of the airplane due to the location of the thrust line),
then it makes for a small increase in pilot workload.
Thus, one can start various combinations of flare height for a given thrust
cut height, get a pilot evaluation of the airplane’s response as well as the man-
euver execution, and then repeat the flare height change with a different
thrust cut height. In the end, the team will select the configuration that pro-
vides the shortest repeatable distance with acceptable pilot workload.
Remember the requirements of §25.101(h)(1).
As a note, flaring at a lower altitude also requires a higher load factor to
arrest the descent [Eqs. (19.12) and (19.13) in Chapter 19], and the airplane
typically only has 1.3-g capability to stall warning in the landing configur-
ation. The testing also needs to be repeated at various CG when the
desired combination of flare height and thrust cut height are selected
940 Operational Aircraft Performance and Flight Test Practices

to see if the combination is still good from a handling point of view at the
other CG.

What to Measure for Air Distance Performance Landing Test

Although in each of the flight testing chapters we discussed instrumenta-
tion later in the chapter, for air distance we bring the subject forward. The air
segment distance is highly dependent on the ability to perform the maneuver
repeatably; one will gain reduced flight testing by monitoring some key par-
ameters during the execution of the test. The first parameter of interest is, of
course, the approach speed. The requirements of §25.125 are such that the
approach airspeed will often be limited by the minimum value of 1.23 VSR ,
which means the airplane will essentially be flying at a constant angle of
attack on approach at all weights. This is good in that one should expect
the behavior of the airplane to be similar (helps with modeling). That air-
speed must be maintained within a test range of +2 kt to help reduce the
test scatter.
The flight path angle on approach dictates the airplane sink rate. The
nominal approach angle is 23 deg for transport category airplanes. FAA
AC 25-7D recommends a tolerance of +0.5 deg. This can be hard to main-
tain if turbulence is present and if the pilot does not have a flight path angle
indication in their field of view. It is therefore recommended to provide the
crew with either a head-up display (HUD) or an experimental display in the
field of view so that the pilot can maintain as stable an approach angle as
possible down to near 50 ft AGL. Then the landing technique and flare
requirements can be properly evaluated. Depending on the data analysis
method (discussed later in this chapter), greater tolerance may be acceptable
to create an air distance model.
Because the landing air distance starts from 50 ft above ground level
(AGL), one must have a reliable way to measure height above ground.
Radar altimeters are usually the best source. Differential global positioning
system (DGPS) is typically a good source for a runway of zero slope; the
height above ground is then the DGPS altitude minus the runway altitude.
The runway slope has a measurable impact on the air distance—not just
on altitude above ground starting point (if using DGPS), but also on the
execution of the air distance test. The air distance from 50 ft to touchdown
is about 1000 ft in length for a nominal 23 deg approach. A 1% runway
downslope from the 50 ft AGL threshold to touchdown will result in a
10-ft runway depression, effectively making it a 60-ft vertical travel and
thus having an impact on time/distance (see Fig. 20.5). The runway slope
will have a nonsymmetric impact on the test results. A downsloping
runway may lead to floating tendencies. (The runway is moving away from
the airplane as the air segment goes on and the pilot flares per normal tech-
nique.) An upsloping runway will be perceived as having a higher sink rate,
and the pilot may flare higher. Runways with minimal slope are best for
 CHAPTER 20 Landing Performance Testing and AFM 941

VREF

3 deg
50 ft Upsloping runway
screen height Zero gradient
Downsloping runway

Fig. 20.5 Impact of runway slope on air distance.

landing air distance testing model development, so the test team should
understand the runway (perform a runway survey prior to the test).
Winds, of course, will have a big impact on the execution of the testing.
First, they will create some turbulence, which makes it harder to maintain
the flight path angle on approach. Then, wind speeds impact the flight path
as compared to the ground (the landing target), and the airplane flight path
angle with respect to the air dictates the amount of thrust required in the
descent [see Eq. 12.32 in Chapter 12], thus somewhat impacting the thrust
spool-down. Finally, if wind is present, there will be a wind gradient as the
airplane approaches the ground, which will also impact the rate of descent
and add to the scatter of the testing. Winds should be minimized (ideally
less than 5 kt, but up to 10 kt with minimum turbulence have been
found acceptable to expedite the testing). The winds should be continu-
ously measured by a dedicated, calibrated wind station that is located
near the expected touchdown point (about 1000 ft from the designated
threshold). Winds should be measured anywhere from 10 ft to 30 ft
above ground.
A good indication of a touchdown is the main landing gear (MLG) wheels
spin-up. On conventional tricycle landing gear configurations, the MLG is
designed to touch down first and absorb the landing loads. On a single-axle
MLG, the wheels spin-up provides a clear and repeatable indication of the
touchdown time. In a good, symmetric landing, both MLGs touch down at
the same time; when testing with a little bit of turbulence or crosswind,
the airplane may touch down in a non-wings-level condition, which results
in one MLG touching down before the other one (see Fig. 20.6). One can
either elect to select the point of first MLG touchdown or use midpoint to
reduce scatter in the test results.
On the larger Part 25 airplanes, one will be faced with multiaxle MLGs
and possible even more than two MLGs (see Fig. 20.7). The delay between
the first set of wheels from the MLG touching down and the last can be any-
where from 0.25 s to over 2 s. With the typical approach speed being around
200 to 300 ft/s, this can significantly add to the scatter of the test results for
the air segment unless a consistent analysis approach is used. This author
recommends using the first set of MLG wheels touchdown (first axle) as
942 Operational Aircraft Performance and Flight Test Practices

Radio height (ft)

250

200
Radio height (ft)

150
50.43216
100

50 0.955514

True airspeed (knot) L MLG (knot) R MLG (knot)

140
121.9802 117.6784
L MLG (kt), R MLG (kt)

120
True airspeed (kt)

100
80
60
40
0 0
20
0 eig
ht
ow
n
30 35 n h 81 45 hd 9 50 55 60
ree 83 uc 55
Sc 41.8 To 9.03
4
Time (s)

Fig. 20.6 Use of wheel spin-up as a means to determine touchdown.

Fig. 20.7 Pick the touchdown point.

 CHAPTER 20 Landing Performance Testing and AFM 943

the reference for the end of the air time and assign the remainder of the time
to the transition segment.
Other indications, should they be recorded, include shock strut com-
pression for MLG, vertical acceleration at the center of gravity (or in the
cockpit), flight path angle and rate of descent [rate of climb (RoC)], and
weight on wheels switches. Figure 20.8 shows the behavior of some of
these parameters for a soft landing (low sink rate at touchdown). The
data presented were sampled at 10 Hz. Note that not all parameters
record a touchdown at the same time or provide a clear enough indication
of a discrete event (like the flight path angle and rate of climb). Weight on
wheels (WoW) switches, for example, require some weight on the wheels to
provide enough input (typically gear compression) to trigger it, but at
touchdown the airplane is still producing a lot of lift. For a softer
landing, the time difference between wheel spin-up and WoW activation
can be distinct. (In Fig. 20.8, there is more than 1 s difference.) This time
difference is somewhat reduced when landings are firmer. This author
prefers to use the wheel speed indicator, when available, as a clear point
for the transition from air to ground; however, one can use any single
value to build the performance model as long as it is used consistently.
(Do not switch parameter midcourse.)

Approach Airspeed Validation

Once the normal landing procedure has been validated at the expected
final VREF conditions, the manufacturer must then validate that this speed
provides acceptable handling characteristics when “normally expected
in-service deviations” occur. The guidance material of FAA AC 25-7D
notes that testing at VREF minus 5 kt should be done at various weights
and CGs to confirm. The AC states, for example, that if the maneuver is “nor-
mally” executed, one should expect a touchdown airspeed about 5 kt less
than normal. One should evaluate elevator authority and the capability to
touch down at no more than 6 ft/s while using the normal landing procedure.

Air Distance Testing and Data Reduction: Acceptable Means of

Compliance
The guidance material of FAA AC 25-7D recommends three methods for
testing and data reduction; these are acceptable means of compliance to
collect flight test data and create a model for air distance. This author rec-
ommends a fourth one. These models are:
• Mathematical upper bound
• Six performance landings
• Parametric landing
• Airtime modeling
 944
RADALT (ft) Pitch (deg)
50 6

Operational Aircraft Performance and Flight Test Practices

40 4.068795 4

RADALT (ft)

Pitch (deg)
30 10.1183 4.784618 2
20 0
10 –0.1233883 –2
0 –4
L MLG (kt) R MLG (kt)
150
24.6201
R MLG (kt)
L MLG (kt)

100
0.105672
50 0.215995 0.105672

0
MLG WoW left (0/1) MLG WoW right (0/1)
1.5
MLG WoW right
MLG WoW left

1.0
0
0.5 0 0 0
0.0
CG Nz (g) Cockpit Nz (g)
1.0
Cockpit Nz (g)

0.0669076 0.0677793 –0.09994004 0.02562412

0.5
CG Nz (g)

0.0
–0.5
–1.0
RoC (fpm) (FPA) (deg)
500 0

(FPA) (deg)
RoC (fpm)

0 –2
–500 –1.078443 –4
–528 –0.5929782 –160
–1000 –6
94 2

36 n
45 e

49 w
55 60 65 70 75
.5 lin

.8 do
61 ent

67 ch
Time (s)
u
Ev

To

Fig. 20.8 Potential signals that can be used to determine touchdown point.
 CHAPTER 20 Landing Performance Testing and AFM 945

Mathematical Upper Bound

The first method defined by the FAA advisory circular is an upper bound
zero-wind airborne distance (in feet) developed from several Part 25 certifi-
cations. The resulting mathematical model, using VREF in KTAS, is

Xa ¼ 1:55 ðVREF  80Þ1:35 þ 800 (20:1)

This equation was developed from the testing of larger Part 25 airplanes.
(The equation does not provide any results below an approach airspeed of 80
KTAS, for example.) Plotting the equation produces Fig. 20.9. Note how this
compares to the fixed distance of 955 ft defined by Eq. (19.1) for a straight
approach at 23 deg from 50 ft to touchdown (no flare), reflecting some of
the impact of the higher approach airspeed and the need to flare to lower
the sink rate prior to touchdown.
One should typically expect the VREF to be between 100 and 145 KCAS for
Part 25 airplanes. Under standard day sea level conditions, one would then
expect, from this method, landing distances between 890 and 1235 ft. One
then also needs to remember that Part 25 airplanes operate over a wide
range of altitudes and temperatures. So 100 KCAS would translate to about
90 KTAS under sea level pressure altitude and 2408C (cold day), and the
145 KCAS would become about 179 KTAS at 10,000 ft pressure altitude and
308C (358C above standard day). The landing air distance would then stretch
from about 835 ft to 1575 ft over this range of altitudes and true airspeeds.
The method also suggests a minimum true airspeed loss of 3 kt at touch-
down from 50 ft that is then used to compute the rest of the landing distance.

2000
1800

1600
1400

1200
Xa (ft)

1000
Equation (19.1)
800
600

400

200

0
80 100 120 140 160 180 200
VREF (KTAS)

Fig. 20.9 Mathematical upper bound model.

946 Operational Aircraft Performance and Flight Test Practices

The guidance material suggests that “an applicant may elect to use the above
relationships instead of measuring the air distance and speed loss.” So, at first
glance, this may seem like an interesting approach to reduce the amount of
testing (schedule cost). The guidance material does, however, continue to say
that “it must be shown by test or analysis that these equations do not provide
non-conservative results.”

Six Performance Landings

The previous method usually results in a conservative airborne distance.
The applicant may decide that it is too penalizing for the performance of the
airplane to be certified. The applicant will often elect to measure the airborne
distance. In such a case, AC 25-7D recommends that at least six landings (per
landing configuration to be certified) be done covering the weight range of
the airplane. (The concept of at least six test conditions reappears often in
our discussion as a minimum acceptable level of demonstration.) The
testing should meet the following criteria:
• The airplane should be on a stabilized approach at 23 deg (average of all
landings should not exceed this value) of flight path and at VREF for a
reasonable amount of time prior to reaching the screen height (suggest
from 200 ft AGL).
• The flight path must not be deliberately increased below 50 ft AGL.
• The target (average) sink rate at touchdown must not exceed 6 ft/s.
The test results in terms of altitude vs distance are similar to those shown in
Fig. 20.10. If the execution of the landing procedure does not result in a clear,

60

50
Height above ground (ft)

40
Path for performance Path for normal
landing at 6 ft/s touchdown rate
30

20

10

0
–200 0 200 400 600 800 1000 1200 1400
Distance from touchdown (ft)

Fig. 20.10 Performance landing vs normal landing.

 CHAPTER 20 Landing Performance Testing and AFM 947

distinct two-path approach (such as shown in Fig. 20.10), then the air dis-
tance model could be shown to simply be an average flight path angle
from 50 ft to touchdown. If instead, the procedure results in a two-path con-
dition, one can create a model similar to that shown in Example 19.5 in
Chapter 19. The speed decay becomes the average of the six landings.
These tests are often all done in a row (single flight), and thus are at a
single altitude. Because the model becomes an average flight path angle to
the ground, altitude will not be a major impact.
The main difference between a normal landing and a performance-type
landing, in terms of air segment, is the reduction in the amount of flare
the pilot uses. Both approaches still require having 23 deg at screen
height. Figure 20.10 is an example of the distance covered by a normal
landing as well as one for a performance landing with minimal flare (sink
rate at touchdown of 6 ft/s).
Compared to the distance from the screen height and touchdown of
the normal landing, the performance landing reduced the distance by
approximately 30 ft (3%) in this example. The main risk associated with
this type of landing is that the pilot could be late to flare and exceed the
maximum design value at touchdown. Such was the case for the crew of a
McDonnell Douglas DC-9-80 during dedicated testing in 1980 (see NTSB
LAX80FA092 accident box).

NTSB Identification: LAX80FA092, DC-9-80, 2 May 1980 crash at

Edwards AFB, N980DC: The pilot failed to flare during an approach to
land. The aircraft was on a certification test flight to determine the horizontal
distance required to land and bring the aircraft to a full stop as required by 14
CFR 25.125. The descent rate at touchdown exceeded the aircraft’s structural
limitations (estimated to have been a 16.2 ft/s touchdown). The empennage
separated from the airplane just aft of the aft fuselage – mounted engines
and fell to the runway. You can see a video of the landing here: https://
youtu.be/COsT6DqkTDc.
Source: https://www.ntsb.gov/investigations/AccidentReports/Reports/
AAR8202.pdf
948 Operational Aircraft Performance and Flight Test Practices

This author often has discussions with authorities on why the landing test
is not performed more like an operational landing—like what the “average”
pilot would do. The answer is simple: Testing for minimum landing distance
provides the most consistent results and, irrespective of how the landing dis-
tance is achieved, the same conservative operational margins will be applied
to the actual landing distance (ALD). There is no benefit to adding margins in
the basic ALD model. The model should be representative of the best, repea-
table airplane landing distance.

Maximum Structural Design Sink Rate

For most touchdowns, the sink rate is not completely reduced, but
rather a small remaining value exists. This small residual sink rate is absorbed
by the landing gear shock struts. Typical operational touchdown sink rate
varies from less than 1 ft/s to approximately 4 ft/s; anything higher is
usually uncomfortable for the passengers. Part of the flight testing should
verify the landing gear characteristics:
§ 25.125 (a)(4) The landing must be made without excessive vertical accel-
eration, tendency to bounce, nose over, ground loop, porpoise, or water loop

Of course, higher sink rates are typically expected for the performance land-
ings (minimize flare). The landing gear must be designed to absorb the fol-
lowing loads as defined in §25.473:
§ 25.473 Landing load conditions and assumptions
(a) For the landing conditions specified in §25.479 to §25.485 the airplane is
assumed to contact the ground—
(1) In the attitudes defined in §25.479 and §25.481;
(2) With a limit descent velocity of 10 fps at the design landing weight
(the maximum weight for landing conditions at maximum descent
velocity); and
(3) With a limit descent velocity of 6 fps at the design take-off weight
(the maximum weight for landing conditions at a reduced descent
velocity).

The 10 ft/s represents the limit load condition for the airplane. The load is
proportional to the sink rate squared, so the ultimate load condition (150%
of limit load) is at a sink rate of 12.25 ft/s. Some minor deformation can
occur between limit load and ultimate load, requiring a hard landing inspec-
tion. Beyond ultimate load, one should expect some permanent damage to
and/or failure of the airplane.

Parametric Landings
If the applicant is willing to perform even more testing, then one may
elect to adopt the parametric landing analysis for the air segment. The para-
metric analysis consists of developing a statistically based model of the air
distance and airtime required from 50 ft AGL to touchdown based on the air-
plane sink rates at the screen height and at touchdown. The method was
 CHAPTER 20 Landing Performance Testing and AFM 949

developed following the DC-9-80 accident (discussed in the accident box)

during the dedicated landing performance testing as a means to reduce the
test risks by reducing the maximum target sink rate at touchdown (forcing
the applicant to not delay flare more than acceptable).
The guidance material of FAA AC 25-7D proposes the following test
limits for dedicated parametric landings:
• The approach flight path should be between 22.5 deg and 23.5 deg.
• The sink rate at touchdown should be within 2 to 6 ft/s, thereby providing
a margin to the maximum design sink rate of 10 ft/s.
• The target speed for the test is VREF , with an acceptable deviation of
approximately +2 kt.
• The flight technique used to perform the parametric landings should be
consistent with the procedure to be used in the AFM.
To define an acceptable model, AC 25-7D suggests that a minimum of 12
landings are required. In reality, the applicant should expect significantly
more testing, usually between 40 and 50 landings (far more than the six per-
formance landings) to acquire sufficient data that remain within the tolerance
listed previously and to build an acceptable model.
Because the parametric landing is a statistical method based on sink rate,
it has been this author’s experience that expanding the flight path tolerances
to fall between 22 deg and 24 deg will still provide an acceptable model with
possibly a better fit for the AFM extrapolation. To gather good data (stable
approach path in particular), the applicant should test in calm air (low turbu-
lence) to decrease the scatter in flight path angle, airspeed, and touchdown rates.
The advantage of this method over the six performance landings is that,
once a good model has been developed, it allows the applicant to use a
23.5-deg approach and 8-ft/s touchdown rate for AFM expansion, even if
the testing methodology suggests that the sink rate at touchdown should
not exceed 6 ft/s during dedicated testing. This allowance was put in place
based on the fact that landing models used to be developed with the six per-
formance landings at the high end of the flight path tolerance band and
minimal flare, and those models have shown acceptable operational safety
when combined with the landing field length (LFL) factor (§121.195) of stop-
ping the airplane within 60% of the runway.
The parametric model developed will take the following form:
50
¼ at þ bt (R=S)50 þ ct (R=S)TD (20:2)
t
V50
¼ aV þ bV (R=S)50 þ cV (R=S)TD (20:3)
VTD
The first equation provides the model for the airtime (50 ft divided by
time from 50 ft to touchdown); the second equation represents the speed
decay of the airplane. All speeds/sink rates are true airspeed in feet per
950 Operational Aircraft Performance and Flight Test Practices

second, and the coefficients are determined using the following equations
(for the speed ratio, replace the 50/t with V50 /VTD ):

[R6  b R1  c R3 ]
a¼ (20:4)
n

[R12  c R11 ]
b¼ (20:5)
R9

[R9 R10  R11 R12 ]

c¼ (20:6)
[R9 R13  R211 ]

The R coefficients making the previous coefficients are

Xn
R1 ¼ 1
(R=S)50 (20:7)

Xn
R2 ¼ 1
[(R=S)50 ]2 (20:8)

Xn
R3 ¼ 1
(R=S)TD (20:9)

Xn
R4 ¼ 1
[(R=S)TD ]2 (20:10)

Xn
R5 ¼ 1
(R=S)50 (R=S)TD (20:11)

Xn 50
R6 ¼ 1
(20:12)
t

Xn
R7 ¼ 1
(R=S)50 (50=t) (20:13)

Xn
R8 ¼ 1
(R=S)TD (50=t) (20:14)

R9 ¼ n R2  R21 (20:15)

R10 ¼ n R8  R3 R6 (20:16)

R11 ¼ n R5  R1 R3 (20:17)
 CHAPTER 20 Landing Performance Testing and AFM 951

R12 ¼ n R7  R1 R6 (20:18)

R13 ¼ n R4  R23 (20:19)

There are a lot of equations to account for in a parametric study, but they
can easily be done with a spreadsheet. Because this is a statistical approach,
feeding good quality data into the model is essential. The focus is then on the
execution of the testing. The best way to get a feel for this approach is to
review an example of test data analysis.

Example 20.1
An airplane executes some performance landing tests, and the following data
are collected. Determine the parametric landing model.

Run R/S50 R/STD V50 VTD t Distance

1 6.72 4.05 205.34 194.52 8.3 1659
2 6.28 4.13 209.74 182.24 9.6 1882
3 5.49 3.80 206.97 193.78 8.9 1783
4 6.67 3.39 200.09 187.83 8.3 1610
5 7.58 4.19 201.29 189.13 7.2 1405
6 8.45 1.62 196.30 184.38 7.1 1351
7 6.45 3.36 194.63 182.86 7.1 1340
8 9.33 4.47 194.55 181.11 6.6 1240

Solution: The sink rate and airspeeds are provided in feet per second of true
airspeed, the distance is in feet, and the airtime is in seconds. Analyzing the
data for airtime, one gets the following coefficients:

R1 101.79 R8 332.37
R2 930.04 R9 798.42
R3 45.94 R10 35.62
R4 183.75 R11 78.67
R5 396.27 R12 386.69 c 20.02868
R6 86.04 R13 94.38 b 0.48715
R7 762.09 a 3.14741

This gives the following airtime model:

50
¼ 3:14741 þ 0:48715 (R=S)50  0:02868 (R=S)TD
t

(Continued)
952 Operational Aircraft Performance and Flight Test Practices

Example 20.1 (Continued)

And the following speed ratio model:

R1 101.79 R8 49.85
R2 930.04 R9 798.4178
R3 45.94 R10 0.279189
R4 183.75 R11 78.67337
R5 396.27 R12 0.600527 c 0.00254
R6 13.01 R13 94.37754 b 0.000502
R7 110.45 a 1.070566

V50
¼ 1:070566 þ 0:000502 (R=S)50 þ 0:00254 (R=S)TD
VTD
The distance is then computed by finding the average airspeed from 50 ft
to touchdown and multiplying by the airtime. So from the planned VREF , one
computes the VTD from the V50 /VTD equation, and the air distance takes on
the following form:
(VREF þ VTD )
Xa ¼ t (20:20)
2
where the airtime t is computed from Eq. (20.2). When one compares the test
data from this example to the model data, we get the following results (feeding
in the actual test data):

Test Model
Run Distance V50 /VTD t Distance V50 /VTD t
1 1659 1.0556 8.3 1565 1.0842 7.9
2 1882 1.1509 9.6 1656 1.0842 8.2
3 1783 1.0681 8.9 1742 1.0830 8.8
4 1610 1.0653 8.3 1528 1.0825 7.9
5 1405 1.0643 7.2 1439 1.0850 7.4
6 1351 1.0646 7.1 1310 1.0789 6.9
7 1340 1.0644 7.1 1512 1.0823 8.1
8 1240 1.0742 6.6 1235 1.0866 6.6

The data are best viewed in chart format (see Fig. 20.11).
Overall, the model fit is very reasonable if the quality of the data was good
(within the test tolerance). The outlier points can be reviewed for acceptability

(Continued)
 CHAPTER 20 Landing Performance Testing and AFM 953

Example 20.1 (Continued)

of execution of the landing procedure by the crew and the weather conditions
at the time of the test.

a) b)
2000 10.0
1900 9.5
1800
9.0
Model distance (ft)

1700 ±200 ft ±1 s

Model time (s)

8.5
1600
1500 8.0
1400 7.5
1300
7.0
1200
1100 6.5

1000 6.0
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Test distance (ft) Test time (s)

Fig. 20.11 Test results and model matching for Example 20.1.

Airtime Modeling (A Flight Path Reconstruction Model)

When I teach my class on landing performance testing, I often joke about
the “complexity” of the parametric landing model and show the image in
Fig. 20.12. The parametric approach fits data into a series of equations to

Fig. 20.12 The magic of parametric landing analysis.

954 Operational Aircraft Performance and Flight Test Practices

best represent the testing results. The problem I have seen over the years is
that people then turn around and use the same model for the next airplane/
design they may be working on saying “it is the best available data” (read
BAD). It is not, of course, and not having a physical sense, you cannot
adjust the parametric model for expected airplane behavior; it is a monster
you have to feed with more data.
What if the parametric modeling is not the best way to produce a landing air
distance model? When the data from Example 20.1 are plotted in terms of dis-
tance vs rate of sink at touchdown, one can hardly see a trend, and the one seen
in this case would indicate an increasing air distance vs increasing sink rate
(counterintuitive); see Fig. 20.13. We know there is direct correlation among
distance, ground speed, and airtime, and this is clearly seen in Fig. 20.13.

2000
Test data × Model Linear (Test data)

1800
×
×
1600 ×
Air distance (ft)

××
×
1400
×
×
1200

1000

800
0 1 2 3 4 5 6
R/STD (ft/s)
12.0
Test data × Model Linear (Test data)

10.0
×
8.0 ×× × ×
×
Airtime (s)

×
×
6.0

4.0

2.0

0.0
600 800 1000 1200 1400 1600 1800 2000
Air distance (ft)

Fig. 20.13 Plot of data from Example 20.1 for trends.

 CHAPTER 20 Landing Performance Testing and AFM 955

The problem with the parametric analysis is that it looks at only the entry
condition (airspeed and sink rate) and the end condition (touchdown) but
not at the actual flight path in-between (see Fig. 20.14). Two landings
could be done with the same exact entry and end conditions but have
greatly different air distances Xa . Unless the performance engineer looks
into the details of the flight path time history, he or she will not get a full
picture and will require significantly more flight test data to make the para-
metric model converge. The parametric method is meant to work well with a
single curved path (per normal landing) where the approach is stable to the
flare point and then a single movement flare is done. It does not review the
entire flight path and the possibility that there was an extended flare before
the same touchdown rate.
The proposed airtime method will be one that is adapted to the landing
procedure to be followed by the crew, that has a physical sense, and therefore
that can be used to extrapolate to given landing conditions similar to the
landings of the parametric analysis. The method is developed using the
material presented in Chapter 19.
The method will consist of dividing the air distance into two segments, air
distance 1 (preflare) and air distance 2 (flare)
Air distance 1 (Xa1 ) – 50 ft to flare height, constant flight path angle
(FPA):
• For data reduction, capture the distance along the runway at 50 ft AGL and
at the start of the flare (must be a distinct pilot command to pull the nose
up, beyond the small FPA adjustments) as well as the height of the airplane
at the start of flare hflare .
• If the procedure calls for a given height, say 30 ft AGL, then expect that
with pilot reaction time and sink rate, the actual start of flare can occur
between 25 ft AGL and 15 ft AGL.
• The average FPA for the segment is simply
 
1 hflare  50 ft
ga ¼ tan (20:21)
Xflare  X50

VREF VREF

50 ft Single Double
50 ft
screen curved path curved path
screen
height 6 ft/s TD 6 ft/s TD
height

Xa Xa

Fig. 20.14 One reason for noncorrelation of parametric landing model and sink rate
at touchdown.
956 Operational Aircraft Performance and Flight Test Practices

• If the throttle is chopped to idle prior to reaching the flare height, its
impact on the flight path angle is captured in Eq. (20.21).
• The target FPA at 50 ft is the nominal approach value (typically 23 deg)
with a tolerance of +0.5 deg. Note that values outside of this should not
necessarily be rejected without investigating the airplane’s behavior and
crew execution of the landing procedure.
• For data expansion, similar to parametric analysis, the performance
engineer will be allowed to use the nominal approach angle minus 0.5 deg
(steeper, so 23.5 deg for a nominal approach angle of 23 deg).
• As we saw in Fig. 19.6 in Chapter 19, this part of the air segment may be
impacted by both the ground effects and the impact of thrust reduction.
• Equation (19.20) will be used to compute the nominal horizontal
distance for data expansion.
• The sink rate at 50 ft will be per Eq. (19.2).
• The flare height for the expansion is as discussed in the following
discussion of air distance 2.
• Note that an engine chop to idle at higher altitudes may actually result in
a more negative FPA than the nominal value minus 0.5 deg. We do not
believe the small difference should be added on top of this tolerance.
Air distance 2 (Xa2 ): This segment of the air distance is highly dependent on
the crew landing procedure, which also makes this method more representa-
tive of the actual airplane behavior.
• For data reduction:
• For the condition where flare height is recommended per AFM
procedure, one should expect an increasing load factor with increasing
approach sink rate (heavy weight, high-density altitudes).
• Pilot reaction time from actual flare height to control column input is
typically of the order of 0.5 s, so pilot input will be lower than
procedure by the equivalent of 0.5 s times the sink rate at flare height.
During the test, actual data are used (approximately 0.5 s in Fig. 19.17
for the 30-ft AGL flare height of the procedure), and an average for all
test landings is determined.
• The “average” load factor for a given landing is simply derived from
the demonstrated air distance from start of flare to touchdown.
• Collect flare height for test point (measured value).
• Collect air distance from flare height to touchdown (measured
value).
• Measure winds.
• You now have two known test values (flare height and distance) and
two unknowns, curved path radius r [Eq. (19.13)] and flight path at
touchdown gTD . One can use Eqs. (19.16) and (19.17) to solve for
the unknowns.
• Equations (19.16) and (19.17) assume an equivalent single curved
path. If the test flight path is a single curved path per Fig. 19.22,
 CHAPTER 20 Landing Performance Testing and AFM 957

then the measured sink rate at touchdown would give a first

estimate of the flight path at touchdown. If the actual path is
elongated or double path, then solving the equations will lead to
an equivalent touchdown flight path.
• From the derived curve path radius, one can then compute the
average load factor for the maneuver [from Eq. (9.13)].
• For conditions where flare height is left to the pilot’s decision, one
should expect the height to increase with increasing sink rate on
approach and that the pilot will apply essentially a constant load factor in
the flare.
• Then, the data reduction will evaluate the average demonstrated flare
load factors.
• The method relies on a single curved path maneuver or in calculating
an equivalent.
• The maximum expected load factor is typically 1.3 g (based on the
maneuver margin requirements of VREF ). One should expect that a
repeatable maximum value would be lower than this peak value
because stall warning in the flare is an unacceptable characteristic.
• Thrust cut height will dictate the speed reduction, which will essentially
be proportional to the airtime from thrust cut height to touchdown.
• One can compute the average deceleration rate computed as the true
airspeed decay divided by airtime to touchdown from cut height.
• This will give a good value if the thrust cut height is performed in a
repeatable manner. If the thrust cut height is left at the discretion of
the crew and is different on every landing, this will not produce a
useful value.
• For data expansion:
• For the procedure with fixed flare height:
• Compute the flare height based on average pilot reaction time from
landing procedure flare height.
• This flare height is used to compute the air distance 1.
• Assume a sink rate at touchdown of 6 to 8 ft/s. (This will give the
equivalent flight path at touchdown, from which one can derive the
curve path radius.)
• Compute the flare distance (air distance 2).
• For the procedure with a fixed average load factor:
• Assume a single curved path with an average load factor. (You can
used highest achieved during testing.)
• Remember that VREF typically offers only 1.3-g capability to stall
warning, so a peak of about 1.2 g.
• Assume a target sink rate at touchdown of 6 to 8 ft/s. This gives the
flight path at touchdown.
• Compute a flare radius and then a flare height.
• This flare height is used to compute the air distance 1.
• Compute the flare distance (air distance 2).
958 Operational Aircraft Performance and Flight Test Practices

• Speed decay during air distance

• An average of the speed decay can be used for expansion, or a
relationship with thrust decay airtime can be established to compute
the airspeed at touchdown.
• This new speed becomes the starting point of the transition segment.
For the airtime method, it is suggested that 6 to 10 landings be demonstrated
using the proposed certification landing procedures and airspeed reasonably
covering the expected weight range of the airplane. The target sink rate on
touchdown should be similar to that of the parametric testing requirements,
namely 2 to 6 ft/s. (It is a performance landing, after all.) Great efforts should
be made to obtain a single curved path flare, which will reduce scatter in the
data reduction, but method can account for a deviation from it.
The total air distance is the sum of two segments, similar to Eq. (19.21).
By observation, this method can now be directly linked to physical aspects of
the maneuver; if the landing procedure is changed, it can be easily adjusted.
One can use the material of Chapter 19 to evaluate the initial targets for
thrust cut height and flare height and run simulations for expected behavior
prior to starting the landing technique development so as to minimize flight
test time.

Recap: Air Distance Instrumentation Required

We reviewed the essential parameters to monitor early in our discussion
of the air segment. We complete the review of the instrumentation here. The
minimum parameters that need to be measured for a performance landing
are as follows:
• Altitude, both barometric and above ground level [radar altimeter
(RADALT) or DGPS] and a means to compute instantaneous sink rate.
• Flight path (ideally, this should be computed in real time and fed back to
the pilot to help reduce the scatter and the number of repeat points).
• Airspeed.
• Knots calibrated airspeed (KCAS) [knots indicated airspeed (KIAS) with
best corrections]
• Knots true airspeed (KTAS)
• Knots true ground speed (KTGS)
• Winds (dedicated experimental weather station with continuously
recorded wind speed, direction, and height above ground), pressure,
and temperature.
• Flight controls and speed brake positions.
• Engine speed and throttle angle position.
• Wheel speed (mostly to determine touchdown point, but also to monitor
wheel skids) and temperature, as well as brake pressure (to determine start
of full braking). Can also include WoW discrete and gear strut pressure.
• Aircraft position (DGPS or INS).
 CHAPTER 20 Landing Performance Testing and AFM 959

• Vertical and longitudinal accelerations.

• Angles of attack and sideslip, pitch and roll attitudes.
• Weight and CG.

Air Distance: Test Risks and Mitigation

Performance landing tests always increase the risk over that of normal
landings. The main reasons for this increased risk include:
• Higher target sink rates at touchdown (2 to 6 ft/s are usually targeted),
which may lead to the actual landing sink rate exceeding the design value of
10 ft/s.
• High derotation rates to minimize time and distance, leading to excessive
nose gear sink rates.
• Maximum braking is desired to get the shortest landing distance possible.
Some of the risk items to be expected during performance landings are:
• Failure of the wheels, tires, and/or brakes
• Structural damage to the airplane
• Loss of directional control [especially during dedicated one engine
inoperative (OEI) or crosswind testing] with subsequent departure from
the runway

Several steps can be taken to mitigate the risks identified:

• Monitor the airplane sink rate.
• Inspect the landing gear assembly and attachment points on a regular basis.
• Monitor the temperature of the wheels and brakes, and review brake
energies prior to test.
• Ensure that the runway to be used has sufficient length (best landing
distance estimates plus margin for the test being performed) and width.
• Make sure ground equipment are a safe distance away from the runway.
• Have other means of deceleration available, such as thrust reversers (if
installed) or spin-chute (if installed).
• Ensure the crew wears protective equipment (helmet, flying suit, and
gloves).

If a dedicated OEI landing is to be performed, the crew should be familiar

with OEI handling characteristics and engine relight procedures.

Testing for the Transition Segment

The transition segment extends from the touchdown to the point where
the nose gear is on the ground and all deceleration devices are extended. The
only devices that an applicant can take credit for performance are as specified
in §25.125.
960 Operational Aircraft Performance and Flight Test Practices

§25.125 Landing
(c) For landplanes and amphibians, the landing distance on land must be
determined on a level, smooth, dry, hard-surfaced runway. In addition—
(1) The pressures on the wheel braking systems may not exceed those
specified by the brake manufacturer;
(2) The brakes may not be used so as to cause excessive wear of brakes
or tires; and
(3) Means other than wheel brakes may be used if that means—
(i) Is safe and reliable;
(ii) Is used so that consistent results can be expected in service; and
(iii) Is such that exceptional skill is not required to control
the airplane.

The part about “safe and reliable” is clearly exemplified in the accident
box “FAA Airworthiness Directive (AD) 2003-07-09, Raytheon Aircraft
Company Model 390 Airplanes” later in this chapter. A small business jet
had several events where the spoilers did not deploy while the airplane was
landing on a short runway. The runway distance with the spoilers inoperative
was longer than the one with the spoilers operative (the baseline published
landing distance). The manufacturer eventually made some changes to
remedy the situation.
The transition segment, for the purpose of performance modeling of the
distance, includes:
• Time to derotate
• Proportional to pitch at touchdown and fuselage length so as not to
exceed 8 ft/s (avoid damage, Fig. 20.16), thus max pitch rate
• Some braking prior to the end of the derotation if §25.125(c)(3)(iii) applies
• Time to select deceleration devices and have them deploy and
become effective
This segment is relatively short and fairly repeatable in execution, so it is
often best to simply do a series of landings and average out the transition
time and the average deceleration during this time. The authorities have gen-
erally accepted the approach that a minimum of six performance landing
transition segments be demonstrated.
The time required for derotation will be a significant part of the transition
time, especially for large airplanes and/or airplanes with a high pitch attitude
on touchdown (see Fig. 20.15). Just like the main landing gear has a design

110 ft
25 ft

5 deg 5 deg

Fig. 20.15 Fuselage length impact on derotation time.

 CHAPTER 20 Landing Performance Testing and AFM 961

Fig. 20.16 Example of forward fuselage damage.

maximum sink rate at touchdown, so does the nose landing gear. AC 25-7D
specifies that the nose gear touchdown rate shall not exceed 8 ft/s for
certification testing.
In Fig. 20.15, both airplanes have a pitch attitude of 5 deg on touchdown,
with the short airplane having its nose gear wheel 2.2 ft off the ground and
the long airplane having its nose gear 9.6 ft off the ground. If both adopt a
nose-down pitch rate of 2 deg/s, the short airplane will have a nose
landing gear (NLG) touchdown rate of less than 1 ft/s and take 2.5 s to
derotate. The long airplane will also take 2.5 s to derotate, but would have
an NLG touchdown rate of about 4 ft/s. There is room to improve the dero-
tation rate of the short airplane, but not much room for the larger plane
without the risk of exceeding the structural limits. Figure 20.16 shows
examples of forward fuselage damage for airplanes that had heavy nose
gear touchdown rates.
We provide Fig. 20.17 as an example of a performance landing transition
segment. The airplane touched down at time 19.8 s on this chart. The ground
spoilers deployed automatically after weight on wheels was sensed and were
fully deployed about 1.1 s after touchdown. Note that the pilot initially mod-
erates the nose-down pitching motion, even pulling a little on the elevator
initially before pushing the nose down (between time 20.9 and 21.5 s) to
accelerate the derotation before pulling on the stick again to ensure no
hard nose-down touchdown (time 21.9 s). Once the NLG is on the ground,
the pilot applies maximum braking, and the brakes become fully effective
(as seen by the longitudinal acceleration Nx becoming stable) by time
 962
Touchdown: 19.8 s
RADALT (ft) KTAS (kt) End transition: 23.3 s

Operational Aircraft Performance and Flight Test Practices

200 150

RADALT (ft)
150

KTAS (kt)
118.5 100
100
117.1 113.3 105.1 50
50
0 0
Nz (g) Elevator (deg)
1.5 20

Elevator (deg)
1.0 10
Nz (g)

0.5
0
0.0
–0.5 –10
–1.0 –20
Pitch (deg) Pitch rate (deg/s)
Pitch = 7 deg
10
Pitch rate (deg/s)
Pitch (deg)

5
0
–5
–10
Nx (g) Ground spoilers (deg) Full braking segment

Ground spoilers
0.0
50
Nx (g)

–0.5

(deg)
–1.0 0
16 18 20 22 24 26 28 30
Ground spoilers deployed: 20.9 s Time (s) NLG touchdown: 21.9 s

Fig. 20.17 Transition segment sequence.

 CHAPTER 20 Landing Performance Testing and AFM 963

23.3 s. The total transition time here was 3.5 s. The airspeed loss during that
segment was 13.4 kt for an average deceleration of 20.2 g.
The average deceleration could be improved if the pilot (or the airplane
system) applied some brake force prior to nose gear touchdown, but this
would add some additional nose-down pitching moment that could lead to
an excessive NLG touchdown rate. Either the pilot needs to carefully modu-
late the brakes or the airplane can be designed with an automatic system that
provides a maximum braking torque while the nose gear is still in the air. We
must also remember that during the early part of the derotation, the airplane
is still at a higher angle of attack, thus producing a great deal of lift, and does
not have much weight on wheels. This reduced weight on wheels may not
provide much braking force and may also lead to wheel skidding.
Figure 20.17 shows an example of automatic spoiler deployment; in con-
trast, Fig. 19.25 in Chapter 19 showed an example of pilot-selected spoilers
during a transition phase. Note that the automatic system tends to deploy
sooner (upon MLG touchdown), but the deployment rate is typically the
same for pilot-activated vs automatic deployment. (The difference in time
can be anywhere from about 1 to 3 s). The spoilers do provide more drag
in the early part of the braking segment, but the main benefit is in killing
the lift on the wing, thus increasing the weight on wheels and thus increasing
the braking capability. This author’s perspective is that as long as the spoilers
are deployed by the time the maximum brake force is applied, the perform-
ance will be best.

FAA Airworthiness Directive (AD) 2003-07-09, Raytheon Aircraft

Company Model 390 Airplanes: This AD is the result of two accidents on
the affected airplanes where a contributing factor was the lift dump spoilers
failing to deploy when commanded after the initial landing. The actions speci-
fied by this AD are intended to require the use of necessary flight information
to prevent runway overruns based on insufficient aerodynamic and wheel
braking if the lift dump spoilers do not operate after landing touchdown.
This could result in reduced or loss of control of the airplane.
What Events Have Caused This AD? The FAA has received information of
an unsafe condition on Raytheon Model 390 airplanes. The current procedure
for an annunciated lift dump failure is to increase landing distance by a factor
of 1.53. In two recent accidents of these airplanes, the lift dump spoilers failed
to deploy when commanded after touchdown. Whether loss of lift dump is
annunciated or unannunciated after touchdown, the pilot (in most instances)
does not have enough time to take effective corrective action.
What Are the Consequences If the Condition Is Not Corrected? Without
requiring the use of necessary flight information, runway overruns based on
insufficient aerodynamic and wheel braking could occur if the lift dump spoi-
lers do not operate after landing touchdown. This could result in reduced or
loss of control of the airplane.
964 Operational Aircraft Performance and Flight Test Practices

Credit for thrust reversers cannot be used on dry runways for the purpose
of computing a shorter landing distance. One should still collect information
on the thrust reverser capability to decelerate the airplane (create a retarding
force model) so as to create a model for use on wet and contaminated
runways. For the transition segment, we collect the time from touchdown
until the reverse thrust is brought up to full reverse.
Propeller-driven airplanes can use propeller discing (ultrafine pitch
setting) as part of the normal landing procedure if this meets the require-
ments of §25.125. Discing puts the propeller into a large drag setting, not
in reverse mode.

Modeling Transition
Once the sequence of events has been demonstrated to be repeatable
during certification flight testing (the famous six performance landings),
one can model the ground distance by the average deceleration and the
demonstrated average time. The average time must account for pilot
actions and may need to be extended somewhat. When expanding the data
for the transition segment, AC 25-7D specifies the following minimum times:
• From touchdown to the activation of the first deceleration device (usually
the brakes), the minimum time should be the average of the demonstrated
times, but should not be less than 1 s.
• For any subsequent devices, use a minimum of 1 s or the average
demonstrated time in successive order.
• For approved automatic devices (e.g., speed brakes), the established
average time demonstrated during testing can be used in lieu of the
minimum of 1 s.
For example, the spoilers could be automatically deploying 1.5 s after touch-
down and the pilot could have applied full braking 1.3 s after touchdown. For
AFM expansion, the first delay time would be 1.3 s for pilot brakes, but the
next delay time would only be 0.2 s for full spoilers instead of 1 s.

Testing for the Ground Run Segment

The ground run segment will start at the time all deceleration devices are
extended (thrust reversers could be the exception) and brakes are applied to
the point of a complete stop. (The complete stop point is required for the
purpose of computing a distance to meet the requirements of §25.125.)
During this segment, the airplane configuration is stable including the
pitch attitude [thus angle of attack (AoA)]. This becomes a braking phase
similar to what was described in Chapters 17 and 18 with the exception
that the flaps are now in the landing configuration and possibly providing
more lift. During the six performance landings, the performance engineer
can extract the deceleration capability of the airplane with this additional
 CHAPTER 20 Landing Performance Testing and AFM 965

lift and compare it with the basic braking model developed for the rejected
takeoff (RTO).

Wheels and Brakes

Per the takeoff testing, the tires must be set to the maximum pressure
desired for certification. This provides for the lowest ground friction coeffi-
cient (conservative). Testing should be done with the normal expected
brake pressure in service.
Part 25.125(b)(2) requires that “The brakes may not be used so as to cause
excessive wear of brakes or tires;”
The guidance material of AC 25-7D requires that at least six measured
landings, covering the entire weight range to be certified, be conducted on
the same set of wheels, tires, and brakes so as to substantiate that the wear
is not excessive. Note that the introduction of antiskid systems has greatly
helped with this aspect. Prior to the introduction of antiskid systems, the
pilots would have to modulate the braking force, which would lead to
more scatter in the braking model and lower braking coefficients.

Total Landing Distance

The total landing distance, the actual landing distance (ALD), is the
as-demonstrated sum of all three segments. The performance engineer
must compare the models created with the actual test distance and verify
that the fit is reasonable (see Fig. 20.18). This serves as validation of the
model when the data are presented to the authorities. This is done by

2500
+200 ft

2000 –200 ft
Model distance (ft)

1500

1000

500

0
0 500 1000 1500 2000 2500
Actual distance (ft)

Fig. 20.18 Comparing the actual test distance to the performance model.
966 Operational Aircraft Performance and Flight Test Practices

feeding in actual test day conditions (winds, temperature, etc.) before the
addition of minimum expansion margins.

Coming in Fast
The landing distance published in the AFM is based on an approved
procedure, including the use of specific approach speeds, VREF . Coming in
for a landing at a speed above VREF will increase the landing distance:
• The air distance from 50 ft to touchdown will increase because of the
higher airspeed. The airplane may actually have an increased tendency to
float prior to touching down.
• The derotation and braking distances will also increase because of the
higher speed.
• The increase in actual landing distance can easily be as much as 50 ft per kt
over VREF if the crew follows the same procedure as for a normal landing (at
the appropriate speed), more if the airplane floats during the flare.
Coming in faster than VREF will also mean that more energy will be trans-
ferred to the brakes during the braking phase. If the airplane lands at an
energy level close to the published maximum landing energy weight, and
no other devices than brakes are used to slow down, the brakes may
absorb more energy than the demonstrated maximum energy of the fusible
plugs (if plugs are installed).
The fusible plug is simply a threaded metal cylinder with a special low
melting point core plug (see Fig. 20.19). As the brakes and wheels absorb
energy during the braking part of the landing, the temperature of the
wheels and tires will increase, leading to an increase in the tire pressure.
The plug is designed to melt at a given temperature that would prevent
this overpressure, per 14 CFR 25.731(d) and 25.735(j):

Melted fuse plug

Fig. 20.19 Melted fuse plug after energy exceedance.

 CHAPTER 20 Landing Performance Testing and AFM 967

25.731 (d) Overpressure burst prevention. Means must be provided in each

wheel to prevent wheel failure and tire burst that may result from excessive
pressurization of the wheel and tire assembly.
25.735 (j) Overtemperature burst prevention. Means must be provided in
each braked wheel to prevent a wheel failure, a tire burst, or both, that may
result from elevated brake temperatures. Additionally, all wheels must meet
the requirements of §25.731(d).

To ensure the plugs don’t melt on every landing, a minimum design require-
ment is established by 14 CFR 25.735 (f)(1):
25.735 (f)(1): Kinetic energy capacity—Design landing stop. The design
landing stop is an operational landing stop at maximum landing weight.
The design landing stop brake kinetic energy absorption requirement of
each wheel, brake, and tire assembly must be determined. It must be substan-
tiated by dynamometer testing that the wheel, brake and tire assembly is
capable of absorbing not less than this level of kinetic energy throughout
the defined wear range of the brake. The energy absorption rate derived
from the airplane manufacturer’s braking requirements must be achieved.
The mean deceleration must not be less than 10 fps2 .

The airplane manufacturer will typically perform one dedicated test to vali-
date the fuse plug energy absorption capability by performing a dedicated
landing at near maximum landing weight (MLW) and targeted maximum
landing energy. FAA AC 25-7D provides a recommended test methodology
for verifying the fuse plug integrity. This author recommends the following
procedure:

(1) Perform a landing that would result in the defined maximum landing
energy following normal landing procedures.
(2) During the landing, the pilot should apply maximum braking force and
not use any reverse thrust.
(3) The airplane must then be taxied at least 3 miles. (This will involve
additional braking along the way.)
(4) Then the airplane needs to be parked in a location that will minimize
the effects of wind cooling. (This author has seen small airplanes be
taxied directly into a hangar.)
(5) The parking brake must be applied if determined to be the most adverse
condition for temperature transfer from brakes to wheel.
(6) The test is over once the temperature peak on the wheel is
reached and cooling of the wheel has started with the fuse plug
not releasing.

Once the maximum landing energy has been provided, the manufacturer
must provide information in the flight manual on the combination of
weight, altitude, temperature, runway slope (downgoing adds more energy
to the brakes), and winds (tailwinds also add more energy) that would
result in an energy in a landing that is greater than the demonstrated value.
968 Operational Aircraft Performance and Flight Test Practices

Coming in Fast: First Phenom 100 Incident

The first Embraer Phenom 100 incident caused no injuries to the two
crewmembers aboard and damage only to the landing gear and flaps when
the light jet overran the runway while landing in the Brazilian beach town
of Angra dos Reis. No passengers were on board at the time of the incident.
A pilot holding for takeoff said the Phenom approached the runway at above-
normal speed. “When he was running out of runway, the captain tried to turn
the aircraft, but it slid backwards off the runway,” he said. Weather was clear at
the time of the incident, but high winds were reported later. According to
Embraer, the damage is repairable, although by one report both main
landing gear came off. According to data from Embraer, the Phenom 100
requires 3125 feet to take off; the field length at Angra Airport is 3001 feet.
The Phenom 100 S/N 49 was delivered just two months prior to the incident.
Source: https://www.ainonline.com/aviation-news/aviation-international-
news/2009-11-02/no-injuries-first-phenom-100-incident

Coming in High

If the airplane crosses the threshold of the runway at an altitude higher

than the expected 50 ft, there will typically be an increase in the landing dis-
tance. The increase in landing distance is approximately 200 ft more runway
required for every 10 ft above the expected 50-ft runway threshold crossing
point if the crew uses the specified landing approach procedure for a
nominal 23-deg approach path.
For example, crossing the runway threshold at 100 ft instead of 50 ft
results in an increase in the minimum landing distance of approximately
1000 ft for a nominal 23-deg approach (see Fig. 20.20).
 CHAPTER 20 Landing Performance Testing and AFM 969

3 deg
100 ft

3 deg
50 ft Flare Flare

Touchdown Touchdown
+1000 ft

Fig. 20.20 Impact of coming in high on landing distance.

For airplanes operating under FAA Part 121, Operating Requirements:

Domestic, Flag, and Supplemental Operations, the airplane cannot be dis-
patched to an airport with a runway less than the landing field length
(LFL, 121.185, 121.195):
§121.195(b) . . .no person operating a turbine engine powered airplane may
take off that airplane unless its weight on arrival, allowing for normal con-
sumption of fuel and oil in flight (in accordance with the landing distance
set forth in the Airplane Flight Manual for the elevation of the destination
airport and the wind conditions anticipated there at the time of landing),
would allow a full stop landing at the intended destination airport within
60 percent of the effective length of each runway described below from a
point 50 feet above the intersection of the obstruction clearance plane and
the runway. . .

So any deviation from being at the proper height at the beginning of the
runway (50 ft AGL) is often masked by the additional operational margin
imposed on the Part 121 operators (but not always, see TSB A05H002 acci-
dent box); however, Part 91 operators are allowed to plan a landing using the
AFM published actual landing distance and should therefore understand the
impact of coming in high.

TSB Accident Report A05H002: Runway overrun and fire, Toronto Pearson
Airport. On 2 Aug. 2005, an Air France Airbus A340-313 arriving at Toronto
Pearson landed long and departed the runway at high ground speed, coming
to a stop in a ravine.
Before departure, the flight crew members obtained their arrival weather
forecast, which included the possibility of thunderstorms. On final approach,
they were advised that the crew of an aircraft landing ahead of them had
reported poor braking action. At about 200 ft above the runway threshold,
while on the instrument landing system approach to Runway 24L with autop-
ilot and autothrust disconnected, the aircraft deviated above the glideslope,
and the groundspeed began to increase. The aircraft crossed the runway
threshold about 40 ft above the glideslope.
970 Operational Aircraft Performance and Flight Test Practices

During the flare, the aircraft traveled through an area of heavy rain, and
visual contact with the runway environment was significantly reduced. The
aircraft touched down about 3800 ft down the runway, reverse thrust was
selected about 12.8 s after landing, and full reverse was selected 16.4 s after
touchdown. The aircraft was not able to stop on the 9000-ft runway and
departed the far end at a groundspeed of about 80 kt.

Landing Smooth
Trying to perform a smooth landing to impress the cabin occupants or for
any other reason may result in a significant increase in the landing distance.
An airplane can typically lose about 1 kt/s while in the flare whereas in the
braking part of the landing run, the change in speed can be around 6 kt/s
with maximum brakes applied. It therefore takes 6 times longer to slow
down in the air than on the ground.
Another impact of extending the air time by flaring more is that the air-
speed will continue decreasing in the air, which will lead to a higher angle of
attack and thus higher pitch attitude at touchdown. This higher pitch attitude
will reduce the tail ground clearance and, for highly swept wings, wing tip
clearance, both of which may lead to ground contact and damage to the
airplane.

TSB Accident Report A12A0082: Runway overrun Transportation Safety

Board of Canada (TSB) report on the August 2012 runway overrun at
St. John’s, Newfoundland, involving a Russian Ilyushin Il-76TD, 13 Aug.
2012 (italics added by author): The TSB found a number of actions that cul-
minated with the 140-ton aircraft rolling off the end of the airport’s 8,500-foot
Runway 11. Despite the use of maximum reverse thrust, the aircraft departed
the hard surface at approximately 40 knots and came to a stop 640 feet beyond
the end of the runway. No injuries were reported to any of the 10 people on
 CHAPTER 20 Landing Performance Testing and AFM 971

board. The first issue centered on the crew’s decision to land on Runway 11
despite a 13-knot tailwind component that exceeded the manufacturer’s limit-
ations. Reported weather at St. John’s at the time of arrival was a 500-foot
ceiling and visibility of four miles in mist. Pilot technique played a role
when, in an apparent attempt to achieve a smooth touchdown, the thrust
levers were never completely retarded to idle until the four-engine transport
was more than 2,000 feet down the runway. The aircraft did not touch down
until it had traveled approximately 3,700 feet beyond the runway threshold,
which left just 4,282 feet remaining. In addition to the wind and pilot-
technique issues, the TSB also determined that the Il-76’s anti-skid system
had been incorrectly serviced sometime before the trip to St. John’s. The
hydraulic lines were installed backward in such a way that the anti-skid
brake pressure was released on the wheels that were experiencing effective
braking and reapplying pressure to wheels that were already skidding. As a
result, the aircraft also experienced reverted-rubber hydroplaning at various
locations along the runway. The Il-76 suffered only minor damage in
the incident.
Source: https://www.ainonline.com/aviation-news/2014-03-03/russian-
transport-overrun-st-johns

Presenting the Information to the Flight Crew

Although one should expect that the landing data presented will be
simpler than the takeoff distance, clear and concise charts/tables and pro-
cedures must be provided to ensure the crew understands the performance
information provided and the resulting margins.

Landing Procedures
The landing distance validated by testing is predicated on the landing
gear wheels crossing the runway threshold at 50 ft above the ground.
Depending on the airplane size and approach pitch attitude, the cockpit
height when the gear crosses the threshold can be significantly different.
Figure 20.21 shows such an example; it was extracted from the
Aerospatiale Concorde flight manual. It may be of benefit to the crew to
provide general information on the cockpit height and aiming point on
approach.
We saw how coming in low resulted in the Global 5000 accident men-
tioned in Chapter 19 (Canada TSB A07A0134, 11 Nov. 2007). For small air-
planes, the distance between the cockpit and the main wheels will minimally
impact the MLG height at threshold, so these details may not need to be
added; however, experience has shown that sometimes this additional infor-
mation may be useful. The manufacturer must cater for “normally expected
in-service deviation.”
972 Operational Aircraft Performance and Flight Test Practices

Visual approach without glide slope guidance – The vertical distance

between the path followed by the pilot’s eyes and that of the
main landing gear is approximately 40 feet with the aircraft in
the approach attitude on a 3 deg approach path. Typical touchdown zone parkings
Where an approach has to be made without any glide slope Concorde
guidance, these eye/wheel characteristics should be borne in 1500 FT. ‘eye path’
mind, and whatever means are available, DME, geographical strong point
features, INS etc, should be used to determine the distance from Touch
touch down in order to establish at least initially, a 3 deg slope 1000 FT. down
to the touch down area. For a given height, the distance in zone
nautical miles to commence a 3 deg glide path is found by dividing
the height in hundreds of feet by 3 e.g. at 1500 feet, the
500 FT.
descent should be commenced 5 nautical miles from touchdown.

On an approach, the ground speed in knots multiplied by 5

gives the rate of descent in feet per minute approximately
equivalent to a 3 deg glide path. Not to scale

747 90’
The sketch of typical touchdown zone markings is included as a Approach conco7rd47 Approximate
concorde 86’
Eye path VC1 e eye weight
VC10 70’
reminder. Until pilots are familiar with the vastly increased 707
0 at threshold
707 65’
pilot eye height at touchdown there will naturally be a tendency All aircraft
for the wheels to touch short of the desired area. On the wheel path
(3 deg approach)
correct Concorde glide slope, the pilot eye path will intercept 50’
the runway at a point about 1800 feet from the threshold. The
approach eye path illustration shows approximate measurements
Threshold 1000’ 2000’
only but is intended to highlight the problem and danger inher-
ent in visual approach judgement on concorde. As can be seen
the situation is almost exactly analogous to the 747.

Pilots vision
reference
Aircra
ft attitu
de
Horizontal 10¼ deg
12½ deg

.R)
(S.V
Limit of nge off
al ra ut
forward
nt visu
e ldc 136’ 100’
vision Sla i
sh Eye Wheel
nd
Runway Wi Approach lights height height

580’ 610’ distance lost

Visual segment Under nose
12½ deg Nose droop
Wheel heights (Assuming aircraft is in centre of ILS glideslope,
over threshold glide slope zero point is 1000 ft. from threshold
and pitch attitude is 10¼ deg

Fig. 20.21 Cockpit height above ground at runway threshold crossing height. Source:
Aerospatiale Concorde flight manual.

We offer the following as an example of proper AFM landing procedure

(this one using manual deployment of devices):
(1) A performance landing is one in which the pilot will follow the VREF
speed very precisely down to 50 ft AGL while carrying just enough power
to maintain a 23-deg approach path.
(2) The aircraft must cross the threshold (50 ft AGL) at the beginning of the
runway where the thrust is rapidly brought to idle.
(3) The pilot will allow the airspeed to decay and will execute a firm
touchdown with minimal flare.
(4) Upon MLG touchdown, the nose must be rapidly derotated, and
maximum braking effort must be initiated immediately upon NLG
touchdown.
(a) The pilot will be required to apply sufficient brake pedal force to
ensure that the antiskid takes over the control of the braking. The
 CHAPTER 20 Landing Performance Testing and AFM 973

pilot must not modulate the brake force, but rather must maintain
the pressure down to the full stop.
(5) The speedbrakes must be deployed immediately after full braking
is initiated.
There is always a balance in the information to be provided, as stated in
the requirements of §25.1585 (shown next with author’s underline). Some
manufacturers have elected to provide minimum information, and so some
flight manuals may not have this information, especially when the large
operational factors are applied and no ALD is provided. But factors must
be explained for the crew to fully appreciate the performance information
provided.
§25.1585 Operating procedures
(b) Information or procedures not directly related to airworthiness or not
under the control of the crew, must not be included, nor must any
procedure that is accepted as basic airmanship.

§25.1587 Performance information

(b) Each Airplane Flight Manual must contain the performance
information computed under the applicable provisions of this part
(including §§25.115, 25.123, and 25.125 for the weights, altitudes,
temperatures, wind components, and runway gradients, as applicable)
within the operational limits of the airplane, and must contain the
following:
(7) An explanation of operational landing runway length factors
included in the presentation of the landing distance, if appropriate.

Approach Airspeed
Approach and landing airspeeds are usually much simpler than the
takeoff airspeeds and can usually be provided on a single page. Temperature
and possibly even altitude do not play much of a role in impacting the air-
speed, and minimum control speed may also not be a player. If the altitude
has an impact on landing speeds, one will notice altitude listed in the infor-
mation such as the chart on the right of Fig. 20.22. Some of the formats that
this author has seen include those found in Figure 20.22 (some incorporated
with the landing distance information or other landing performance
information).

Landing Distance
There are many landing distances that can and will be provided in a flight
manual. The actual landing distance (ALD) is the landing distance demon-
strated in flight testing from which a certification model, with appropriate
certification margins added, was developed and accepted by the authorities.
This distance satisfies §25.125(a):
 974
Uncorrected landing field length (ft)

(×1000 lb)

(×1000 kg)
80

Operational Aircraft Performance and Flight Test Practices

Zero slope, No wind 35
Flaps LDG 75
33 Normal Abnormal Abnormal
A/I off, dry Rwy 70 Abnormal
31 Weight = 55,000 lb flaps 20°
Altitude Weight VAPPR Temperature (°C) flaps 39° flaps 10° flaps UP
(feet) (lb) (KCAS) ISA ISA+ 5 ISA+10 ISA+15 ISA+20 65 29 F39° F39°
Altitude VApp VREF VApp VREF VApp VREF VApp VREF
8000 105 2959 2996 3033 3070 3107 Shake Push
60

Landing weight
8500 108 3072 3111 3150 3189 3228 27 137
9000 111 3201 3243 3284 3325 3366
0 109 102 132 141 136 147 142 155 150
0 9500 114 3330 3374 3417 3460 3504 55 25 EF 2000 109 102 137 132 141 136 147 142 155 150
9800 116 3408 3453 3497 3542 3586 VR 4000 109 102 137 132 142 137 148 143 156 151
10,000 117 3460 3505 3551 3596 3642 50 23
10,400 119 3563 3610 3658 3705 3752 6000 110 102 138 133 143 138 149 144 157 152
21
8000 105 3022 3060 3098 3137 3176 45 8000 111 103 140 135 144 139 151 146 159 154
8500 108 3139 3179 3219 3260 3302 19
9000 111 3272 3314 3357 3400 3444
10,000 112 104 141 136 145 140 152 147 161 156
40
1000 9500 114 3405 3450 3494 3540 3586 17 12,000 113 105 142 137 147 142 154 149 163 158
9800 116 3484 3531 3577 3624 3671 35
10,000 117 3538 3585 3632 3680 3728
14,000 114 107 143 138 148 143 155 150 165 160
15
10,400 119 3644 3693 3742 3791 3841 15,000 114 107 144 139 149 144 156 151 166 161
30
8000 105 3089 3111 3167 3208 3248 Caution: Fuseplug release possible
8500 108 3210 3236 3293 3335 3378 Caution: Tire speed limit
9000 111 3347 3376 3434 3479 3524 with max. braking
2000 9500 114 3484 3516 3576 3623 3670
9800 116 3566 3601 3661 3709 3758
10,000 117 3621 3657 3717 3767 3817 90 100 110 120 130 140 150
10,400 119 3730 3770 3830 3882 3934 Indicated airspeed (kt)
Note : VAPPR = VREF
Landing distance - feet (unfactored) - DRY
Associated conditions: Simplified effects of
Thrust : As required to maintain 3 deg approach angle slope & wind on distance
to 50 ft. retard to idle at 50 ft.
Approach speed : VREF 1% Downhill +200
Flaps : Down 10 kt Tailwind +1100
Anti-skid : Normal 10 kt Headwind –140
Braking : Maximum See correction charts
Lift dump : Extended after touchdown
if more detail required
Pressure altitude : Sea level
V REF ~ KIAS Static air temperature ~ °C ISA+37
Weight Copilot or
PFD standby –40 –30 –20 –10 0 10 20 30 40 50 52
12,500 lb
5670 kg 121 120 2822 2913 3005 3099 3193 3288 3384 3481 3578 3675 3694
11,600 lb
5262 kg 117 116 2738 2816 2895 2972 3050 3134 3219 3306 3393 3481 3498
11,000 lb
4990 kg 114 113 2678 2754 2831 2906 2981 3058 3136 3214 3292 3370 3385
10,000 lb
4536 kg 109 108 2572 2645 2717 2788 2859 2929 2998 3067 3135 3203 3216
9000 lb
4082 kg 104 102 2460 2528 2597 2664 2731 2796 2862 2926 2991 3054 3067
8000 lb
3629 kg 98 96 2335 2398 2461 2523 2586 2647 2708 2767 2827 2886 2898

Fig. 20.22 Approach and landing speed information formats for a flight manual.
 CHAPTER 20 Landing Performance Testing and AFM 975

§25.125 Landing
(a) The horizontal distance necessary to land and to come to a complete stop
(or to a speed of approximately 3 knots for water landings) from a point
50 feet above the landing surface must be determined (for standard
temperatures, at each weight, altitude, and wind within the operational
limits established by the applicant for the airplane):
(1) In non-icing conditions; and
(2) In icing conditions with the most critical of the landing ice
accretion(s) defined in Appendices C and O of this part, as
applicable, in accordance with §25.21(g), if VREF for icing conditions
exceeds VREF for non-icing conditions by more than 5 knots CAS at
the maximum landing weight.

Note that this only requires provision of standard day temperature con-
ditions. One must remember that on days warmer than standard, the airplane
true airspeed will be higher than standard day temperature true airspeed, and
thus, the actual landing distance will be larger than what can be published in
the AFM. Some manufacturers actually provide temperature corrections. We
also note that the wind correction, §25.125(f), adds some conservatism to the
distance:
§25.125 Landing
(f) The landing distance data must include correction factors for not more
than 50 percent of the nominal wind components along the landing path
opposite to the direction of landing, and not less than 150 percent of the
nominal wind components along the landing path in the direction
of landing.

Part 91 operators (noncommercial) are authorized to use ALD to plan a

landing, but one must understand the real meaning of the distance provided.
Some examples of AFM/POH (pilot operating handbook) ALD landing
charts and tables are provided in the figures that follow. Figure 20.23
shows only ALD. Correction factors are provided as simple factors below
the table. As one can see, the information is much simpler to present than
the takeoff distance.
Figure 20.24 provides another format for ALD information, this one for a
specific altitude and flap configuration. It incorporates temperature and
winds’ impact on ALD. Figure 20.25 also presents ALD information with
accompanying information on what ALD truly represents for this chart.
Some manufacturers will offer correction for runway slope on the chart
(when within +2%); see Fig. 20.26.
The usual additional (or sometimes the only one provided for commercial
airplanes) other landing distance is the landing field length (LFL), which is
defined by the requirements of §121.195. The LFL corresponds to the
runway distance under which an airplane can land and come to a complete
stop within 60% of the available runway (see Fig. 20.27) (70% for turboprops).
The LFL value is used to decide if an airplane can be dispatched to an
intended runway (author’s underline).
976 Operational Aircraft Performance and Flight Test Practices

SHORT FIELD LANDING DISTANCE

AT 2550 POUNDS
CONDITIONS:
Flaps 30 deg
Power off
Maximum braking
Paved, level, dry runway
Zero wind
Speed at 50 Ft: 61 KIAS

0°C 10°C 20°C 30°C 40°C

Grnd Total Grnd Total Grnd Total Grnd Total Grnd Total
Press Roll Ft To Roll Ft To Roll Ft To Roll Ft To Roll Ft To
Alt Ft Clear Ft Clear Ft Clear Ft Clear Ft Clear
in 50 Ft 50 Ft 50 Ft 50 Ft 50 Ft
Feet Obst Obst Obst Obst Obst
S.L. 545 1290 565 1320 585 1350 605 1380 625 1415
1000 565 1320 585 1350 605 1385 625 1420 650 1450
2000 585 1355 610 1385 630 1420 650 1455 670 1490
3000 610 1385 630 1425 655 1460 675 1495 695 1530
4000 630 1425 655 1460 675 1495 700 1535 725 1570
5000 655 1460 680 1500 705 1535 725 1575 750 1615
6000 680 1500 705 1540 730 1580 755 1620 780 1660
7000 705 1545 730 1585 760 1625 785 1665 810 1705
8000 735 1585 760 1630 790 1670 815 1715 840 1755

Note:
1. Short field technique as specified in Section 4.
2. Decrease distances 10% for each 9 knots headwind. For operation
with tail winds up to 10 knots, increase distances by 10% for each
2 knots.
3. For operation on dry, grass runway, increase distances by 45% of
the “ground roll” figure.
4. If landing with flaps up, increase the approach speed by 9 KIAS and
allow for 35% longer distances.

Fig. 20.23 Example of ALD.

§121.195 Airplanes: Turbine engine powered: Landing limitations: Des-
tination airports

(b) Except as provided in paragraph (c), (d), or (e) of this section, no person
operating a turbine engine powered airplane may take off that airplane
unless its weight on arrival, allowing for normal consumption of fuel and
oil in flight (in accordance with the landing distance set forth in the
Airplane Flight Manual for the elevation of the destination airport and the
wind conditions anticipated there at the time of landing), would allow a
full stop landing at the intended destination airport within 60 percent of
the effective length of each runway described below from a point 50 feet
above the intersection of the obstruction clearance plane and the runway.
For the purpose of determining the allowable landing weight at the
destination airport the following is assumed:
(1) The airplane is landed on the most favorable runway and in the most
favorable direction, in still air.
(2) The airplane is landed on the most suitable runway considering the
probable wind velocity and direction and the ground handling
characteristics of the airplane, and considering other conditions such
as landing aids and terrain.

Note the additional information for wet runway. This extra distance is the
LFL increased by a distance of 15%, so the landing distance on wet runway
 CHAPTER 20 Landing Performance Testing and AFM 977

FLAP 31 3000 FT
Uncorrected actual landing distance
dry level runway
WAI off

VAPP (KIAS) 98 101 104 106 109 111 114 116 118 119 124
VREF (KIAS) 93 96 99 101 104 106 109 111 113 114 119
OUTSIDE LANDING WEIGHT, LBS (KG)
AIR TOWER 8500 9000 9500 10000 10500 11000 11500 12000 12500 12725 13950
TEMP WIND (KT) (3856) (4082) (4309) (4536) (4763) (4990) (5216) (5443) (5670) (5772) (6328)
–10 2645 2715 2790 2877 2963 3049 3134 3212 3289 3323 3507
–40 °C 0 2238 2301 2363 2426 2489 2552 2614 2672 2730 2756 2894
10 2114 2174 2234 2294 2354 2414 2475 2531 2586 2610 2744
ISA – 49 °C 20 1996 2053 2111 2168 2225 2283 2341 2394 2447 2471 2600
30 1884 1939 1993 2048 2103 2158 2213 2264 2315 2337 2461
–10 2759 2848 2942 3035 3127 3219 3311 3394 3476 3514 3711
–20 °C 0 2338 2406 2474 2542 2610 2678 2747 2809 2872 2900 3077
10 2209 2275 2340 2405 2470 2536 2601 2662 2722 2749 2895
ISA – 29 °C 20 2087 2149 2211 2274 2336 2399 2462 2520 2578 2604 2744
30 1970 2029 2088 2148 2208 2268 2328 2384 2439 2464 2599
–10 2891 2992 3092 3191 3290 3388 3485 3574 3662 3702 3913
0 °C 0 2437 2511 2584 2658 2731 2805 2879 2949 3030 3067 3263
10 2304 2375 2445 2516 2586 2657 2728 2794 2858 2888 3057
ISA – 9 °C 20 2177 2244 2312 2379 2447 2515 2583 2646 2709 2737 2889
30 2055 2120 2184 2249 2314 2379 2444 2505 2565 2592 2738
–10 2959 3064 3166 3269 3370 3471 3572 3663 3754 3795 4012
10 °C 0 2487 2564 2640 2716 2792 2868 2946 3031 3115 3153 3354
10 2352 2425 2498 2571 2645 2718 2792 2859 2927 2957 3146
ISA + 1 °C 20 2222 2292 2362 2432 2503 2573 2644 2709 2774 2803 2961
30 2098 2165 2232 2299 2367 2435 2503 2565 2628 2656 2807
–10 3027 3134 3240 3345 3450 3554 3658 3752 3846 3888 4112
20 °C 0 2537 2615 2695 2773 2852 2932 3025 3112 3199 3238 3446
10 2399 2475 2551 2627 2703 2779 2855 2925 2995 3033 3236
ISA + 11 °C 20 2267 2340 2413 2485 2558 2632 2705 2773 2840 2870 3034
30 2141 2210 2280 2350 2420 2490 2561 2626 2691 2720 2877
–10 3094 3204 3314 3422 3530 3637 3743 3841 3937 3980 4211
30 °C 0 2586 2668 2750 2831 2913 3005 3103 3194 3283 3323 3538
10 2447 2525 2604 2682 2761 2840 2919 2991 3077 3115 3325
ISA + 21 °C 20 2313 2387 2463 2539 2614 2690 2766 2836 2906 2937 3117
30 2184 2256 2329 2401 2473 2546 2619 2687 2754 2784 2947
–10 3161 3274 3386 3498 3609 3719 3828 3929 4028 4072 4310
40 °C 0 2636 2720 2804 2889 2978 3080 3182 3275 3367 3408 3629
10 2494 2575 2656 2737 2819 2900 2982 3068 3158 3198 3413
ISA + 31 °C 20 2358 2435 2513 2592 2670 2748 2827 2900 2972 3004 3203
30 2228 2302 2376 2452 2527 2602 2678 2748 2817 2849 3018
–10 3217 3332 3447 3562 3675 3788 3900 4003 4104 4150 4394
49 °C 0 2678 2764 2851 2938 3039 3144 3248 3343 3438 3480 3706
10 2534 2617 2700 2784 2868 2952 3042 3135 3227 3268 3489
ISA + 40 °C 20 2396 2476 2556 2637 2717 2798 2879 2954 3028 3061 3277
30 2264 2340 2417 2495 2572 2650 2728 2800 2871 2903 3078
NOTE: Shaded fields indicated conditions exceeding quick turn-around or climb limits.
NOTE: Weight values above the maximum structural limit of 12725 Ibs are provided for emergency landings only.

Fig. 20.24 ALD format 1.

becomes (ALD  1.92). Figures 20.28 and 20.29 are other examples of LFL
charts including additional corrections.
Some manufacturers may elect to provide a table format rather than a
chart format (see Fig. 20.30). The information is easier to read but usually
produces more pages in the flight manual.
 978
Operational Aircraft Performance and Flight Test Practices
UNFACTORED LANDING DISTANCES UNFACTORED LANDING DISTANCE (m)
Unfactored landing distance is the actual distance to land the airplane ENGINE ICE PROTECTION OFF - WINGSTAB OFF
from a point 50 ft above runway threshold to complete stop, using the FLAP FULL
landing technique described in the beginning of this section.

Unfactored Landing Distance tables are presented for a set of ALTITUDE

pressure altitudes and winds for the conditions below: Weight –1000 ft 0 ft
(kg) WIND
- Dry runway;
–10 kt 0 kt 10 kt 20 kt –10 kt 0 kt 10 kt 20 kt
- Zero slope; 5600 759 636 597 559 773 649 610 572
- Ice Protection System: 5800 774 650 611 573 788 663 624 585
For Wingstab OFF, Engine Ice Protection OFF or ON, 6000 790 665 626 587 804 679 639 600
6200 805 680 640 601 820 693 653 614
For Wingstab ON, Engine Ice Protection ON;
6400 821 694 654 615 835 708 668 628
- Temperature: ISA (1); 6600 836 709 668 629 851 723 682 642
- No SAT effect; 6800 851 722 682 642 866 737 696 656
- No vREF Overspeed; 7000 865 736 695 655 881 751 710 669
7200 880 750 709 669 897 766 724 683
- No Drag Index. 7400 895 765 723 682 912 780 738 697
7600 909 778 736 695 926 794 752 710
NOTE: 1) Wing and Stabilizer Ice-Protection system is auto-inhibited
7800 924 791 749 708 942 808 765 723
outside its operational envelope. Refer to its operational
8000 943 806 763 722 964 823 780 738
envelope in Section 2 for further information.
8200 963 820 777 735 985 837 793 751
NORMAL OPERATION CAUTION: SHADED AREAS REPRESENT CONDITIONS WHERE
The required landing distance for dispatch is the unfactored landing THE MAXIMUM LANDING WEIGHT OR CLIMB LIMITED
distance increased by a factor according to the operating regulations. WEIGHT WAS EXCEEDED.

Fig. 20.25 ALD format 2.

 CHAPTER 20 Landing Performance Testing and AFM 979

LANDING TOTAL DISTANCE - FLAPS 40°

FROM 50 FT; (STANDARD UNITS)
ASSOCIATED CONDITIONS: EXAMPLE:
APPROACH AT 1.3 Vs1 WEIGHT ~ LB VAPP ~ KIAS ALTITUDE ~ FT 6000 FT
REFER TO THE SPEED SCHEDULE TABLE 6400 67 OAT ~ °C 18 °C
AVERAGE BRAKING TECHNIQUE 7300 72 WEIGHT ~ LB 7716 LB
GROUND IDLE AFTER TOUCH DOWN 8200 76 HEADWIND COMPONENT ~ KT 8 KT
RUNWAY SURFACE: TARMAC 9100 80 UPHILL COMPONENT ~ % 1%
SEE SECTION 2 - LIMITATIONS 10,000 84 LANDING TOTAL DISTANCE ~ FT 1970 FT
3400

Pres

Reference line

Reference line

Reference line
30 3200
°C 20 10,000 su

d
~

in
re alt
A 10

ll
IS

Tailw
3000

hi
0

wn
itud

Landing total distance ~ FT

–10 8000

Do
–20 2800
e~f

–30 6000
–40 2600
t

4000 Up
2000 hil
l
2400

He
0

a
dw
2200

in
d
2000

1800

1600
–60 –40 –20 0 20 40 60 11,000 9000 7000 5000 0 10 20 30 40 0 2 4
Outside air temperature ~ °C Weight ~ LB Wind component – KT Slope ~ %

Fig. 20.26 ALD with runway slope correction.

Notes
Consult using airline for specific operating
procedure prior to facility design
Zero runway gradient
2.50 Zero wind
8
Pressure Altitude
feet (meters)
2.25 10,000 (3049)
8000 (2439)
7 6000 (1829)
4000 (1219)
Far landing runway length

2000 (609)
2.00 Sea level
(1000 meters)

1000 feet

6
1.75

1.50 5

1.25
4
Dry runway
Wet runway
1.00
3
300 320 340 360 380 400 420 440 460
1,000 pounds

140 150 160 170 180 190 200 210

(1000 kilograms)
Operational landing weight

Fig. 20.27 LFL chart 1.

980 Operational Aircraft Performance and Flight Test Practices

Maximum landing weight - field length limited Maximum landing weight - field length limited
Flaps 22 deg - no ice encounter Flaps 22 deg - no ice encounter
Chart 1 of 2 Chart 2 of 2

5600 5600
3400

Ref. Line

Ref. Line

Ref. Line
Ref. Line
Pressure altitude - ft 5400 5400
3300
5200 5200
13800 3200
5000 5000
13000
3100
4800 4800
12000
3000
4600 4600
11000
2900 4400 4400
10000
2800 4200 4200
9000
2700 4000 4000
8000

7000 2600 3800 3800

6000 3600 3600
2500

Landing field length - m

5000 3400
3400
2400

Transfer scale

Transfer scale
4000
3200 3200
3000 2300
2000 3000 3000
1000 2200
SL 2800 2800
2100
–1000 2600 2600
2000 2400 2400
1900 2200 2200
1800 2000 2000

1700 1800 1800

1600 1600
1600
1400 1400
1500
1200 1200
1400
1000 1000
1300
800 800
1200 600 600
1100 400 400
26000 30000 34000 38000 42000 46000 50000 54000 –10 0
10 20 30 –3 –2 –1 0 1 2 3 Dry Wet Fact. Unfact.
Weight - lb Head Down Up Runway Factor
Tail
Wind - kt Slope - % condition correction

Fig. 20.28 LFL chart 2.

Note that operators of Part 135 airplanes, operating on-demand services,

are allowed to use a reduced landing distance factor. Under §135.385, the
safety factor is reduced from 60% to 80% (author’s underline).
§135.385 Large transport category airplanes: Turbine engine powered:
Landing limitations: Destination airports

(f) An eligible on-demand operator may take off a turbine engine powered
large transport category airplane on an on-demand flight if all of the
following conditions exist:
(1) The operation is permitted by an approved Destination Airport
Analysis in that person’s operations manual.
(2) The airplane’s weight on arrival, allowing for normal consumption of
fuel and oil in flight (in accordance with the landing distance in the
Airplane Flight Manual for the elevation of the destination airport
and the wind conditions expected there at the time of landing), would
allow a full stop landing at the intended destination airport within 80
percent of the effective length of each runway described below from a
point 50 feet above the intersection of the obstruction clearance plane
and the runway. For the purpose of determining the allowable landing
weight at the destination airport, the following is assumed:
(i) The airplane is landed on the most favorable runway and in the
most favorable direction, in still air.
(ii) The airplane is landed on the most suitable runway considering
the probable wind velocity and direction and the ground
handling characteristics of the airplane, and considering other
conditions such as landing aids and terrain.
(iii) The operation is authorized by operations specifications.
 Performance
Landing field length Category A brakes
and speed
Antiskid operative
12,000
Flaps 40

Note change 11,000

Automatic speed brake operative, for manual in scale

(Tailwind)
speed brakes, increase field length by 450 ft.
Wet runway distance can also be obtained by 10,000
multiplying the dry runway distance by 1.15

CHAPTER 20
For use with autoland, increase speed by 9000
5 Knots (VREF + 5).
See the landing distance adjustment chart (for
automatic landing).

Landing field length - FT

8000
2

0 ft

Landing Performance Testing and AFM

100 1
e– 7000
ud
ltit 1
ea
sur 9

(Headwind)
res 8 2
ort p 7
6 6000
A irp 5
4
3
2
1
0
–1
160 5000

Airspeed (VREF) – knot IAS

150
t
000 f
de–1 140 4000
altitu
ssure 9
p o r t pre 130
Air –1
120 3000

Ref. Line

Ref. Line

2 Wet
1 Dry
110
100 2000
32 36 40 44 48 52 56 60 64 –15 –10 –5 0 20 40 Forecast runway
Gross weight – 1000 kg Wind – knots condition

Fig. 20.29 LFL chart 3.

981
982 Operational Aircraft Performance and Flight Test Practices

Dry and wet runway

Flaps 30°
Actual, factored and wet landing distances
Uncorrected landing distance – FT
ALT WT Actual Factored Wet ALT WT Actual Factored Wet
FT LB distance distance distance FT LB distance distance distance
28000 2290 3820 4390 28000 2600 4330 4980
30000 2400 4000 4600 30000 2730 4550 5230
Sea 32000 2510 4180 4800 32000
6000 34000
2860 4760 5480
34000 2620 4360 5010 2990 4980 5730
level 36000 2730 4540 5220 36000 3120 5200 5980
38000 2830 4710 5420 38000 3250 5410 6220
40000 2940 4890 5620 40000 3380 5620 6470
42000 3040 5070 5830 42000 3510 5840 6720
28000 2340 3890 4480 28000 2660 4430 5090
30000 2450 4080 4690 30000 2790 4650 5350
32000 2560 4260 4900 32000 2930 4880 5610
1000 34000 2670 4450 5120 7000 34000 3060 5100 5860
36000 2780 4640 5330 36000 3200 5320 6120
38000 2890 4820 5540 38000 3330 5540 6370
40000 3000 5000 5750 40000 3460 5770 6630
42000 3110 5180 5960 42000 3600 5990 6890

Fig. 20.30 Table format of landing distance including ALD, LFL, and wet runway.

So an airplane with an expected ALD of 2400 ft at the destination airport

would normally need to dispatch (under Parts 121 and 135) to a runway
with a minimum length equal to the LFL of 2400/0.60 or 4000 ft. If the oper-
ator is authorized by the FAA, that reduced landing distance factor would
mean that the operator could plan to land on runways as short as 2400/
0.80 or 3000 ft—a significant difference (and reduced safety margin) and a
large increase in accessibility to smaller airports.

Time of Arrival Landing Performance

The operational regulations discussed previously stipulate that the air-
plane shall not dispatch if the predicted landing performance would not
allow the airplane to land within a prescribed distance. (Sections 91.1037,
121.195, and 135.385 prescribe landing performance requirements that
must be met at the time of takeoff.) This approach (pun intended) has
proven to be acceptable for many years, and no formal requirements, certifi-
cation or operational, exist to mandate the crew to review the airplane’s per-
formance once it reaches the destination airport. This changed following the
accident of a Southwest Airlines 737-700 in Chicago in 2005 (see accident
box, “NTSB/AAR-07/06, Runway Overrun and Collision, Southwest Airlines
Flight 1248”). This accident clearly showed that the distance needed to safely
complete the landing at the time of arrival may be different if the runway,
runway surface condition, meteorological conditions, approach guidance,
airplane configuration, airplane weight, approach speed, or use of airplane
ground deceleration devices differs from that used to show compliance
with §§91.1037, 121.195, or 135.385. We note that as of 2020, the Flight
Test Harmonization Working Group (FTHWG) was working to define
 CHAPTER 20 Landing Performance Testing and AFM 983

Part 25 requirements to incorporate the time of arrival performance; once the

recommendations from the group are complete, they will go through formal
rule-making procedures with the FAA, EASA, and TCCA.

NTSB/AAR-07/06, Runway Overrun and Collision Southwest Airlines

Flight 1248 Boeing 737-7H4, N471WN, 8 Dec. 2005: . . .Southwest Airlines
(SWA) flight 1248, a Boeing 737-7H4, N471WN, ran off the departure end of
runway 31C after landing at Chicago Midway International Airport, Chicago,
Illinois. The airplane rolled through a blast fence, an airport perimeter fence,
and onto an adjacent roadway, where it struck an automobile before coming to
a stop. . ..
The National Transportation Safety Board determined that the probable
cause of this accident was the pilots’ failure to use available reverse thrust
in a timely manner to safely slow or stop the airplane after landing, which
resulted in a runway overrun. This failure occurred because the pilots’ first
experience and lack of familiarity with the airplane’s autobrake system dis-
tracted them from thrust reverser usage during the challenging landing.
. . .[C]ontributing to the accident was the pilots’ failure to divert to another
airport given reports that included poor braking action and a tailwind com-
ponent greater than 5 knots. Contributing to the severity of the accident
was the absence of an engineering materials arresting system, which was
needed because of the limited runway safety area beyond the departure end
of runway 31C.
[The airplane was operating under FAA 14 CFR Part 121, which requires
operators to perform preflight landing distance calculations before they depart
on a flight, in part, to determine the maximum takeoff weight at which the air-
plane can depart, travel to the destination, and safely land on the available
landing distance at the destination and/or alternate airport. Although pre-
flight landing distance assessments are standardized by federal regulations,
the assessments do not attempt to comprehensively account for the actual
conditions, configuration, and pilot techniques that exist and/or occur at
the time of arrival.]

Following the accident and the recommendations of the NTSB, the

Takeoff and Landing Performance Assessment (TALPA) initiatives were
launched (https://www.faa.gov/about/initiatives/talpa/). Out of this initiat-
ive, the FAA produced the guidance material in Advisory Circular AC 25-32,
Landing Performance Data for Time-of-Arrival Landing Performance
Assessments. This AC provides guidance and standardized methods that
manufacturers/applicants and airplane operators can use when developing
landing performance data for time-of-arrival landing performance assess-
ments for transport category airplanes. It applies landing factors (safety
margin) based on actual airplane weight and configuration selected,
984 Operational Aircraft Performance and Flight Test Practices

Table 2. Runway surface condition—pilot-reported braking action—

wheel braking coefficient correlation matrix
Runway Pilot-reported
condition Runway surface condition description braking Wheel braking coefficient
code action
6 • Dry — 90% of certified value used to
comply with § 25.1251.
5 • Frost Good Per method defined in § 25.109(c).
• Wet (includes damp and " (3 mm)
depth or less or water)
" (3 mm) depth or less of:
• Slush
• Dry snow
• Wet snow
4 –15 °C and colder outside air temperature: Good to 0.203
• Compacted snow Medium2

3 • Wet ("Slippery When Wet“ Medium2 0.163

runway)
• Dry snow or wet snow (any depth)
over compacted snow
Greater than " (3 mm) depth of:
• Dry snow
• Wet snow
Warmer than –15 °C outside air
temperature
• Compacted snow
2 Greater than " (3 mm) depth of: Medium2 to (1) For speeds below 85% of the
• Water Poor hydroplaning speed4: 50% of
• Slush the wheel braking coefficient
determined in accordance with
§ 25.109(c), but no greater than
0.163; and
(2) For speeds at 85% of the
hydroplaning speed4 and above:
0.053.
1 • Ice Poor 0.083

0 • Wet ice Nil Not applicable. (No operation in

• Water on top of compacted snow Nil conditions.)
• Dry snow or wet snow over ice
1100% of the wheel braking coefficient used to comply with § 25.125 may be used of the testing from which that

braking coefficient was derived was conducted on portions of runways containing operationally representative
amounts of rubber contamination and paint stripes.

Fig. 20.31 Table 2 of FAA AC 25-32.

runway selected (runway slope), actual winds and temperature, runway

surface conditions/braking action (pilot report), pressure altitude, icing con-
dition, final approach speed, and deceleration devices used. This AC also pro-
motes the use of consistent terminology for runway surface conditions used
among data providers and FAA personnel (see Fig. 20.31). The AC also aligns
the FAA requirements with those used for decades by the EASA (CS25.1591
and AMC 25.1591) and TCCA [§525.1581(g)] on the definition of contami-
nated runways.
The landing distance defined by the margins provided by FAA AC 25-32
is labeled operational landing distance (OLD). The guidance mentions that
the landing distance for a time-of-arrival landing performance assessment
may be determined analytically from the landing performance model
 CHAPTER 20 Landing Performance Testing and AFM 985

developed to show compliance with §25.125. Some of the additional margins

that are included in FAA AC 25-32 include:
• Minimum airtime defined by the test results supporting §25.125 but not
less than 7 s (similar to AMC 25.1591).
• The air distance should be determined at a speed of 98% of the
recommended speed over the landing threshold multiplied by the
airtime. The touchdown speed would then be 96% of the threshold
speed.
• The touchdown rate of descent should be in the range of 1 to 4 ft/s.
• The transition distance should be based on the recommended procedures
for use of the approved means of deceleration, in terms of both sequencing
and any cues for initiation. Reasonably expected time delays should also be
taken into account.
• The calculation of the final stopping configuration distance should be
based on the braking coefficient associated with the runway surface
condition or pilot-reported braking action, including the effect of
hydroplaning, if applicable (see Fig. 20.31).
• Landing distances used for time-of-arrival landing performance
assessments may include credit for the stopping force provided by reverse
thrust, consistent with the procedures established for its use. The
procedures should include all of the pilot actions necessary to obtain the
recommended level of reverse thrust, maintain directional control and safe
engine operating characteristics, and return the reverser(s), as applicable,
to either the idle or the stowed position.
The AC goes on to mention that the data for time-of-arrival landing perform-
ance assessments should represent expected landing performance by a
trained flight crew of average skill following normal flight procedures and
training. Contaminated runway landing performance data approved by
either the Joint Aviation Authorities or EASA in compliance with either
their contaminated runway type certification or operating requirements are
acceptable when using the optional process identified in this AC with
some exception. (See the AC for more details.)
As an example of constantly evolving regulations, EASA mandated in 2019
that landing distance at time of arrival (LDTA, equivalent to FAA OLD) be
made available to all European operators by Nov. 2020. The information for
this mandate can be found in operating rule CAT.OP.MPA.303:

(1) No approach to land shall be continued unless the landing distance

available (LDA) on the intended runway is at least 115 % of the landing
distance at the estimated time of landing.
(2) Performance information for the assessment of the LDTA shall be based
on approved data contained in the AFM. When approved data contained
in the AFM are insufficient in respect of the assessment of the LDTA,
they shall be supplemented with other data which are either determined
in accordance with the applicable certification standards for aeroplanes
or determined in line with the AMCs issued by the Agency.
986 Operational Aircraft Performance and Flight Test Practices

The definition of LDA for the FAA is found in FAA AC 150/5300-13A,

Airport Design: LDA—the length of runway declared available and suitable
for satisfying landing distance requirements. The LDA begins at the
threshold. When the runway safety area (RSA), runway object free area
(ROFA), approach runway protection zone (RPZ), and threshold siting
requirements are met, the threshold is normally placed at the beginning of
the runway. When these requirements are not met, the threshold may
be displaced.
EASA also introduced the requirement that the AFM contain a statement
allowing the use of reduced landing distance factor (RLDF) such as what is
allowed by FAA §135.385 (80% factor rather than the 60% factor of LFL) to
allow the operational use of that factor. The operational requirement was
spelled out in EASA CAT.POL.A.255:

(a) An aeroplane operator may conduct landing operations within 80 % of

the landing distance available (LDA) if it complies with the following
conditions:
(1) the airplane has an MOPSC of 19 or less;
(2) the airplane has an eligibility statement for reduced required landing
distance in the AFM;
(3) the airplane is used in non-scheduled on-demand commercial air
transport (CAT) operations;
(4) the landing mass of the aeroplane allows a full-stop landing within
that reduced landing distance;
(5) the operator has obtained a prior approval of the competent
authority.

At the time of writing this chapter, the FAA and TCCA did not have such
a requirement.

Steep Approach
This author has been involved in several steep approach certification test
campaigns from as early as 2002 (see [1]). Steep approaches are defined as
those requiring an approach path down to the threshold of 24.5 deg or
steeper. The main difference in steep approach landing distances is most
often limited to the air distance, with the transition and full braking segments
remaining essentially the same.

London City Airport

London City Airport (LCY) is a single-runway (4948 ft long) airport
located in the Royal Docks in the center of London’s financial district. The
feasibility of operation by short takeoff and landing (STOL) airplanes was
demonstrated during the summer of 1982 when a de Havilland Dash 7 was
landed on the cleared docks (see Fig. 20.32).
The airport was then built on the same location and started services
in Oct. 1987. The airport is surrounded by industry and houses/apartments,
 CHAPTER 20 Landing Performance Testing and AFM 987

Fig. 20.32 The de Havilland Dash 7’s first landing on the docks in London, 27 June 1982.

so special procedures were put in place for noise abatement purposes. For
approach, airplanes are required to maintain a flight path angle of
25.5 deg down to threshold altitude (50 ft AGL). Access to this airport
has resulted in special certification regulations being developed by the
various authorities to address the safety requirements of the greater approach
angle.
The airport’s only runway is 4948 ft (1508 m) long, but has a declared
landing distance available (LDA) of only 4327 ft (1319 m). This is usually
not a problem for propeller-driven airplanes, but it is becoming increasingly
common for jet airplanes (mostly commercial flights, all business jets) to
operate from this airport.

Regulations and Guidance Material

Some of the additional requirements for steep approach certification, as
compared to normal performance landing tests, are that the airplane must
be capable of demonstrating approaches and landing from an approach
flight path angle 2 deg steeper than the angle at which certification is
sought while being able to reduce the touchdown rates to 3 ft/s or less.
This additional angle has been deemed the minimum safe margin for steep
approach operation. For LCY, this means that landing demonstration must
be done at up to 7.5-deg approach angles.
This author has often debated the merit of such a large abuse angle with
the authorities. The original airplane to go to London City airport, the Dash 7,
had enough thrust deficit on approach to achieve that angle. The airplane
also had an approach speed of around 80 kt. For that airplane at that
approach speed, a wind speed change of 20 kt (considered a strong change
in wind on approach) represents about a 2-deg change in approach angle.
The problem with keeping 2 deg is that airplanes with much higher approach
speeds (as normally expected for jets) must now demonstrate significantly
more margin than the earlier airplanes; for example, a jet airplane coming
in to land with an approach airspeed of 124 kt subjected to a 10-kt wind
change would, in fact, only have a flight path angle change of about 1 deg
[Eq. (12.32)]. This author contends that either greater margins are required
988 Operational Aircraft Performance and Flight Test Practices

of faster airplanes or not enough from slower airplanes. That argument has so
far not been accepted by the authorities.
The combined requirements of demonstrating a 27.5-deg approach
angle with a sink rate of less than 3 ft/s at touchdown and the short
runway at London City present a special challenge to the certification team,
some of which this author presented in [1] and that will be discussed next.
In the late 1990s and early 2000s, the reference material for steep
approach came in the form of the Joint Aviation Authorities (JAA) of
Europe NPA25-267. Now, most authorities have formally documented the
steep approach requirements in the following certification requirements
and/or guidance material (not all harmonized yet):

• EASA CS25 Book 1 Appendix Q, Additional Airworthiness Requirements

for Approval of a Steep Approach Landing (SAL) Capability, and EASA
CAT.POL.A.245, Approval of Steep Approach Operations (https://www.
easa.europa.eu/sites/default/files/dfu/Annexes%20to%20Regulation.pdf).
• TCCA AC 525-011, Approval of Steep Approach Landing Capability of
Transport Category Aeroplanes
• FAA AC 25-7D, section 42.4, Criteria for Approval of Steep Approach to
Landing

All three agencies agree that this material applies for an approach path
steeper than (or equal to) 24.5 deg, that a demonstration of a safe landing
from 2 deg steeper than the nominal value must be successfully demon-
strated, and that the sink rate must be reduced to 3 ft/s or less no later
than at touchdown.

Flight Path Capability and Abuse Case

In order to show compliance with the abuse case landing scenario (2 deg
steeper than nominal approach), the manufacturer must understand the
descent flight path capability of the airplane. The most adverse CG location
for a descent is the aft CG (least drag) for a conventional airplane layout (tail
in the back) and light weight. The most restrictive of the three agencies in
terms of capability demonstration is the FAA, which has requirements to
successfully demonstrate the landing while the engine is “off idle.” This state-
ment is a little vague and often is a qualitative evaluation (not clearly quantifi-
able). The FAA has accepted in the past that although the approach is
performed at the abuse angle, the throttle should not remain at idle for any
significant amount of time, which is meant to say the pilot has the ability to
modulate the thrust to maintain the flight path. Depending on the engine’s
response time at near-idle setting, this can easily represent a 0.2- to 0.4-deg
additional margin over the 2-deg-steeper approach. The manufacturer can
perform an initial investigation (this is not a certification requirement, more
of a program risk reduction) of the airplane’s actual descent capability by
 CHAPTER 20 Landing Performance Testing and AFM 989

adopting the proposed steep approach landing configuration and airspeed

while the engines are at idle to determine the actual descent capability of
the airplane.
Most Part 25 airplanes have relatively low drag, even in the approach
configuration, and they may have marginal or no capability to achieve
the flight path angle required for steep approach. One should expect
that the abuse condition may be achieved from one of the following
combinations:

• Normal landing configuration and normal landing reference speed

(ideal)
• Normal landing configuration with increased landing reference speed
• This typically increases the airplane’s drag and reduces the thrust (due to
lapse rate), thus increasing the approach path capability.
• The case of the Learjet 45 [1] showed that a constant approach speed
SVREF of 124 KIAS provided the required approach path at all weights
while the airplane was flying in a normal landing configuration.
• Normal landing configuration with extra drag devices deployed
• Ideally, just speed brakes deployed with no impact on airplane lift
capability (see the Avro RJ airplane’s tail split airbrake or HondaJet tail
airbrake).
• Often, the manufacturer will need to deploy wing spoilers to create the
extra drag, but this will also reduce the wing lift capability; thus, there
will be a need to increase the approach airspeed.
• Another approach is the direct lift control (DLC), where the flight path
angle is maintained by dynamically adjusting the wing spoilers rather than
pitching the nose up or down.

In all cases, if the approach speeds are increased to achieve the desired abuse
approach angle, this does increase the airplane rate of descent and thus will
impact the flare height/required load factor in the flare. We will address this
in the following section. This author quoted the approach speed in the steep
approach configuration SVREF , which must be equal to or larger than the
minimum requirements of 1.23 VSR . EASA has labeled this airspeed
VREFðSALÞ .
Headwind does provide some relief to the approach requirements. The
approach path to London City airport is one that is fixed with respect to
the ground; we call this the glide path. We showed the impact of the wind
on the climb angle in Eq. (12.32), and we use the same equation here to quan-
tify the impact of the wing on the flight path angle (still the angle with respect
to the air mass and a function of the airplane’s excess thrust). Approaching
with a headwind can relieve the flight path requirements by half a degree
or more (see Fig. 20.33), meaning that the airplane must carry more thrust
to descend. This leads to an engine thrust setting that is higher and usually
a more responsive engine when thrust increase or decrease is needed
990 Operational Aircraft Performance and Flight Test Practices

Final approach ICT 1000 ft AGL (2300 ft MSL)

124 KIAS, ISA conditions (128.2 KTAS)
VW 30 kts head wind
10

Glide path (deg) - following the

20 kts head wind
RDapp SVREF 9 10 kts head wind
8 Abuse approach angle no wind
5.5 deg (GP) 7
␥ Nominal approach angle
6

PLASI
5
4
Approach flight path wrt 3
the air mass with 5.5 deg fixed ILS beam 2
head wind (relative to ground) 1
0
0 1 2 3 4 5 6 7 8 9 10
Flight path (deg)

Fig. 20.33 Impact of headwind on flight path angle for London City approach.

(helps with flight path control). Flight test requirements include demonstrat-
ing the abuse landing capability with calm wind conditions (less than 5 kt).

Flare Capability
An important concern for steep approach is the ability of the airplane to
reduce the sink rate to reasonable values prior to touchdown. The certifica-
tion requirements do ask for strong flare capability in terms of equivalent
touchdown sink rate of equal to or less than 3 ft/s. This needs to be achieved
no later than at touchdown. Note that the sink rate could be reduced below
this value and still have a sink rate greater than 3 ft/s at touchdown, and thus
still be compliant [1]. Assuming the airplane is capable of instantaneous load
factor increase from the approach value of near 1.0, the flare height would
change proportionally [1/(n 2 1)], the peak load factor having the greatest
contribution on the ability to turn the corner, as shown in Chapter 12, Eq.
(12.39). We offer Fig. 20.34 as a means to quantify flare height for a given
true airspeed on approach. We compare the normal approach angle of
23 deg with that of the nominal approach of 25.5 deg in London City and

SVref Flt path RS 50 ft n r RS_TD ␥ TD h flare

KTAS (deg) (ft/s) (g) (ft) (ft/s) (deg) (ft)
120 –3 –10.61 1.25 5079 –3 –0.85 3.6
Normal

120 –3 –10.61 1.2 6340 –3 –0.85 4.5

120 –3 –10.61 1.15 8435 –3 –0.85 5.9
120 –3 –10.61 1.1 12595 –3 –0.85 8.9
120 –5.5 –19.42 1.25 5015 –3 –0.85 16.5
Steep

120 –5.5 –19.42 1.2 6240 –3 –0.85 20.6

120 –5.5 –19.42 1.15 8258 –3 –0.85 27.2
120 –5.5 –19.42 1.1 12206 –3 –0.85 40.2
120 –7.5 –26.45 1.25 4938 –3 –0.85 33.2
Abuse

120 –7.5 –26.45 1.2 6122 –3 –0.85 41.2

120 –7.5 –26.45 1.15 8053 –3 –0.85 54.2
120 –7.5 –26.45 1.1 11762 –3 –0.85 79.2

Fig. 20.34 Flare height vs approach angle and load factor in the flare.
 CHAPTER 20 Landing Performance Testing and AFM 991

the 2-deg abuse condition ( –7.5-deg approach). Note how large the sink rate
is at 50 ft AGL. For the normal approach, barely any flare is required to
reduce the sink rate below the structural limit of 10 ft/s; it is not that case
for the abuse condition.
Of course, the airplane cannot increase the load factor instantaneously.
Instead, the curved path will be a combination of peak load factor achieved
and the g-onset rate (how fast the peak load is achieved). One more thing
is certain: For the steep approach angle, the flight path will more often be
a single curved path as the pilot focuses on reducing the sink rate prior to
touchdown. On conventional airplanes (reversible or irreversible flight con-
trols), the g-onset rate depends on the pilot’s ability to deflect the elevator, the
elevator power, and the CG location (forward CG being most adverse), so this
value is typically large (0.4 to 0.6 g/s), but the pilot is responsible for main-
taining a margin to stall warning (1.23 VSR typically providing 1.3-g capa-
bility) in the flare. One should expect lower peak load factor of the order
of 1.15 g to 1.25 g at best. For fly-by-wire (FBW), the airplane’s control
laws (CLAWS) may be capable of going to the peak available g (the airplane
is protected), but the CLAWS logic may result in a lower g-onset rate. We
offer the following charts to illustrate the impact of the maneuver execution
on the flare height. Figure 20.35 could represent a conventional airplane with
an approach speed of 125 KTAS on a 25.5-deg approach. The airplane is

1600 10
9
8
7
1400 Pitch 6
5
4
3
1200 Targets: 2 Pitch (deg), sink rate (ft/s)
1
Sink rate TD = 6 ft/s 0
–1
1000 Peak g = 1.15 –2
Distance (ft)

–3
g-onset = 0.6 g/s
hpull = 40.2 ft

–4
–5
800 –6
Air distance = 792 ft –7
–8
Airtime = 3.79 s –9
600 –10
–11
–12
0.45 s –13
400 –14
–15
–16
SVREF = 125 KTAS –17
200 –18
Approach = –5.5 deg –19
–20
–21
0 –22
0 10 20 30 40 50 60
Height (ft)

X Straight path SR Pitch

Fig. 20.35 Steep approach flare with large g-onset rate, average peak g.
992 Operational Aircraft Performance and Flight Test Practices

1600 10
9
8
Pitch 7
1400 6
5
4
Targets: 3
1200 2

Pitch (deg), sink rate (ft/s)

Sink rate TD = 6 ft/s 1
0
Peak g = 1.2 –1
1000 g-onset = 0.2 g/s –2
Distance (ft)

–3

hpull =38.8 ft
–4
–5
800 –6
Air distance = 728 ft –7
–8
Airtime = 3.48 s –9
600 –10
–11
–12
0.55 s –13
400 –14
–15
–16
SVREF = 125 KTAS –17
200 Approach = –5.5 deg –18
–19
–20
–21
0 –22
0 10 20 30 40 50 60
Height (ft)
X Straight path SR Pitch

Fig. 20.36 Larger peak load factor, but lower g-onset rate.

capable of a 0.6-g/s g-onset, and the pilot limits the pitch increase to an
equivalent 1.15-g peak load factor. The airplane touchdown rate target is
6 ft/s (a performance landing). To achieve this maneuver, the pilot
must initiate the flare at a height of 40.2 ft, about 0.45 s after crossing
the threshold.
The maneuver is repeated with the same airplane, now equipped with an
FBW system. The airplane can achieve at least 1.2 g, but the g-onset rate is
limited to 0.2 g/s. This does provide the pilot a little more reaction time
(0.55 s vs 0.45 s for the conventional plane) due to the higher load factor
achieved, but the airplane sink rate does not reduce below 10 ft/s until the
airplane is at about 7 ft above the ground (see Fig. 20.36).
The low onset rate does not afford flexibility in reaction time, especially if
the airplane deviates from its nominal approach angle. We offer Fig. 20.37 as
an example that repeats the conditions of Fig. 20.36, but with an approach
path only 0.5-deg steeper. To ensure the airplane touchdown rate has the
same target, the pilot would have had to flare only 0.2 s after crossing the
threshold altitude.
One can see that the airplane in Fig. 20.37 approaching the landing with
the abuse case of a 27.5-deg approach angle would be unable to flare at or
below 50 ft and reduce the sink rate at touchdown to less than the
maximum 10-ft/s structural limit, let alone meet the certification require-
ments of less than 3 ft/s at touchdown. This author successfully argued to
 CHAPTER 20 Landing Performance Testing and AFM 993

1600 10
9
8
Pitch 7
1400 6
5
4
Targets: 3
1200 2
Sink rate TD = 6 ft/s 1

Pitch (deg), sink rate (ft/s)

0
Peak g = 1.2 –1
1000 g-onset = 0.2 g/s –2
Distance (ft)

–3

hpull = 45.5 ft
–4
–5
800 –6
Air distance = 736 ft –7
–8
Airtime = 3.52 s –9
600 –10
–11
–12
0.2 s –13
400 –14
–15
–16
SVREF = 125 KTAS –17
200 –18
Approach = –6.0 deg –19
–20
–21
0 –22
0 10 20 30 40 50 60
Height (ft)
X Straight path SR Pitch

Fig. 20.37 Impact of 0.5-deg steeper approach angle to conditions of Fig. 20.36.

the JAA during the Learjet 45 certification that, at the increased sink rates
seen in the abuse conditions of 27.5 deg and selected SVREF, the much
increased sink rate (near 30 ft/s) on approach would be a sufficient incentive
for the crew to start flaring higher than the threshold of 50 ft and that the
flare height could be 1.5 times threshold height (75 ft). Indeed, some auth-
ority pilots have said in the past that even the “normal” approach angle
(–5.5 deg) was a rapid descent, and they had a clear tendency to want to
reduce the FPA prior to the threshold.
The TCCA AC requests that the abuse case flare be done at the threshold
height but has allowed the higher flare height by issue papers provided
higher density altitudes (higher true airspeed on approach) be tested.
The FAA AC 25-7D stipulates that, for flight test safety reasons, when
conducting the 2-deg-steeper approach path angle test condition, the pilot
may begin to flare the airplane (or reduce the approach angle) at a reasonable
height somewhat higher than the normal steep approach flare height. If this is
done, it should be shown by analysis that there is sufficient pitch control to
arrest the descent rate if the flare were to be initiated at the normal steep
approach flare height, keeping in mind the criteria of less than 3 ft/s at
touchdown.
This author finds the authorities’ focus on the 2-deg-steeper approach
abuse case interesting and believes that the focus should really be placed
more on providing proper flight path guidance (like a HUD) to the pilot
994 Operational Aircraft Performance and Flight Test Practices

rather than asking for this significant flight path capability margin. If one
observes Fig. 19.2, it could be understood that the FPA could deviate some-
what because the airplane is far from the runway and the higher sink rate
would offer less capability to the crew; however, having guidance in the
form of a HUD with FPA throughout the approach, for example, would
give the pilot time to adjust the FPA (put the FPA marker on the landing
point on the runway) and track trends in deviation from the target. It
would result in a smaller overall deviation and remove the need to be able
to flare this much. A lower margin of 1-deg steeper would then be sufficient.
Most jet airplanes do not have the 22-deg margin for normal landing,
especially not for angles closer to the 24.5-deg limit of the normal range.
Also, airplanes with slower approach speeds may not need as much flare
capability, although they are more impacted by wind variation. Consider
that the sink rate on approach for an airplane with a VREF of 141 KTAS at
a 23.5-deg approach angle would be 14.5 ft/s and would require a flare
height of 22 ft (for a target touchdown sink rate of 6 ft/s) under a 1.15-g
peak load factor condition. Another airplane approaching at a VREF of 85
KTAS at 25.5 deg would have a sink rate of 13.8 ft/s and would require a
flare height of 19 ft under the same peak load factor of 1.15 g, yet this
second airplane must also demonstrate a descent capability of 2-deg
steeper. More improvement is possible in the steep approach requirements.
EASA requirements state that one may adjust threshold height down to
35 ft (improves landing distance) or increase it up to 60 ft AGL (improves
flare capability). TCCA does not support a screen height higher than 50 ft,
stating in its AC: “Screen heights greater than 50 feet are not allowed. . .
Transport Canada considers that an aircraft must be able to safely conduct
an approach to landing using this criteria for any glidepath.”

Performance Landings
Once the handling capability of the airplane has been successfully
demonstrated at the higher approach angle, one can proceed with
“normal” performance landing testing. The regulations for steep approach
allow the use of parametric landings analysis with some exceptions—the
expansion must be done at the nominal approach angle (not nominal
minus 0.5 deg, i.e., not steeper than nominal) and no more than 6-ft/s touch-
down speed for expansion. EASA, the baseline authority for steep approach,
requires the following (note how similar the words are to §25.125, yet they
include additional requirements for the steep approach):
EASA (SAL) 25.3 Steep Approach Landing Distance
(Applicable only if a reduced landing distance is sought, or if the landing
procedure (speed, configuration, etc.) differs significantly from normal oper-
ation, or if the screen height is greater than 50 ft.)
(a) The steep approach landing distance is the horizontal distance necessary
to land and to come to a complete stop from the landing screen height
 CHAPTER 20 Landing Performance Testing and AFM 995

and must be determined (for standard temperatures, at each weight,

altitude and wind within the operational limits established by the
applicant for the aeroplane) as follows:
(1) The aeroplane must be in the all-engines-operating or
one-engine-inoperative steep approach landing configuration,
as applicable.
(2) A stabilised approach, with a calibrated airspeed of VREF(SAL) or
VREF(SAL)-1 as appropriate, and at the selected approach angle
must be maintained down to the screen height.
(3) Changes in configuration, power or thrust, and speed must be made
in accordance with the established procedures for service operation
(see AMC 25.125(b)(3)).
(4) The landing must be made without excessive vertical acceleration,
tendency to bounce, nose over or ground loop and with a vertical
touchdown velocity not greater than 6 ft/sec.
(5) The landings may not require exceptional piloting skill or alertness.
(b) The landing distance must be determined on a level, smooth, dry,
hard-surfaced runway (see AMC 25.125(c)). In addition,
(1) The pressures on the wheel braking systems may not exceed those
specified by the brake manufacturer;
(2) The brakes may not be used so as to cause excessive wear of brakes or
tyres (see AMC 25.125(c)(2)); and
(3) Means other than wheel brakes may be used if that means
(i) Is safe and reliable;
(ii) Is used so that consistent results can be expected in service; and
(iii) Is such that exceptional skill is not required to control
the aeroplane.
(c) Reserved.
(d) Reserved.
(e) The landing distance data must include correction factors for not more
than 50 % of the nominal wind components along the landing path
opposite to the direction of landing, and not less than 150 % of the
nominal wind components along the landing path in the direction
of landing.
( f) If any device is used that depends on the operation of any engine, and if
the landing distance would be noticeably increased when a landing is
made with that engine assumed to fail during the final stages of an
all-engines-operating steep approach, the steep approach landing
distance must be determined with that engine inoperative unless the use
of compensating means will result in a landing distance not more than
that with each engine operating.

The authorities have limited the landing at steep approach airports for air-
planes capable of completing the landing within 60% of the runway (LFL).
EASA and other authorities stipulate that the manufacturer does not
need to change the landing air distance model, but the steep approach
has a potential to reduce the total landing distance by a few hundred feet.
Figure 20.38 shows distance based on the approach angle with no flare.
This, combined with the requirement to land within 60% of the landing dis-
tance available (LDA) at London City, makes the difference between an
acceptable payload and sometimes no landing capability, so most manufac-
turers will go for the reduced air distance testing and performance
model update.
996 Operational Aircraft Performance and Flight Test Practices

SVREF SVREF

3 deg 5.5 deg

50 ft 50 ft

955 ft 522 ft

Fig. 20.38 Potential improvement in air distance for steep approach.

Because of the LFL requirements imposed for steep approach landing

models, the use of parametric analysis has been deemed acceptable. One
thing that comes out of the model for approaches of 25.5 deg is that the
air distance may actually reduce slightly as the airplane weight and approach
speed increase; the steeper approach angle and the essentially fixed flare
height (near 50 ft AGL) lead to this result.

Flight Test Risks and Mitigations

Even when one gets used to the higher sink rate on approach, steep
approach testing always involves greater test risk than normal landing tests.
The most important risks associated with steep approach testing include:
• High sink rate on touchdown
• Failure of wheels, tires, and brakes
• Structural damage
• Simulated OEI conditions with possible loss of directional control and/or
loss of thrust on operating engines
• Stall warning activation and possible stall in the flare
• Elevator limited in the flare
The mitigations for these risks are:
• Perform buildups in flight path angle (from low to high) and flare height
(from high to low) at mid-CG, and then proceed to forward CG.
• Perform SVREF tests before SVREF minus 5 tests.
• Monitor wheel, tire, and brake temperatures prior to a test and the energy
required for the following test.
• Use a runway of sufficient length.
• Move ground equipment a safe distance from the runway centerline.
• Perform AEO testing prior to OEI testing.
• Have flight crew establish test abort criteria (off glide path, speed, stabilized
approach, etc.).
• Mandate the crew wearing protective equipment.
• Define criteria for stable approach.
 CHAPTER 20 Landing Performance Testing and AFM 997

AAIB Bulletin: 8/2008, Avro RJ100, HB-IYU, 18 Aug. 2007, London City
Airport: The commander was carrying out an ILS approach to Runway 28 at
London City Airport, with the approach stabilised from the glideslope capture
at 3,000 ft. At between 50 and 30 ft above the runway the pilots felt the aircraft
“dropping” and the commander, who was the pilot flying, pulled back on the
control column to prevent a hard landing. The pitch attitude of the aircraft
increased to a maximum of 9.38 and the lower aft fuselage briefly contacted
the runway before MLG contact, causing significant damage.
GROUND SPEED (kt)

130 SWITCH TO TAILWIND Vref + 5

120 Vref
450
GUSTS Vref – 7
CAS (kt)

110 20 400
PITCH (° +ve nose up)

AIRSPEED AND ALTITUDE LESS

AOA (°)

100 10 350 ACCURATE AT HIGH AOA

90 0 300
NORMAL ACCEL (g)

10 –10 250
9.3 DEGREE PEAK
PRESSURE ALTITUDE (ft)

0 3 200
RADIO ALTITUDE (ft)

2.3 g
–10 2 150

1 100

0 50

0
SQUAT SWITCHES LEFT RIGHT NOSE
09:39:37 09:39:41 09:39:45 09:39:49 09:39:53 09:39:57 09:40:01
UTC (2 seconds per square)

Transport Canada Steeper Requirements

Transport Canada imposes an additional flight condition, the steeper
approach. This approach is one that falls between 23.5 deg and 24.5 deg,
inclusive. (Author’s note: there is often a discussion on what type of approach
to use when an airport calls for an approach exactly at 24.5 deg, steep or
steeper. It does not matter because the applicant will have demonstrated
the approach using both the steeper and steep approach procedures.)
For this approach, Transport Canada recognizes that the baseline certifi-
cation requirements of Part 25 are sufficient for the landing configuration;
thus, no additional testing is required in that sense. The requirements for
steeper (a Transport Canada working note), however, do require that the
applicant demonstrate that the airplane is capable of at least a 1-deg-steeper
approach path than the nominal value to be certified (the minimum margin),
something that is not required for normal approaches. Thus, for a certifica-
tion to an approach angle of 24.5 deg, the airplane must be capable of a flight
path angle of at least 25.5 deg in the selected approach configuration and fol-
lowing the desired approach airspeed, under the most adverse weight and
center of gravity to be certified.

Reference
[1] Asselin, M.Hinson, M.Storrer, B., “Learjet Model 45 Steep Approach Certification.”
SAE 2004-01-1805, 2004.
 ABOUT THE AUTHOR
Mario Asselin is currently Chairman of Asselin, Inc. (www.asselininc
.com) and a FAA Flight Analyst DER (Design Engineering Representative)
for Part 23 Small Airplanes and Part 25 Transport Category Airplanes. Mr.
Asselin also holds concurrently the position of Engineering Fellow at Bom-
bardier Aviation, a Transport Canada DAD (Design Approval Designee),
Learjet ODA (Organization Designation Authorization) Unit Member
(UM), and Chairman of the Safety Review Board at the Bombardier
Flight Test Center. He is a member of the Manufacturers Flight Test
Council (MFTC), of GAMA EPIC (Electric Propulsion and Innovation
Committee), of the Manufacturers Icing Certification Group (MICG),
and of the Electric Flight Test Committee. Prior to this, he was Director
Engineering, Operations, and Quality at the Bombardier Flight Test
Center (BFTC) in Wichita, Kansas, where he oversaw the Flight Test
Center Engineering, Operations, Safety, Quality, and Corporate Shuttle
activities. Mr. Asselin also held a position of Senior Manager for Flight
Sciences and Flight Test Engineering & Operations at Honda Aircraft Cor-
poration and was Vice President Engineering at Sino Swearingen Aircraft
Corporation, now known as Syberjet Aircraft, where he oversaw the
SJ30-2 flight testing and certification. Prior to this, Mario was Learjet
Chief of Stability and Control at BFTC, and Chief Technical Engineer for
the aerodynamic design and certification of Bombardier’s CRJ-900 (now
an MHI product). Mario’s early career included 14 years in the Canadian
Air Force and a short stay at CAE’s military division.
Mr. Asselin has taught courses in aircraft performance, stability and
control, and aerodynamics for the Royal Military College of Canada in
Kingston, Ontario. He has also taught in Montreal at McGill University,
École de Technologie Supérieure (ETS) and Concordia University of Mon-
treal. He has been an instructor for Kansas University aerospace short
courses since 2004 where he teaches “Operational Aircraft Performance
and Flight Test Practices” and monitored “Airplane Performance: Theory,
Applications and Certification.” He holds a Baccalaureate of Engineering
degree in mechanical engineering from the Royal Military College of

999
1000 Operational Aircraft Performance and Flight Test Practices

Canada and a Master of Applied Science (M.Sc.A.) in aerothermodynamics

(aircraft icing) from École Polytechnique of Montreal.
Building on this experience, Mario has also contributed to the pro-
gression of multiple projects including the Heart Aerospace (www
.heartaerospace.com) ES-19 all electric regional airplane and will also par-
ticipate in the preparation of the certification basis for the TranscendAir
(www.transcend.aero) Vy400 tilt-wing aircraft and the JetPack Aviation
SpeederTM .
INDEX
absolute ceiling, aircraft, 532 airplane configuration, 125
abused takeoff testing, 797–798 airplane flight manual (AFM), 138
accuracy, 123 airplane geometry, 171–175
actual landing distance (ALD), 965–971, airfoil shapes, 171–172
973–982 wing plan form, 172–175
actuator disk theory, 336–349 airplane load factor, 515
ADCs. See Air data computers air segment, landing, 893, 896–924
AEO. See All engines operating eye-to-MLG sink rate management,
aerodynamic compensation, 121 919–920
aerodynamic stall, 186–192 flare, 911–915, 938–940
high-altitude stall, 189–192 float, tendency to, 920–921
on propeller-driven planes, 187–188 ground effects, 905–911
air, 1 steady approach, need for, 923–924
density, 3 testing, 936–959
pressure, 1–2 airtime modeling, 953–958
temperature, 2–3 approach airspeed validation, 943
viscosity, 3–4 and data reduction, 943–953
air data computers (ADCs), 21, 106–107, flare technique, 938–940
121, 476 instrumentation required, 958–959
air data errors, flight testing for, 72–114 landing technique development,
free stream static-pressure calibration 936–938
method, 89–109 mathematical upper bound, 945–946
alternative tower fly-by, 98 parameters, 940–943
chase plane, 92 parametric landings, 948–953
display lag, 105–107 risks and mitigation, 959
experimental system, lag in, 101–105 sink rate, 948
nose boom, 107–109 six performance landings, 946–948
pacer airplane, 89–92 thrust cut height impact, 900–904
tower fly-by, 92–98 thrust reversers in flight, use of, 921–923
trailing cone, 98–101 total air distance, 921–923
position errors, source of, 73–74 touchdown, 915–919
system leak, 109–110 airspeed, 39–66
total temperature probe calibration, altitude effects (TAS vs. M), 44
110–114 calibrated, 53–60
true airspeed calibration method, 74–89 vs. EAS, 57–60
closed course method, 74–83 vs. Mach, 57
GPS method, 83–89 vs. TAS, 60–61
airfoil shapes, 171–172 equivalent, 52–53

1001
1002 Operational Aircraft Performance and Flight Test Practices

airspeed (Continued ) RVSM space, 29–30

TAS vs. EAS, 53 temperature variation with, 17–18
IAS, 67–70 angle of attack display in cockpit, 192–193
indicators, 61–66 anti-icing system, 206
KCAS, 62–63 antiskid systems, 837–838
KTAS, 63–64 approach ice, 199
Machmeter, 63 ASRs. See Altimeter setting regions
Mach number and, 41–43, 50–52 ATC. See Air traffic control
speed of sound and, 40–41 atmosphere, 1
TAS, 40–43 layers, 4–5
temperature effects (TAS vs. M), 44–46 properties, 1–5, 10–11
total air temperature, 50–52 standard, 5–11
total pressure, 48–50 geopotential vs. geometric altitude,
airtime modeling, 953–958 5–6
air traffic control (ATC), 412 modeling, 6–7
ALD. See Actual landing distance stratosphere, 9–10
all engines operating (AEO), 372–374, troposphere, 7–9
526, 547–548, 564, 573, 590, atmospheric ratios, 11–14
593, 601, 605
climb requirements, 595 balanced field length (BFL), 879–880
takeoff performance barometric altimeter, 21–25
acceleration from VLOF to V35, 698–700 barometric setting, 21–25
acceleration from VR to VLOF, battery-powered airplanes, 500–502
695–698 BFL. See Balanced field length
forces, 685–691 BLI. See Boundary layer ingestion
ground run, 681–707 Bonanza V-35, 521
minimum unstick speed VMU, 693–694 boundary layer ingestion (BLI), 351–352
MLG and NLG compression, braking force, 832, 834–838
682–683 braking performance, 849
pitch attitude, target, 704–707 Brayton cycle, 322
pitch rate, target, 702–704 Bureau d’Enquêtes et d’Analyses
rolling friction force, 684–685 (BEA), 244
rotation speed (VR), 691–692 bypass ratio, 328–329
runway gradient, 692–693
takeoff, 680 calibrated airspeed (CAS), 53–60, 67
takeoff safety speed V2, 700–702 vs. EAS, 57–60
trim setting, 707 vs. Mach, 57
alternative tower fly-by, 98 vs. TAS, 60–61, 67
altimeter setting regions (ASRs), 26–27 carbon dioxide (CO2), 321, 333
altitude in aviation, 13, 15, 17–18, 67 cargo and checked baggage, 155–156
cruising, 28–29 CAS. See Calibrated airspeed
GPS, 37 center of gravity (CG), airplane, 137–146
pressure, 18–35 envelopes, 156
barometric altimeter, 21–25 fuel weight and, 147–153
barometric setting, 21–25 lateral, 143–145
measurement, 19–21 vertical, 145–146
operational regulations, 25–28 certification regulations, 68–70
temperature effects on measurement, Cessna 172, 521
30–35 chase plane, 92
radar, 35–37 clearway, 722–725
 Index 1003

climb gradient (grad), 518, 578 hydrogen plane, 462–465

climb performance model, 513–569 parameters, 412–413
angles for, 513–515 payload-range diagram, 447–449
certification requirements, 547–550 performance triangle, 414
climb gradient (grad), 518 range, 411–412
data reduction, 551 reciprocating/turboprop, range
descending flight, 538–539 equation, 437–440
development, 513–526 specific air range, 414–426
energy height, 544–547 airspeed on, 416–421
example, 517 altitude on, 421–426
flight crew, information to, 564–569 turbojet/turbofans, range equation,
flight path angle 426–437
climb gradient vs., 518–520 cruise altitude restrictions, 433–434
vs. specific excess thrust, 517 optimum cruise altitude, 435–437
flight path variation, 539–544 stepped climb, 434–435
forces for, 513–515 wind on range, impact of, 444–446
influencing factors, 527–537 airspeed for, 446–447
acceleration, 534–535 cruise performance triangle, 414
altitude, 529–534 cruising altitude, 28–29
climb ceilings, 532–534
drag, 528 deep stall, 263
thrust, 528 deicing boots, 207–210
weight, 527–528 density, 3
winds, 535–537 descending flight, 538–539
load factor, 539–544 DGPS. See Differential global positioning
maximum climb angle, 518–522 system
rate of climb, 522–526 DGPS runway survey analysis method,
schedules, 538 772–775
testing for, 551 diesel, 321
acceleration in climb and winds, 564 differential global positioning system
data analysis and corrections, (DGPS), 940
563–564 dipping, 150
geometric vs. pressure altitude, 563 display lag, 105–107
instrumentation, 562 drag model, 289–313
level flight acceleration, 559–560 for basic performance assessment,
sawtooth climb test, 551–558 298–299
validation climb test, 560–562 flight crew, presenting information to,
T/W and Em values, 516 502–507
closed course method, 74–83 cruise altitude selection, 503
coffin corner, 392 reserve fuel planning, 505–507
combat ceiling, aircraft, 533 SAR charts, 503–505
contaminated runway, 726–730 simplified vs detailed flight
cruise of airplanes planning, 507
drag model testing, 467–511 (See also from flight test results, 467–468
Drag model) fuel (energy) management/
electrical propulsion, 451–461 conservation, 507–511
flight endurance, 449–451 aerodynamic cleanliness, 511
fuel reserves, 440–444 ground operation, 510–511
fuel-weight fraction, 412–413 high-speed, testing for, 487–493
hybrid-electric, 461–462 monitoring, 492
1004 Operational Aircraft Performance and Flight Test Practices

drag model (Continued) advantages, 333–334

Reynolds number effects, 492–493 carbon dioxide, impact on, 335–336
scope, 490–491 efficiency, 333–334
verification, 491–492 environment, impact on, 335–336
low-speed, testing for, 467–487 electro-expulsive separation system
acceleration correction, 479–480 (EESS), 210–211
air data correction, 476 Electro-Mechanical Expulsion Deicing
altitude variation correction, 477–479 System (EMEDS), 210–211
center of gravity location, 476 EMAS. See Engineered materials arresting
corrected test weight, 474–476 systems
drag polar fitting, 480–486 EMEDS. See Electro-Mechanical Expulsion
instrumentation, 472 Deicing System
monitoring, 472–474 empty weight, 133
posttest, 474 energy height, 544–547
power effects, 486–487 energy maneuverability, 649–654
preparation, 471 engine compression ratio, 315
scope/purpose, 471 engineered materials arresting systems
minimum power required condition, (EMAS), 846–847
310–311 engine out departure procedures
power required for flight, 310 (EODPs), 610
SAR, testing for, 493–499 en route ice, 198
on electric airplanes, 496–497 EODPs. See Engine out departure
instrumentation for, 499 procedures
integrated model verification, 497 equivalent airspeed (EAS), 52–53
power leakage, 495–496 CAS vs., 57–60
secondary effects, 495–496 TAS vs., 53
validation flight, 497–499 ETM. See Excess thrust manager
slow flight speed domain, 311–313 ETOPS. See Extended-Range Twin-Engine
sources, 289–298 Operational Performance
excrescence drag, 291–292 Standards
skin friction drag, 291 excess thrust manager (ETM), 352–353
trim drag, 295–296 excrescence drag, 291–292
vortex drag, 292–295 Extended-Range Twin-Engine Operational
wave drag, 296–298 Performance Standards
steady level flight assumption, 299–310 (ETOPS), 571
aerodynamic efficiency, 300–301 eye-to-MLG sink rate management,
airspeed/drag coefficients/weight, 919–920
drag variation with, 303–310
minimum drag condition, 301–303 FAA 14 CFR/EASA CS/TCCA CAR Part
tree, 290 23, 237–241
zero fuel level, testing for, 499–502 failure ice, 200
on battery-powered airplanes, FARs. See Federal Aviation Regulations
500–502 fastest turn, 640–643
drag rise Mach number, 297 FBW. See Fly-by-wire
ducted fan, 347–349 Federal Aviation Regulations (FARs), 68
ferry range, of airplane, 448
EAS. See Equivalent airspeed FIKI. See Flight into known icing
EESS. See Electro-expulsive final takeoff ice, 198
separation system flaps out envelope, 401–403
electric motors engines, 333–336 flare technique, 911–915, 938–940
 Index 1005

flight envelope, 355–410 experimental system, lag in, 101–105

flaps out and gear down envelopes, nose boom, 107–109
401–403 pacer airplane, 89–92
flaps down airspeeds and altitude, tower fly-by, 92–98
401–403 trailing cone, 98–101
high-speed limit, 386–398 position errors, source of, 73–74
buffet, 388–392 system leak, 109–110
characteristics, 394–395 total temperature probe calibration,
design dive speed, 392–394 110–114
flutter speed, 397–398 true airspeed calibration method, 74–89
maximum operating airspeed/Mach, closed course method, 74–83
387–388 GPS method, 83–89
out-of-trim characteristics, 395–396 float, tendency to, 920–921
requirements, 398–399 fly-by-wire (FBW) airplanes with envelope
reversal speed, 397–398 protection, 242–244, 991
low-speed limit, 357–386 flying test beds (FTBs), 890–891
buffet boundary, 362–363 free stream static-pressure calibration
crosswinds, 383–384 method, 89–109
maneuvering speed VA, 380–383 air data lag, 103
maneuver margin, alternative tower fly-by, 98
360–362 chase plane, 92
minimum control speed, 366–377 display lag, 105–107
air VMCA, 366–368 experimental system, lag in, 101–105
dynamic VMCA, 368–370 nose boom, 107–109
ground VMCG, 374–377 pacer airplane, 89–92
landing VMCL, 372–374 tower fly-by, 92–98
one-engine-inoperative speed VSSE, trailing cone, 98–102
371–372 freight, 155
tailwinds, 384–386 FSs. See Fuselage stations
takeoff/landing speed restrictions, FTBs. See Flying test beds
362–366 FTHWG. See Flight Test Harmonization
trim, 378–380 Working Group
maximum altitude, 398–400 fuel weight, 147–153
placard speeds, 404 fuel-weight fraction, 412–413
regulation-based limit, 355–357 fuselage stations (FSs), 140
RVSM, 404–410
temperature envelope, 400–401 gas turbines, 321–333
flight into known icing (FIKI), 204 carbon dioxide, impact on, 333
flight manual, 125–126 efficiency, 323–324
flight path variation, 539–544 engine, evolution of, 324
flight test equipment/ballast, 156–160 environment, impact on, 333
Flight Test Harmonization Working maximum thrust/power available vs.
Group (FTHWG), 242 altitude, 332–333
flight testing, 156–160 principle of operation, 322
flight testing for air data errors, 72–114 turbofan cycle, 328–330
free stream static-pressure calibration turbojet cycle, 325–328
method, 89–109 turboprop cycle, 330–332
alternative tower fly-by, 98 turboshaft cycle, 332
chase plane, 92 gear down envelope, 401–403
display lag, 105–107 geopotential vs. geometric altitude, 5–6
1006 Operational Aircraft Performance and Flight Test Practices

glaze ice, 194–195 hydrogen plane, 462–465

glide path, 989 IAS. See Indicated airspeed
Global Positioning System (GPS), 37, ICAO. See International Civil Aviation
83–89 Organization
altitude, 37 ice-contaminated tailplane stall (ICTS), 201
course vs. speed course, 88–89 ice protection systems (IPS), 204–211
example, 87–88 anti-icing system, 206
four-leg course, 86–87 deicing system, 207–211
horseshoe course method, 85 deicing boots, 207–210
method–induced error, 85 electro-expulsive systems, 210–211
two-leg method, 85 icing, 193–204
GPS. See Global Positioning System approach, 199
Ground Course Test Card, 79 behaviors, 194
ground run distance modeling/testing, conditions, 195–199
769–796 en route, 198
DGPS runway survey analysis method, failure, 200
772–775 final takeoff, 198
instrumentation, 795–796 holding, 198–199
OEI testing, 786–792 ICTS, 201
procedure, 779–780 intercycle, 200–201
rapid rotation, 794–795 IPS, 204–211
rolling friction force coefficient, 776–777 landing, 199–200
runway survey, 770–772 preactivation ice, 195–198
speed spread model, 792–794 ridge, 200–203
thrust, modeling excess, 775–780 stall in, 244–245
VLOF to V35, 783–786 takeoff, 198
VR to V35, 780–781 types, 194–195
VR to VLOF, 781–783 ICTS. See Ice-contaminated tailplane stall
wind considerations, 777–779 impact pressure, 1
ground run segment, landing, 929, vs. dynamic pressure, 46–48
964–965 indicated airspeed (IAS), 67–70, 128–129
inner wing stall in turn, 654–656
head-up display (HUD), 940 instrument error, 72
high-speed drag model testing, 487–493 intercycle ice, 200–201
monitoring, 492 International Civil Aviation Organization
Reynolds number effects, 492–493 (ICAO), 31
scope, 490–491 IPS. See Ice protection systems
verification, 491–492
high-speed limit, flight envelope, 386–398 KCAS. See Knots calibrated airspeed
buffet, 388–392 knots calibrated airspeed (KCAS), 57–61,
characteristics, 394–395 71–72, 562
design dive speed, 392–394 KTAS, 419
flutter speed, 397–398 measures, 62–63
maximum operating airspeed/Mach, knots (kt), 40, 42
387–388 knots true airspeed (KTAS), 40–42, 44–46,
out-of-trim characteristics, 395–396 54–56, 63–64, 114, 419–422
requirements, 398–399 vs. KCAS, 60–61
reversal speed, 397–398 vs. KEAS, 53–54
holding ice, 198–199 measures, 63–66
HUD. See Head-up display knots true ground speed (KTGS), 480, 551
 Index 1007

KTAS. See Knots true airspeed for ground run segment, 964–965
KTGS. See Knots true ground speed risks and mitigations, 996–997
landing, 893–932 steep approaches, 986–997
air segment, 893, 896–924 flare capability, 990–994
eye-to-MLG sink rate management, flight path capability and abuse case,
919–920 988–994
flare, 911–915, 938–940 London City Airport, 986–987
float, tendency to, 920–921 performance landings, 994–996
ground effects, 905–911 regulations and guidance material,
steady approach, need for, 923–924 987–988
testing, 936–959 risks and mitigations, 996–997
thrust cut height impact, 900–904 time of arrival, 982–986
thrust reversers in flight, use of, for total landing distance, 965–971
921–923 coming in fast, 966–968
total air distance, 921–923 coming in high, 968–970
touchdown, 915–919 smooth landing, 970–971
basic modeling, 893–896 for transition segment, 959–964
full braking segment, 894 Transport Canada Steeper
ground run segment, 929, 964–965 requirements, 997
operational requirements, 930–932 for wheels, 965
parametric, 948–953 lateral CG, 143–145
testing, 935–997 level flight acceleration, 559–560
total landing distance, 929–930, 965–971 lift, 176–177
transition segment, 893–894, 924–928 aerodynamic perspective, 211–215
events, 927–928 aerodynamic stall, 186–192
ground distance, 927–928 high-altitude stall, 189–192
testing, 959–964 on propeller-driven planes, 187–188
landing ice, 199–200 angle of attack display in cockpit, 192–193
landing performance testing, 935–997 curve and distribution, 177–186
for air segment, 936–959 configuration effects, 183–185
airtime modeling, 953–958 lift vs. weight coefficients, 186
approach airspeed validation, 943 required lift coefficient vs. airspeed,
and data reduction, 943–953 183–185
flare technique, 938–940 equation, 176–177
instrumentation required, 958–959 icing, 193–204
landing technique development, approach, 199
936–938 behaviors, 194
mathematical upper bound, 945–946 conditions, 195–199
parameters, 940–943 en route, 198
parametric landings, 948–953 failure, 200
risks and mitigation, 959 final takeoff, 198
sink rate, 948 holding, 198–199
six performance landings, 946–948 ICTS, 201
for brakes, 965 intercycle, 200–201
distance test campaign, preparing for, IPS, 204–211
935–936 landing, 199–200
flight crew, information to, 971–982 preactivation ice, 195–198
approach and landing airspeeds, 973 ridge, 200–203
landing distances, 973–982 takeoff, 198
landing procedures, 971–973 types, 194–195
1008 Operational Aircraft Performance and Flight Test Practices

lift (Continued) maximum ROC, 525–526

manufacturing tolerance, 211–215 maximum take-off weight (MTOW), 133
vs. weight coefficients, 186 maximum zero fuel weight (MZFW), 133
load factor, airplane, 515 mean geometry chord (MGC), 173
London City Airport (LCY), 986–987 mean sea level pressure (MSLP), 22
low-speed drag model testing, 467–487 MGC. See Mean geometry chord
acceleration correction, 479–480 MIL-F-8785C, 241–242
air data correction, 476 minimum control speed, air VMCA testing,
altitude variation correction, 477–479 750–761
center of gravity location, 476 dynamic, 758–760
corrected test weight, 474–476 p-factor, 750, 752
drag polar fitting, 480–486 risks and mitigations, 760–761
instrumentation, 472 rudder, bank angle effect on, 757
monitoring, 472–474 static, 754–755
posttest, 474 thrust asymmetry, 751
power effects, 486–487 wing lift/tail aerodynamics, prop wash
preparation, 471 impact on, 753–754
scope/purpose, 471 minimum unstick speed VMU testing,
low-speed limit, flight envelope, 357–386 761–769
buffet boundary, 362–363 analysis, 766–767
crosswinds, 383–384 data, 765–766
maneuvering speed VA, 380–383 example, 767–769
maneuver margin, 360–362 geometry, 764
minimum control speed, 366–377 pitch angle, 762
air VMCA, 366–368 mixed-phase ice, 194–195
dynamic VMCA, 368–370 MLG. See Main landing gear
ground VMCG, 374–377 MLW. See Maximum landing weight
landing VMCL, 372–374 MRW. See Maximum ramp weight
one-engine-inoperative speed VSSE, MSLP. See Mean sea level pressure
371–372 MTOW. See Maximum take-off weight
tailwinds, 384–386 MZFW. See Maximum zero fuel weight
takeoff/landing speed restrictions,
362–366 NASA Quiet Short-Haul Research Aircraft
trim, 378–380 (QSRA), 487
National Business Airplane Association
Mach, Ernst, 41 (NBAA) range, 412
Machmeter, 63 National Transportation Safety Board
Mach number, 41–43, 297, 487–488 (NTSB) recommendations,
and TAT, 50–52 244–245
main landing gear (MLG), 682–683 nautical mile, 40
maneuvering speed, 638–640 NLG. See Nose landing gear
manifold absolute pressure (MAP), 320 nose boom, 107–109
manufacturing and maneuver-based nose landing gear (NLG), 682–683
errors, 114–121
MAP. See Manifold absolute pressure OAT. See Outside air temperature
mass moment of inertia, 146–147 one engine inoperative (OEI)
mathematical upper bound model, certification regulations, 578–590
945–946 approach climb, 588–589
maximum landing weight (MLW), 133 final segment climb, 587–588
maximum ramp weight (MRW), 133 first segment climb, 579–583
 Index 1009

second segment climb, 583–586 pacer airplane, 89–92

third segment climb, 586–587 parametric landings, 948–953
check climbs, 591 passenger weight, 153–155
climb information to flight crew, perfect gas, 4
609–611 PEs. See Position errors
clear procedure, 615–616 p-factor, 750, 752
close-in charts, 617 phugoid, 469
conversions, of climb units, 611–612 pilot operating handbooks (POHs), 520
net climb gradients, 612–615 pitot probe off-angle tolerance, 75
on climb performance, 571–578 pitot-static tubes, 49
climb requirements, 578–590 POHs. See Pilot operating handbooks
conversions, of climb units, 611–612 position errors (PEs), 67, 72
flight test, 590–591 source of, 73–74
data reduction, 593–594 preactivation ice, 195–198
mitigation, 594–595 precision of sensor, 123
monitoring, 593 pressure, 1–2, 13
risk, 594–595 pressure altitude, 18–35
scope for, 591–593 barometric altimeter, 21–25
for thrust after 8 seconds, 595–596 barometric setting, 21–25
validation, 593–594 measurement, 19–21
gliding flight, 617–620 operational regulations, 25–28
information to flight crew, 620 temperature effects on measurement,
single engine airplanes, certification 30–35
requirements for, 618–620 propeller installation, 353
net climb gradients, 612–615 propeller thrust, 336–347
obstacle clearance, 598–608 propulsive efficiency, engine, 327
bank angle impact on climb
gradient, 608 radar altimeter, 36
climb information to flight crew, radar altitude, 35–37
609–611 range, airplane, 411–412
driftdown, 606–608 rate of climb (ROC), 522–526
en route climb ceiling, 604–606 airspeed, 538
takeoff performance maximum, 525–526
critical engine failure speed reciprocating engines, 314–321
VEF, 710 carbon dioxide, impact on, 321
ground run, 707–719 efficiency, 316–320
maximum brake on speed V1, environment, impact on, 321
710–712 maximum power available vs. altitude,
modeling VEF to V35, 713–719 320–321
takeoff speed summary, 719–720 Otto cycle, 314–315
takeoff trim settings, 708 reduced vertical separation minimum
VEF, event at, 713–719 (RVSM), 70–71, 404–410
takeoff weight limited by climb airspace, 29–30
requirements, 596–598 certification, 408–409
operating weight empty (OWE), 133 flight envelope, 409–410
operational ceiling, aircraft, 532–533 rejected takeoff (RTO), 813–848
Oswald coefficient, 294 antiskid systems for, 837–838
Otto cycle, 314–315 braking distance, 838–839
outside air temperature (OAT), 3, 50–52 braking force, 832, 834–838
OWE. See Operating weight empty causes, 813–814
1010 Operational Aircraft Performance and Flight Test Practices

rejected takeoff (RTO) (Continued) hysteresis, 123

critical speeds, review of, 819–823 precision, 123
development timeline, 850 resolution, 123
distance definition, 816–819 sensitivity, 123
EMAS technology in, 846–847 SFC. See specific fuel consumption
example, 817–819 sink rate, 948
requirements for, 816–817 six performance landings, 946–948
runway safety area, 846 skin friction drag, 291
runway surface conditions, 839–844 slipstream, 336
contaminated runways, 843–844 specific air range (SAR), 414–426
default braking coefficient vs. drag model testing, 493–499
contaminants, 844 on electric airplanes, 496–497
grooved runways, 842–843 instrumentation, 499
sequence for OEI condition, 816 integrated model verification, 497
statistics, 814 power leakage, 495–496
stopway, 845–846 secondary effects, 495–496
takeoff speeds, 819–820 validation flight, 497–499
testing, 860–868 flight crew, presenting information,
analysis, 861–866 503–505
braking coefficient, 861–866 specific fuel consumption (SFC),
data expansion, 867–868 413–414
for ground VMCG, minimum control speed of sound, 40–41
speed, 850–859 speed schedule, 538
instrumentation required, 860–861 spray impingement drag, 733–735
maximum kinetic energy, 868–872 SSEC. See Static source error correction
maximum wheel speed, 872 stall speeds, 217, 221–224
model validation, 866–867 data expansion, 286
risks and mitigations for, 861 data reduction phase, 267–285
takeoff field length, 872–891 accelerometer correction, 270–273
transition phase, 823–833 air data in stall, calibration of,
V1 critical speed, 820–823 268–269
wheel slip during, 836 altitude and test weight effects,
worn brakes, 844–845 283–285
Reynolds number, 492–493 CG correction, 277–279
ridge ice, 200–203 configuration effects, 267–268
rigging, flight controls, 214 entry rate correction, 282–283
rime ice, 194–195 production air data system, 269–270
ROC. See Rate of climb test weight verification, 273–277
roughness drag. See Excrescence drag thrust correction, 279–282
RSA. See Runway safety area definition, 221–224
RTO. See Rejected takeoff flight crew, presenting information to,
runway safety area (RSA), 846 286–287
RVSM. See Reduced vertical separation production tolerances, 285–286
minimum vs. stall identification, 227–230
testing, 250–266
SAT. See Static air temperature attitude recovery device, 264–265
sawtooth climb test, 551–558 instrumentation, 255–258
sensors preparation, 253–255
accuracy, 122–123 risks, 259–264
calibration, 123–125 telemetry monitoring, 258
 Index 1011

test card, 259 OEI testing, 786–792

test progression, 265–266 procedure, 779–780
VS. VCLmax, 250–253 rapid rotation, 794–795
stall testing, 217–287 rolling friction force coefficient,
accident, 262–263, 265–266 776–777
Aft stick elevator stop–defined stall, runway survey, 770–772
237–244 speed spread model, 792–794
FAA 14 CFR/EASA CS/TCCA CAR thrust, modeling excess, 775–780
Part 23, 237–241 VLOF to V35, 783–786
fly-by-wire (FBW) airplanes with VR to V35, 780–781
envelope protection, 242–244 VR to VLOF, 781–783
MIL-F-8785C, 241–242 wind considerations, 777–779
characteristics, 224–245 minimum control speed, air VMCA,
importance, 217–218 750–761
program timeline, 217 dynamic, 758–760
pusher-defined stall, 230–237 p-factor, 750, 752
poststall pusher, 235–237 risks and mitigations, 760–761
prestall pusher, 233–235 rudder, bank angle effect on, 757
for repeatability, 218–221 static, 754–755
stall in icing, 244–245 thrust asymmetry, 751
stall speed wing lift/tail aerodynamics, prop wash
definition, Part 23 small airplane, impact on, 753–754
221–224 minima unstick speed VMU, 761–769
vs. stall identification, 227–230 analysis, 766–767
testing, 250–266 data, 765–766
stall warning, 246–250 example, 767–769
stall turn speed, 643–644 geometry, 764
stall warning, 246–250 pitch angle, 762
standard pressure regions, 27–28 procedure validation, 746–747
static air temperature (SAT), 3, 50–52 risks/mitigations, 807–811
static pressure, 1 rotation speed VR, 742–743
static source error correction (SSEC), takeoff safety speed V2, 742–743
121–122 telemetry, use of, 807–811
steep approaches, 986–997 timeline, 741
stopway, 845–846 trim setting, 747–749
takeoff field length (TOFL)
takeoff distance, 677–679 balanced field length, 879–880
takeoff distance modeling/testing, minimum control speeds, 874–876
741–749 RTO expansion, 876–879
airspeed indication calibration, takeoff speeds, AFM, 881–891
743–746 military approach, 890
certification regulations, 796–807 special test airplanes, 890–891
abused takeoff testing, 797–798 takeoff distance, 882–885
early rotation, 798–803 takeoff weight, 885–890
mistrim takeoff, 803–806 thrust/power setting, 876
ground run distance modeling/testing, unbalancing takeoff, 880–881
769–796 winds for AFM, 874
DGPS runway survey analysis method, takeoff ice, 198
772–775 takeoff path, 572–573, 579, 583, 586–587,
instrumentation, 795–796 598–600, 678, 716
1012 Operational Aircraft Performance and Flight Test Practices

takeoff performance model, 679–719 tendency to float, 920–921

acceleration-go, 720–722 thrust-drag bookkeeping, 350–353
AEO aerodynamic impact on airplane, 353
forces, 685–691 boundary layer ingestion, 351–352
ground run, 681–707 excess thrust manager, 352–353
minimum unstick speed VMU, 693–694 propeller installation, 353
MLG and NLG compression, thrust model, 313–350
682–683 ducted fan, 347–349
pitch attitude, target, 704–707 electric motors engines, 333–336
pitch rate, target, 702–704 advantages, 333–334
rolling friction force, 684–685 carbon dioxide, impact on, 335–336
rotation speed (VR), 691–692 efficiency, 333–334
runway gradient, 692–693 environment, impact on, 335–336
takeoff, 680 gas turbines, 321–333
takeoff safety speed V2, 700–702 carbon dioxide, impact on, 333
trim setting, 707 efficiency, 323–324
VLOF to V35, acceleration from, engine, evolution of, 324
698–700 environment, impact on, 333
VR to VLOF, acceleration from, maximum thrust/power available vs.
695–698 altitude, 332–333
certification regulations, 722–739 principle of operation, 322
clearway, 722–725 turbofan cycle, 328–330
contaminated runway, 726–730 turbojet cycle, 325–328
displacement drag, 730–733 turboprop cycle, 330–332
dry snow compression drag, 735–736 turboshaft cycle, 332
Part 23, 736–738 propeller, 336–347
spray impingement drag, 733–735 propulsion system tree, 313
thrust, contaminant impact on, 736 reciprocating engines, 314–321
V-Go vs. V-Stop, 738–739 carbon dioxide, impact on, 321
wet runways, 725–726 efficiency, 316–320
OEI environment, impact on, 321
critical engine failure speed VEF, 710 maximum power available vs. altitude,
ground run, 707–719 320–321
maximum brake on speed V1, 710–712 Otto cycle, 314–315
modeling VEF to V35, 713–719 TOFL. See Takeoff field length
takeoff speed summary, 719–720 total air temperature (TAT), 3, 50–52,
takeoff trim settings, 708 110–114
VEF, event at, 713–719 total landing distance, 929–930, 965–971
takeoff distance, 720–722 coming in fast, 966–968
TAT. See Total air temperature coming in high, 968–970
temperature, 2–3, 13, 17–18 smooth landing, 970–971
air, 2–3 total pressure, 48–50
airspeed total temperature probe calibration,
temperature effects (TAS vs. M), 110–114
44–46 touchdown, 915–919
total air temperature, 50–52 tower fly-by, 92–98
effects on pressure measurement, 30–35 trailing cone, 98–101
flight envelope, 400–401 transition segment, landing, 893–894,
OAT, 3, 50–52 924–928
static air temperature, 3, 50–52 events, 927–928
 Index 1013

ground distance, 927–928 minimum radius of turn, 644–647

testing, 959–964 under OEI conditions, 674–675
Transport Canada Steeper, 997 reference system, 625–631
trim drag, 295–296 stall turn speed, 643–644
tropopause, 5 unrestricted, 672–674
troposphere, 4–5, 7–9
true airspeed calibration method, 74–89 unreliable airspeed (URA), 126–128
closed course method, 74–83 URA. See Unreliable airspeed
GPS method, 83–89 USAF Learjet C-21, 101
true airspeed (TAS), 40–43
vs. altitude in standard atmosphere, validation climb test, 560–562
44–46 vertical CG, 145–146
CAS vs., 60–61, 67 Vickers Viscount accident in Stockholm,
EAS vs., 53 Sweden, 201, 204
impact pressure vs. dynamic pressure, viscosity, 3–4
46–48 vortex drag, 292–295
vs. Mach in standard atmosphere, 44–46
Mach number and, 41–43, 50–52 wave drag, 296–298
total air temperature, 50–52 weight, aircrafts, 131–132, 134
total pressure, 48–50 AFM, 138–139
units, 40, 42 BEW in-service, 166
turbofan cycle, 328–330 breakdown, 133–134
turbojet cycle, 325–328 buildup, 135
turboprop cycle, 330–332 cargo and checked baggage, 155–156
turboshaft cycle, 332 center of gravity, 137–146
turn performance model envelopes, 156
assumptions, 625 fuel weight and, 147–153
buffet envelope, 663–672 lateral, 143–145
EASA additional requirements, vertical, 145–146
670–672 definitions, 133–134
information to flight crew, 669–670 fuel, 147–153
building, 623–625 mass moment of inertia, 146–147
climbing turn parameter variations, 624 minute, 164–166
energy maneuverability, 649–654 passenger, 153–155
equilibrium of forces, 625–631 on performance, 131, 166–169
fastest turn, 640–643 procedures and documentation, 160–163
flight envelope, altitude effects on, scales, 135–137
647–649 theory and practice, 134–137
generic grid for, 630 weight a minute, 164–166
inner wing stall in turn, 654–656 wet runways, 725–726
load factor wind aspect ratio (AR), 173
on stall speed, 632–635 wind tunnel test, 43
on thrust required, 635–638 wing, 172–175
maneuvering speed, 638–640 worn brakes, 844–845
maneuver margin and certification
regulations, 657–663 Yeager, Charles E., 41
en route, 661
flight testing for, 661–663 zero fuel level, testing for, 499–502
landing, 659–660 on battery-powered airplanes, 500–502
low-speed, 663 zero fuel weight (ZFW), 134
takeoff, 657–659 ZFW. See Zero fuel weight
Supporting Materials
To download supplemental material files, please go to AIAA’s electronic
library, Aerospace Research Central (ARC), at arc.aiaa.org. Use the menu bar
at the top to navigate to Books . AIAA Education Series; then, sort alphanu-
merically by title to navigate to the desired book’s landing page.
A complete listing of titles in the AIAA Education Series is available from
AIAA’s electronic library, Aerospace Research Central (ARC), at arc.aiaa.org.
Visit ARC frequently to stay abreast of product changes, corrections, special
offers, and new publications.
AIAA is committed to devoting resources to the education of both practicing
and future aerospace professionals. In 1996, the AIAA Foundation was founded.
Its programs enhance scientific literacy and advance the arts and sciences of
aerospace. For more information, please visit www.aiaafoundation.org.
ABOUT THE BOOK
Operational Aircraft Performance and Flight Test Practices is intended to serve as a single source
reference, from the basic theory to practical cases, for certification flight testing and operational
performance monitoring. It provides more real-life examples than are offered in traditional textbooks.
To this end, the first part of every subject begins with the development of the basic performance
equations based on a given set of assumptions. Following this, we deconstruct the maneuver being
analyzed to validate its components (create small flight test packages to validate models and expected
airplane behavior). Then, flight test considerations are discussed including required instrumentation,
flight test risk and risk mitigation, data scatter, data reduction and presentation of the performance
information to the flight crew. Several examples of flight test results help the reader better understand
what they may be faced with when collecting data to create performance models.

There is a major focus on FAA 14 CFR Part 25 certification requirements as a basis for discussion since
the format of this set of regulations is publicly available and clearly shows the basic items that go into
demonstrating that an airplane is safe to operate. The general approach is also valid for certification
under Part 23 or military specifications and, from time to time, the reader will encounter these additional
requirements to show variance into airplane certified performance. Throughout the book, readers will
find review of accidents and incidents that have occurred on various airplanes. It is not the purpose of this
book to assign blame or liability, but rather to share experiences and lessons learned to hopefully aid in
the prevention of future accidents and incidents as well as changes in regulations to continue improving
aviation safety.

ABOUT THE AUTHOR

MARIO ASSELIN is currently Chairman of Asselin, Inc. (www.asselininc.com) and a FAA
Flight Analyst DER (Design Engineering Representative) for Part 23 Small Airplanes and Part 25
Transport Category Airplanes. Mr. Asselin also holds concurrently the position of Engineering
Fellow at Bombardier Aviation, a Transport Canada DAD (Design Approval Designee), Learjet ODA
(Organization Designation Authorization) Unit Member (UM), and Chairman of the Safety Review
Board at the Bombardier Flight Test Center. He is a member of the Manufacturers Flight Test Council
(MFTC), of GAMA EPIC (Electric Propulsion and Innovation Committee), of the Manufacturers Icing
Certification Group (MICG), and of the Electric Flight Test Committee. Prior to this, he was Director of
Engineering, Operations, and Quality at the Bombardier Flight Test Center (BFTC) in Wichita, Kansas.
Mr. Asselin also held a position of Senior Manager for Flight Sciences and Flight Test Engineering &
Operations at Honda Aircraft Corporation and was Vice President Engineering at Sino Swearingen
Aircraft Corporation, now known as Syberjet Aircraft. Mr. Asselin’s early career included 14 years in the
Canadian Air Force and a short stay at CAE’s military division.

Mr. Asselin has taught courses in aircraft performance, stability and control, and aerodynamics for the
Royal Military College of Canada in Kingston, Ontario. He has also taught at McGill University, École
de Technologie Supérieure (ETS), Concordia University of Montreal, and Kansas University. He holds
a Baccalaureate of Engineering degree in mechanical engineering from the Royal Military College of
Canada and a Master of Applied Science (M.Sc.A.) in aerothermodynamics (aircraft icing) from École
Polytechnique of Montreal.

American Institute of Aeronautics and Astronautics

aiaa.org

ISBN: 978-1-62410-592-0