USAAMRDL-TR-7S-46

INTEGRATED ACCESSORY SYSTEMS FOR SMALL GAS TURBINE

ENGINES

Pratt A Whitney Aircraft

Division of United Technologies Corporation
Florida Research and Development Center
Box 2691
West Palm Beach, Fla. 33402

N
H

January 1976
o
< Final Report for Period 1 May 1973 - 30 June 1975

M 2
Approved for public release; we
distribution unlimited.
I^Ü^J
Prepared for
EUSTIS DIRECTORATE
U. S. ARMY AIR MOBILITY RESEARCH AND DEVELOPMENT LABORATORY
Fort Eustis, Va. 23604
•

EUSTIS DIRECTORATE POSITION STATEMENT

This report contains detailed information on the design and performance trade-offs
associated with advanced control and accessory systems for small gas turbine engines.
The information contained in the report and the findings and conclusions cited are
considered to be accurate and appropriate to the work conducted. The results of
this work will be considered in future planning of any work related to gas turbine
engine controls or accessories.

Mr. Roger G. Furgurson of the Propulsion Technical Area, Technology Applications

Division saved as project engineer for this effort.

CflW :T

fi

DISCLAIMERS

The findings in this report are not to be construed as in official Department of the Army position unless so
designated by other authorized documents.

When Government drawings, specifications, or other data are used for any purpose other than in connection
with a definitely related Government procurement operation, the United States Government theicby incurs no
responsibility nor any obligation whatsoever; and the fact that the Government may have formulated, furnished,
or in any way supplied the said drawings, specifications, or other data is not to be regarded by implication or
otherwise as in any manner licensing the holder or any other person or corporation, or conveying any rights or
permission, to manufacture, use, or sell any patented Invention that may in any way be related thereto.

Trade names cited in this report do not constitute an official endorsement or approval of the use of such
commercial hardware or software.

DISPOSITION INSTRUCTIONS

Destroy this report when no longer needed. Do not return it to the originator.

■-" TihiiBaifnii ■•■••"•'■'- II>I-.

I
 UNrT.ASfiTT-TPn
SECURITY CLASSIFICATION OF THIS PACE (Whm Dmtm Enlmrmd)
READ INSTRUCTIONS
)CUMENTATION PAGE BEFORE COMPLETING FORM

5D COVERED

INTEGRATED ACCESSORY SYSTEMS FOR SMALL

s
GAS TURBINt ENGINES,' ^ S 4
{nj ^W//r~FR-6983 ^ I
**X."T" *SmSSSSSSSiS3k mSSSmm

DAAJÄ2-73-C -Maf jt
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK
AREA * WORK UNIT NUMBERS
irfrM & Wi.itnev Aircraft, Division of United Technologies
Corporation. Florida Research and Development Center 62209A 1F262209^H76 00
Box 2691, West Palm Beach, Florida 33402 007EK
11. CONTROLLING OFFICE NAME AND ADDRESS

Eustis Directorate, U. Li. Army Air Mobility Research (//) Jan—» «76/ ^gi
and Development Laboratory \^L'Sm wSSSSSSSm K-tfrfp/)
Fort Eustis, Virginia 23604 284 läCÖy*
IlimilTnnil" 'Tltl"^-"^ f tUrmnl Imm
Irom Conlrolllnä Olllc
Conlrolllnf Otllc*) IS. SECURITY CLASS, fa'">/4
^
'6 Unclassified
IS«. DECLASSIFICATION/DOWNORAOING
SCHEDULE

It. DISTRIBUTION STATEMENT {ot Oil» Rmpott)

Approvci^ |orpij^|ii- n jiiiini illitiihutinn imlimiTi rt

toj
i-F-lc.: tv/-H*i-'/coc
17. DISTRIBUTION STATEMENT (ol Ihm mbtlrmcl mfnd In Block 20, II ilUmrmnl Irom Riport)

II. SUPPLEMENTARY NOTES

It. KEY WORDS (Conllnu* on nrormo old* II n*c»Mry «id Idmntll- by block numbor)

Helicopter Engine Controls, Accessories, Oil Systems, Vulnerability

»ABSTRACT (Cmtlmm mt mrtmo •«« « nmtmmm—r amä liontlly by block numbor)

The Intent of the Integrated Accessory Systems for the Small Gas Turbine Engines
Program was to optimize the controls and accessory systems for the next generation
of small turboshaft gas turbine engines for Army helicopter applications. A 1977
development time frame was considered. A primary objective of the program wis
to reduce vulnerability of the controls and accessory components without severely
compromising other Important design criteria.

DD | jJuTn 1473 EDITION OF I NOV »» IS OBSOLETE

UNL \SSIFIED
SECURITY CLMStFICATlu« OF THIS PACE (Whmn Dmtm Enlmrod)

VO/ S*'^ o
 TTTJPT ASSTffTFn
WCUWTV CLMIIFIOTIOM OF THU gAgjgtg gjjj MlttM*)

During the Phase I conceptual design phase, analyses and preliminary designs of
Control and Accessory (CtA) configurations for both front- and rear-drive engines
were accorapltshed.
Phase fl Included preliminary design of two candidate control and accessory systems
and seectlon of one system for detailed design. A tower shaft control and accessory
drive vlth an air turbine starter was selected, and a detailed system design of all
C&A components was accomplished. The critical Items of the C&A system were
Identified and test programs were recommended for Phase III.
Phaso ni Included testing of the following critical Items: (1) fuel pump Inducer (Inlet
sucKon tests), (2) high-speed oil pump (cavltatlon tests), (3) electronics cooling
tf ohnlques (performance tests), (4) power turbine overspeed sensor (performance
tests), and (5) starter overrunning clutch (endurance tests).

UNCLASSIFIED
»tCUHITY CLASSIFICATION OF THIS PAGE(lrh*n D«>« Enl.r.rfJ
 SUMMARY

The intent of this program was to optimize the controls and accessory systems
for the next generation of small turboshaft gas turbine engines for Army heli-
copter applications. A 1977 development time trame was considered for the
engine. A primary objective of the program was to reduce vulnerability of the
controls and accessory components without severely compromising other impor-
tant design criteria. The effort was accomplished by closely coordinating the
study, selection, and design of the controls and accessory (C&A) components
with consideration for their integration within the engine.

The program was divided into three phases. During the Phase I conceptual design
phase, analyses and preliminary designs of C&A configurations for both front-
and rear-drive engines were accomplished. The candlate C&A systems were
evaluated using weighted selection criteria, and two candidate rear-drive C&A
systems (Tower Shaft/Air Turbine Starter, Cluster Gearbox/Air Turbine Starter)
were recommended for further detailed design study.

With Army approval of the two recommended candidate C&A systems, the Phase II
detailed design effort was initiated. The early part of the Phase II effort was
directed toward analyses and selection of one system for detailed design. A tower
shaft C&A drive with an air turbine starter was selected, and a detailed system
design of all C&A components was accomplished. The critical items of the C&A
system were identified and test programs were recommended for Phase III.

The Phase HI critical Item test programs were selected by the Army, and de-
velopment testing of the following critical Items was accomplished: (1) fuel pump
inducer (Inlet suction tests), (2) high-speed oil pump (cavltatlon tests), (3) elec-
tronics cooling techniques (performance tests), (4) power turbine overspeed
sensor (performance tests), and (5) starter overrunning clutch (endurance tests).

Recommendations for future C&A system development programs were made.

i i
 PREFACE

This effort Is reported according to the terms of Contract DAAJ02-73-C-0003,

which is under the technical direction of Mr. R. G. Furgurson, Eustis Directorate,
U. S. Army Air Mobility Research and Development Laboratory, Fort Eustis,
Virginia.
The program was conducted by the Advanced Controls Group, Florida Research
and Development Center, Pratt & Whitney Aircraft, West Palm Beach, Florida.
Mr. B. C. Miller was the Program Manager. Technical coordination at P&WA
was provided by Mr. J. D. Estes; Mr. A. White coordinated the activities at the
Chandler Evans Corporation.

This report carries the internal designation of FR-6983.

 CONTENTS

SECTION PAGE

SUMMARY 3

PREFACE 4

ILLUSTRATIONS 7

TABLES 14

INTRODUCTION 16

A. Problem Statement and Approach 16

B. Description of Approach 16

n REQUIREMENTS DEFINITION - FRONT-DRIVE

POWER TURBINE 26

III SELECTION CRITERIA FOR CANDIDATE SYSTEMS .... 37

IV CONCEPTUAL DESIGN - FRONT-DRIVE ENGINE 39

A. Selection of Ten Candidate Systems 39

B. Front-Drive Study Conclusions 96

CONCEPTUAL DESIGN - REAR-DRIVE ENGINE 98

A. Engine Definition Studies 98

B. Baseline Engine Definition 104
C. Candidate C&A Drive/Starter Descriptions 106
D. Rear-Drive Study Conclusions 130

VI SELECTION OF SYSTEM FOR DETAILED DESIGN 132

A. Candidate Systems 132

B. Evaluation, Rating, and Selection of One System ... 137

VII DETAILED DESIGN OF SELECTED SYSTEMS 151

vni REAR-DRIVE ENGINE WITH INTEGRATED C&A

COMPONENTS 194

A. Final Engine Design Layout 194

B. C&A Component Critical Items 197
C. Engine Design Critical Items 202
D. Long-Range Engine Design Critical Items 203
 I
I

CONTENTS (Continued)

SECTION PAGE

IX TESTING OF C&A CRITICAL ITEM COMPONENTS . . k . . 204

A. Vane Element Oil Pump - Sundstrand Aviation .... 204

B. Starter Overrunning Clutch - AiReseareh 218
C. Electronic Control Components to Housing
Heat Transfer - Chandler Evans Inc 233
D. Fuel Pump Inducer - Chandler Evans Inc 241
E. Power Turbine Overspeed Governor - Chandler
Evans Inc 259

X CONCLUSIONS 270

XI RECOMMENDATIONS 271

APPENDIX - Engine Description 273

SELECTED BIBLIOGRAPHY 278

6
 ILLUSTRATIONS

FIGURE PAGE
1 Program Outline for Phase I -
Conceptual Design - Front Drive Engine ,,. 17
Program Outline for Phase I -
Conceptual Design - Rear Drive Engine 21
Preliminary Program Outline for Phase II -
Detail Design Phase 23
4 Baseline Engine Configuration 27
5 Tilting Pad Air Journal Bearing
Rear Compartment Conceptual Design 30
6 SemlreverseFlow Inlet Particle Separator 32
7 Inlet Particle Separator Alternate Inlet
Configurations 34
8 Engine Rotor Parasitic Losses 35
9 Compressor Map During Starting 35
10 Torque and Horsepower vs Ignition Speed 36
11 Starting Torque and Speed Requirement
(Sea Level/Cold Day) 36
12 Candidate C&A Drive (Tower Shaft Drive) 46
13 Candidate C&A Drive (Cluster Gearbox Drive) 47
14 Integral Electric Starter/Generator 4g
15 Candidate C&A Drive (Pneumatic Drive) 49
16 Candidate Pneumatic Drive Systems so
17 Turbine Horsepower Turndown Ratio 51
18 Integral Pumping Element 53
19 External Starter Systems 55
20 Integral Starter Systems 56
21 Logic Matrix for Selection of Five Systems 57
22 Integral Engine Starter (APU Bleed Air Supplied) 66
23 Impingement Starting Characteristics 67
24 Tank Volume Requirements for Impingement Starting
Gas Required for Two Starts 68
25 Impingement Starter 69
26 Variable Inlet Guide Vanes Used To Close
Compressor Inlet 71
27 Torque Characteristics of Ram Start/Closed IGV's 72
 ILLUSTRATIONS (Continued)

FIGURE PAGE

28 Compressor Map of Ram Start/Closed IGV's 72

29 Engine Coast-Down Characteristic of
IGV's Open 73
30 Tower Shaft and Pneumatic Turbine Drive Systems 78
31 Candidate C&A Drive 79
32 Hydraulic Starter and Accessories Drive 81
33 Pneumatic Starter and Accessories Drive 82
34 Tower Shaft/Gas Turbine Starter 84
35 Tower Shaft Gearbox/Impingement Starter 85
36 Cluster Gearbox/Air Turbine Starter 87
37 Cluster Gearbox/impingement Starter 88
38 Hydraulic Drlve/Hydraullc Starter 89
39 Air Bearing Engine Configuration 99
40 Alrframe Gearbox-Supported Power Turbine 101
41 Integral Start Turbine 105
42 Baseline Engine Configuration 107
43 Tower Shaft Drive/Air Turbine Starter 115
44 Tower Shaft Drlve/lmplngement Starter 116
45 Cluster Gearbox Drive/Air Turbine Starter 118
46 Cluster Gearbox Drlve/lmplngement Starter 119
47 Single-Speed Drlve/lmplngement Starter 120
48 Single-Speed Drlve/lmplngement Starter/
Integral Alternator 121
49 Front Drive Engine With Pad for Electrical Starter 131
50 Tower Shaft/Air Turbine Starter Candidate 133
51 65,000-rpm Centrifugal Pump 134
52 Tower Shaft Drive Fluid Controller and Fuel Pump 135
53 Cluster Gearbox/Air Turbine Starter Candidate 136
54 65,000-rpm Centrifugal Pump/Alternator 137
55 Cluster Gearbox Drive Fluid Controller
and Fuel Pump/Alternator 138
56 Tower Shaft Drive System Bearing
DN's and Gear Pitch-Line Velocity 143
 ILLUSTRATIONS (Continued)

FIGURE PAGE

57 Cluster Gearbox Drive System Bearing

DN and Gear Pitch-Line Velocity 144
58 Tower Shaft Configuration With a
38-deg Inlet Blockage 147
59 Engine Electronic Control 154
60 Torque Sensor Mounting Configuration 156
61 Torque Sensor Assembly 157
62 Optical Pyrometer 158
63 Turbine Blade Cooling 159
64 Optical Pyrometer Installation 159
65 Engine Inlet Air Temperature Sensor 160
66 Fluid Controller Implementation 161
67 Engine Fluid Controller and Bleed Valve 162
68 Engine Fluid Controller and Bleed Valve 163
69 Power Turbine Overspeed Sensor
Implementation 164
70 IGV Driving Link Adjustment 165
71 65,000-rpm Centrifugal Pump 166
72 Estimated Discharge Pressure
for 65,000-rpm Centrifugal Fuel Pump 167
73 Estimated Temperature Rise
for 65,000-rpm Centrifugal Fuel Pump 167
74 Estimated Efficiency for
65,000-rpm Centrifugal Fuel Pump 168
75 Engine Alternator Tower Shaft Assembly 169
76 Estimated Starting Torque
and Speed Requirements 170
77 Air Turbine Starter 171
78 ATS Performance 172
79 Engine Lubricating System Schematic 173
80 Lubricant/Scavenge Pump Element 176
81 15,000-rpm Oil Pump 177
82 Redesign of IPS Can Reduce Increase In
Compressor Inlet Air Temperature Resulting
From Engine Lubrication System Heat Load 179
83 Tower Shaft Drive System Bearing DN's and
Gear Pitch-Line Velocity 181
9
 ILLUSTRATIONS (Continued)

FIGURE PAGE
84 Air/Oil - Mist Lubrication System 183
85 Control and Accessory Gearbox
Bearing Compartment 184
86 Front Bearing Compartment 185
87 Rear Bearing Compartment 186
88 OU-Mlst System Plumbing 187
89 Estimated Shroud Static Pressure Variation 188
90 Estimated Compressor Shroud Air
Temperature Variation 188
91 Finned Inlet Duct Heat Exchanger Size
Required to Cool Bearing OU-Mlst Airflows 189
92 Finned Inlet Duct Heat Exchanger Size
Required to Cool Bearing OU-Mlst Airflow
From 660 to 250oF 190
93 Emergency Operation Alr/Oil-Mlst
Lubrication System 191
94 Air Temperature Effect on Heat Exchanger Size 192
95 Final Tower Shaft Drive/
Air Turbine Starter Engine Cross Section 195
96 Gearboxless Engine (1982 Time Frame) 198
97 Maximum P» mp Speed vs Minimum Inlet
Pressure fo * Various Cross Vane Pump Designs 206
98 Oil Pump - Test Schematic 207
99 OH Pump Test Arrangement 207
100 Pump Performance Variation With Altitude 208
101 Oil Pump Cavltatlon Tests
(Fluid Temperature 75-80oF) 209
102 Oil Pump Cavltatlon Tests
(Fluid Temperature 1750F) 210
103 Oil Pump Cavltatlon Tests
(Fluid Temperature 250oF) 211
104 Oil Pump Element Liner Showing
Material Removed From Inlet Opening 212
105 Pump Performance Variation With Altitude -
Increased Lubrication Element Inlet Opening 213
106 Oil Pump Cavltatlon Tests
(Fluid Temperature 70-80oF) 214

10
 ILLUSTRATIONS (Continued)

FIGURE PAGE

107 Oil Pump Cavltation Tests

(Fluid Temperature 1750F) 215
108 Oil Pump Cavltation Tests
(Fluid Temperature 250oF) 216
109 Oil Pump Components Before
50-hr of Endurance Testing at 250<,F 217
110 Oil Pump Components After
50-hr of Endurance Testing at 250oF 217
111 Phase I - Overrunning Clutch Test Rig 220
112 Overrunning Clutch Test Stand Schematic 223
113 Clutch Endurance Test Rig 223
114 Clutch Test Rig After
40-hr of Overrunning at 65,000 rpm 224
115 Quill Shaft and Spline After
40-hr Overrunning at 65,000 rpm 224
116 Clutch Components After
40-hr Overrunning at 65,000 rpm 225
117 Clutch Components After 137.5-hr of
Overrunning at 65,000 rpm (Overrunning
Bearing Failure) 226
118 Clutch Test Rig After 137.5-hr Overrunning
at 65,000 rpm (Overrunning Bearing Failure) 226
119 Phase II - Clutch Endurance
and Decoupler Test Rig 228
120 Clutch Components After
200-hr of Overrunning at 65,000 rpm 230
121 Inner Race and Bearings After 686-hr
of Overrunning at 65,000 rpm (With Oil Plug Failure) .... 230
122 Clutch Test Rig After
1,000-hr of Overrunning at 65,000 rpm 231
123 Inner Race at Conclusion of Endurance Test 232
124 Clutch at Conclusion of 1,000-hr Over-
running at 65,000 rpm Showing Clutch Sprag 232
125 Thermal Plane Heat Transfer Concept 235
126 Thermal Plane Printed Circuit Board 236
127 System Test Schematic 236
128 Printed Circuit Board Layout 238

11
 ILLUSTRATIONS (Continued)

FIGURE PAGE

129 Component Temperatures vs Thermal Thickness 239

130 Electronic Computer With
Thermal Plane PCB Implementation 240
131 Percentage Reduction In
Temperature Rise vs Thermal Plane Thickness 242
132 Inducer Test Rig Configuration 244
133 Inducer Pump Test Fixture 244
134 Inducer Test Schematic 245
135 Inducer Head/Flow Calibration 245
136 Inducer 10,000 ft V/L Performance 246
137 Inducer Sea Level V/L Performance 246
138 Inducer 20,000 ft V/L Performance 247
139 Cutback Inlet of the Inducer 247
140 Inducer Head Flow Calibration With Cutback Inlet 248
141 Axial/Conical Inducer Head vs T^low Calibration 249
142 Axial/Conical Inducer Sea Level V/L Performance 249
143 Axial/Conical Inducer 10,000 ft V/L Performance 250
144 Axlal/Conlcal Inducer 20,000 ft V/L Performance 250
145 Conical With Jet Inducer Test Conditions 252
146 Conical With Jet Inducer Sea Level V/L Performance .... 253
147 Conical With Jet Inducer
10,000 ft and 20,000 ft V/L Performance 253
148 Conclal With Jet Inducer
V/L Performance With Direct Air Induction 254
149 Mixed Flow Impeller -
Estimated Head vs Flow Performance 256
150 Comparison of Boost Inducer Cross Sections 256
151 Predicted Fuel Pump Total Temperature Rise 257
152 Power Turbine Overs peed Governor 260
153 Power Turbine Overspeed Sensor Test Setup 262
154 Power Turbine Overspeed Sensor and Test Drive 263
155 Overspeed Sensor Room-Temperature Calibration 264
156 Pneumatic Characteristic of Power
Turbine Overspeed Governor 265

12
■

ILLUSTRATIONS (Continued)

FIGURE PAGE

157 Variation in Power Turbine Governor Pressure Gain

With Ambient Temperature ., 266
158 Transient Response Tests 267
159 Failed CantQever Section 268
160 Improved Sensor Design 269
A-l Typical Mission Cycle 274

13
 TABLES

TABLE PAGE
1 Component Accessory Drive Systems Matrix « 40
2 Identification Notes for the Starter -
C&A Drive System Matrix 41
3 Ten Candidate Systems 43
4 Tower Shaft Gearbox Turbine Bleed Performance 51
5 Starter Concept Evaluation 59
6 Starter Concepts Evaluation Data 60
7 Self-Contalned Starter Evaluation Rating 62
8 Accessory Integration Program Starter
Concepts Evaluation Data 63
9 Starter Concept Evaluation 65
10 Bleed From APU For Engine Starting
(Integral Starter Evaluation Rating) 74
11 Integral Starter Evaluation Data 75
12 Controls and Accessory Drive System
Evaluation Rating 77
13 Fuel Pump Evaluation Rating 79
14 Hybrid Control and Accessory Drive System 83
15 Rating Summary 90
16 Reliability (Rating = 23%) 90
17 Vulnerability (Rating = 19%) 92
18 Weight and Volume (Rating = 10%) 92
19 Performance (Rating - 9%) 93
20 Maintainability (Rating - 7%) 94
21 Maintainability Comments 94
22 Installation Flexibility (Rating = 6%) ^7
23 Baseline Engine Heat Generation Summary 102
24 Bearing Configuration Candidates 103
25 Candidate Component and Accessory Drive Systems no
26 Rating Summary 122
27 Reliability (Rating = 23%) 122
28 Vulnerability (Rating - 19%) 124
29 Weight and Volume (Rating = 10%) 125
30 Performance (Rating = 9%) 127

14
 TABLES (Continued)

TABLE PAGE

31 Maintainability (Rating = 7%) 128

32 Installation Flexibility (Rating = 6%) 129
33 Components Weights and Volumes TS/ATS System 135
34 Components Weights and Volumes CGB/ATS System .... 138
35 Overall System Evaluation 138
36 Reliability Summary 140
37 Vulnerability Analyses 142
38 Alternator Design Requirements 145
39 Drive System and Components
Development Risk Rating 145
40 Comparison of C&A Drive System Components 145
41 Cost 146
42 Combined Weights and Volumes for
Candidate Configurations 146
43 Weight and Volume . . . 147
44 Performance 148
45 Maintainability Evaluation Comments 149
46 Englne/Alrframe Interface
and Component Accessibility 150
47 Sensor Requirements 153
48 Fuel Control Requirements 153
49 Engine Lubrication System Design Data 175
50 Small Gas Turbine Engine
Heat Generation* Summary 178
51 Performance Penalty Heating
of Engine Inlet Air Due to Engine Heat Generation 178
52 Air .md Oil Flow Requirements 182

15
 SECTION I

INTRODUCTION

A. PROBLEM STATEMENT AND APPROACH

.
Engine technology for military helicopters is advancing toward more reliable, ^
higher performing, lighter weight engines with low vulnerability and simplified
maintenance. Much of the advance in this technology has been accomplished by
increasing the rotational speed of the major engine parts, which has reduced engine
size. To avoid large, heavy, and extremely complex control and accessory (C&A)
systems, it has been necessary to keep the state of the art of the C&A systems
abreast of engine development.
Previously funded Army programs have contributed to the reduction of C&A com-
ponent cost, size, and improvement of packaging. The Integrated Accessory Sys-
tems for the Small Gas Turbine Engines program was directed toward the continued
advancement and optimization of small gas turbine engine C&A systems through
study and evaluation techniques to (1) combine the component functions into inte-
grated modules and (2) integrate component functions into the basic engine design.
The program closely coordinated the design of the C&A components and the
accessory drive system with the design of the engine Interfaces before the engine
configuration was firmly established.
The major problem In meeting the goal of this program was to establish a C&A
system that incorporates, at its proper level, each of the established perform-
ance and design criteria. Reliability is of major importance, but. In a military
application, a reliable component that is also highly vulnerable to combat
threats Is a poor compromise. Integratina the components within the engine
to reduce the exposed area, providing high-speed or dual-function components,
or locating the components so that the critical items are shielded by the engine
structure were means of achieving such reduced vulnerability. The develop-
ment risk and cost of these integration techniques were considered to reflect
the growing concern for cost reduction. The impact on overall engine perform-
ance and weight was reflected in the system design. Since an Army helicopter
must be field maintainable, the C&A implementation also considered component
repair or replacement. The problem was therefore not just one of component
integration, but component integration In a manner which did not severely com-
promise other Important operational criteria.

B. DESCRIPTION OF APPROACH

The program was divided Into three phases: (1) Conceptual Design, (2) Detailed
Design, and (3) Critical Item Fabrication and Test. These phases are described
In the following paragraphs.

1. Conceptual Design Phase

a. Conceptual Design Phase - Front Drive Engine
(A program outline for this phase is shown In Figure 1.)

16
i , u
 ESTABLISH ESTABLISH CONTROL ESTABLISH CANDIDATE
ESTABLISH COMPONENT ESTABLISH CANDIDA
PRELIMINARY ENGINE AND ACCESSORY STARTER AND DRIVE
TECHNOLOGY LEVEL C&A SYSTEMS
CONFIGURATION REQUIREMENTS TECHNIQUES

Requirements: Components: Approach: Approach:

• 3 lb/sec • Fuel Control • Review Previous C&A • Review Air Force Program
• 600 shp • Fuel Pump Contract Effoits "Gas Turbine Advanced
• Front End Drive • Lubrication System • Review Current and Accessory Drive System

:sv 65,000 rpm

= 26,000 rpm
• TIT - 2350oF
•
•
Electrical Power Generator
15-hp, 20,000 rpm Accessory
Pad
Pioposed C&A Technology
Programs
• Coordmute Technology Review
Study"
• Define Candidate Urivi'
Techniques
• PR = 10.0 • Variable Geometiy Actuators With C&A Vendors, • Describe All Candidate C&A
• Compressor Bleed Valve Systems in Terms ol
• Anti-Icing Valvn Accessory Drive Tec! "ique
Approach:
• Ignit.on System and Engine Interlace
• Scale Configuration • Engine Starter
From Current
and Previous Approach:
Small Fngme
• Define Mission
Programs
• Establish Engine Performance
• Maintain Flexible
at Extreme Flight Points
Configuration
• Determine C&A
Requirements

Figure 1. Program Outline for Phase I - Conceptual Design - Front Drive Engine

17
 CONDUCT TRADEOFF
LISH CANDIDATE EVALUATE 10 ANALYZE FIVE STUDY SUMMARIES
ESTABLISH CANDIDATE
TER AND DRI\/r CANDIDATE C&A CANDIDATE AND SELECT TWO
C&A SYSTEMS
ECHNIQUES SYSTEMS C&A SYSTEMS FRONT DRIVE
C&A SYSTEMS

Requirement Requirement: Requirement:

Air Force Program Evaluate the 10 Continue Analysis and Select Two C&A Systems
s Turbine Advdnceri Most Promising C&A Mechanical Design Studies for Preliminary Design
pssory Drive System Systems and Perform of FiVv C&A Systems To Based on Analysis and
Jv" the Necessary Trades to Provide Trades To Reduce Mechanical Design
Candidate ürivt' Select the Best Candidates 10 Two Systems Integration Efforts
Tniques Five Systems
e All Candidate C&A
:ems in Terms o\ Approach Approach Approach:
essory Driv» Tecl'iique • Perform Prellmlnsry Analv is • Perform Mechanical • Determine Trade
Engine Interfati' To Si^e .he Candidate Integration Studies To Study Rating FiCtor
C&A Components Define Interfaces • Evaluate Candidate
• Prepare Engineering Sketches • Modify Engine Design To C&A Systems
of Possible Interface Improve Integration • Prepare Summary
Techniques Where Possible Charts of Rat igs
• Analyze Each To Determine • Evaluate the C&A • Select Two Systems
the Advantages and Systems To Determine • Review with Army
Disadvantages and Trades Technical Representative
Relative Merits "nteria Shall Include:
• Select Five C&A Systems 1. Reliability
for Continued Study 2. Vulnerability Resistance
3. Development Risk
4. Cost
5. Weight and Volume
6. Performance
7. Maintainability
8. Installation Flexibility

e Engine

cÄ-

ij
 1. A preliminary front drive engine component sizing and flow-
path arrangement was established. The basin engine com-
ponents were guided by the specified engine description
(Appendix A).

2. The required control and accessory components and their

performance requirements were defined. These components
included the following:

Fuel Control
Fuel Pump
Engine Lubrication System
Starter
Electrical Power Generation for Engine Use
Variable-Geometry Actuator
Aircraft Accessory Pad (15 hp at 20,000 rpm)
Compressor Bleed Valve
Anti-Ice Valve
Ignition System.

3. A review of previous USAAMRDL and other Government-

sponsored programs was made, and component vendors
were surveyed to establish component technology levels
compatible with the engine development time frame of
1977.

4. The candidate control and accessory drive techniques were

defined. This effort utilized the results of the Turbine En-
gine Advanced Accessory Drive System Study (A FA PL
Contract F33615-72-C-1170) where applicable.

5. Candidate integrated control and accessory systems ap-

plicable to a front drive engine were Identified, and 10
systems were selected on a qualitative basis for prelimi-
nary design evaluation.
6. A preliminary analytical and mechanical design analysis
was conducted to evaluate the ln candidate schemes and
to recommend 5 candidates for further analysis.

7. Analytical and mechanical Integration studies of the five

selected systems were performed, and two integrated sys-
tems were recommended for further analysis.
b. Conceptual Design Phase - Rear Drive Engine

(A program outline of this phase is shown in Figure 2.)

1. A preliminary rear drive engine configuration was defined.

The control and accessory components were the same as
defined for the front drive arrangement.
2. Candidate control and accessory drive techniques were defined,
and six systems were selected on a qualitative basis for pre-
liminary evaluation.
19

WrtrnTMl PiGB BLANK-NOT FILMED

 3. Analytical and mechanical design analyses were conducted
to evaluate the six systems and recommend two systems for
further analysis.

4. An assessment of both the front and rear C&A drive/engine

starter system candidates was made, and a recommenda-
tion was made to the Army for approval to proceed with
the Detailed Design Phase.

2. Detailed Design Phas'4

(A program outline for this phase is shown in Figure 3.)

1. Preliminary design layouts were provided that define the
packaging, interfaces, volumes, environment and mechanical
implementation of the two candidate systems selected in tho
Conceptual Design Phase.
2. Design analysis support was provided for the layout effort
to establish the component sizes and weights and the effects
on overall system performance. Preliminary reliability,
maintainability and vulnerability assessme its of the two
systems were made. Evaluations of relative cost, develop-
ment risk and installation flexibility were also made.

3. Trade-off studies of the two systems on the basis of selec-

tion criteria and rating factors were conducted, and one
integrated system was recommended for design.

4. Preliminary control and accessory system component speci-

fications were established that identified the interface and
performance requirements.
5. A design of the selected system control and accessory com-
ponents, through a combined P&WA snd vendor effort, and
with sufficient detail to identify any high-risk areas, was made.

6. A design of the selected C&A/englne interfaces, insufficient

detail to show mechanical implementation and to identify any
high-risk areas, was made.

7. The high-risk components or subsystems were Identified

and ranked in order of priority from a standpoint of further
required development.
8. A test plan and a cost estimate for a test program to evalu-
ate each Identified high-risk component or subsystem were
prepared.

20
 1
ESTABLISH ESTABLISH E
ESTABLISH CONTROL ESTABLISH
1 PRELIMINARY CANDIDATE STARTER CANDIDATE
AND -^ C
| ENGINE ACCESSORY AND C8iA
| CONFIGURATION DRIVE TECKNIOUES SYSTEMS Cl
REQUIREMENT

t
Requirements:
1 Hequiremen
Same As Phase 1
• 3 lb/sec Analysis arv
• 600 shp Design S
• Rear Drive Six C&A
• Ngg-65,000 rpm Provide
• NPT - 36.000 rpm Candida»
• TIT - 2350°?
• PR = 10
Approach:
Perform
Integ
Defir
Prepare
of f(
Tech
Perform
ToS
C&A
Evaluate
Systi
Trad
Crite
1. R(
2. V
3. D
4. O
5. W
6. P<
7. M
8. Ir

Figure 2. Program Outline for Phabe I - Conceptual Design - Rear Drive Engine

/
21
II

CONDUCT TRADE-OFF EVALUATE REAR

EVALUATE
STUDY SUMMARIES vs FRONT DRIVE
SIX
AND SELECT TWO SYSTEMS AND PHASE II
CANDIDATE
REAR DRIVE MAKE RECOMMENDATION
C&A SYSTEMS
C&A SYSTEMS FOR DETAIL DESIGN

Requirement: Requirement:
Analysis and Mechanical Select Two C&A Systems
Design Studies of for Preliminary Design
Six C&A Systems To Based on Analysis and
Provide Trades To Reduce Mechanical Design
Candidates to Two Systems Integration Efforts

Approach: Approach:
• Perform Mechanical • Determine Trade
Integration Studies To Study Rating Factor
Define Interfaces • Evaluate Candidate
• Prepare Engineering Sketches C&A Systems
of Possible Interface • Prepare Summary
Techniques Charts of Ratings
• Perform Preliminary Analysis • Select Two Systems
To S'ze the Candidate • Review with Army
C&A Components Technical Representative
I Evaluate the C&A
Systems To Determine
Trades
Criteria Shall Include:
1. Reliability
2. Vulnerability Resistance
3. Development Risk
4. Cost
5. Weight and Volume
6. Performance
7. Maintainability
8. Installation Flexibility

^2
 COMPLETE PRELIMINARY
CONDUCT TRADEOFF RECOMMENDED ONE SYSTEM
LAYOUTS OF TWO
ANALYSIS FOR DETAILED DESIGN
CANDIOATE SYSTEMS

Requirements Requirements Requirements:

Provide Preliminary Design Layouts Perform Design Analysis and Trade Select One C&A System Based
That Define Packaging Interfaces, Studies Necessary To Evaluate on the Trade Study Results
Volumes, Environment, and Relative Merits of Each System Review Selection With Desinnated
Mechanical Implementation of the Technical Representative
Two Candidate C&A Systems

Approach: Approach Approach

• Obtain the Design Characteristics • Evaluate t'">e Performance • Apply Rating Factors to Trade-off
of Each Control and Accessory Characteristics Study Results
• Define Engine Interfaces for • Review Layouts To Determine C&A • Prepare Trade Study Results and
Selected C&A Systems Integration Impact on Engine Design Selection Information for Technical Review
• Define Required Engine • Evaluate the C&A System Based on
Configuration Changes the Following General Performance
• Prepare Two Preliminary Layouts and Design Criteria:
1. Reliability
2. Vulnerability Resistance
3. Development Risk
4. Cost
5. Weight and Volume
6. Performance
7. Maintainability
8. Installation Flexibility

Figure 3. Preliminary Program Outline for Phase II - Detail Design ^hase

23
 SUBMI"1" PRIORITY LIST OF
COMPLETE DESIGN OF IDENTIFY AND RANK
RECOMMENDED
SELECTED SYSTEM THE HIGH-RISK AREAS
TcST .'ROGRAMS

Requirements: Raauirements: Requirements:

Continue Layout Design To Further Identify High Risk Components or Prepare and Submit Test Program
Define the Final Integrated Subsystams and Rank in Order Plans in Priority Listing
C&A System ot Priority

Approach: Approach: Approach:

• Perform Detailed Analysis of • Critique Final C&A SyMem to • Determine Tests Required for
Engine Interface Identify High Risk Components Experimental Evaluation of
• Complete Design Layout ot and Subsystems High Risk Components and
Critical Interfaces • Review High Risk Items With C&A Subsystems
• Complete Component Vendors to Determine Any Related • Define Required Rigs for Tests
Design Layout Technology Programs • Provide Cost Estimates and
Recommended Test Prograr.s

nscnina rax BUNK-WT „„,„

, . ■-

J
3. Fabrication and Test Phase
1. Test rigs to allow the selected tests to be conducted were
fabricated. Tesf rigs and C&A hariware in existence from
previous contracts were reviewed and utilized to the fullest
extent possible.
2. The component or subsystem functional and endurance tests
were conducted in accordance with the approved test plan to
fully evaluate the high-risk aspects of the component or sub-
system. Critical interface conditions for the components
which were identified in the Detailed Design Phase were simu-
lated.

3. The initial test results were evaluated, and the components

were modified and retested as required for full evaluation,

4. After completion of all tests, the test results were evaluated

for relative success or failure to meet the previously estab-
lished performance and endurance goals.

PRECEDirn PATrS BUNK-WT fllMSD

 SECTION n
REQUIREMENTS DEFINITION - FRONT-DRIVE POWER TURBINE

A baseline engine configuration was established as a starting point for the C&A
optimization studies. The detailed engine requirements are outlined in the engine
description, included in Appendix A. The engine requirements are summarized
below:

3 lb/sec Airflow
600-hp Power Turbine
1977 Development Time Frame
Dual Engine Installation
Front Drive Power Turbine
15 hp, 20,000 rpm, PTO
Emergency Oil System - 6-min Capacity

A baseline was derived from previous and current small engine design activities
at FRDC. The engine technology was selected to be consistent with the develop-
ment time frame.

A cross section of the basic engine without controls and accessories is shown
in Figure 4. Since it was the intent of this study to optimize the overall C&A
system, this layout was used only as a starting point for the study to establish
the basic engine component sizes, flowpath, and bearing configurations. The
engine design characteristics are outlined below;

65,000 rpm - Gas Generator

26,000 rpm - Power Turbine
2350° F - Turbine Inlet Temperature
10:1 Pressure Ratio - Compressor
Scmireverse Flow, Ejector Type - Inlet Particle Separator (IPS)

The engine configuration consists of a single-stage centrifugal compressor, an

annular combustor, and a single-stage, cooled, axial turbine. The power turbine
is a two-stage, uncoolcd, axial turbine with the shaft concentric with the gas
generator rotor.

The centrifugal compressor has a separate inducer, pipe diffusers, and a variable
inlet guide vane assembly. The centrifugal compressor offers the advantage of
being less vulnerable than an axial or axial/centrifugal design.

The combustor is a full-annular, radial-inflow design using advanced cooling

techniques and air-atomizing fuel nozzles. Combustor liner cooling is accom-
plished through the use of a finned, double-wall-const ruction technique,
FINWALL?.
i
5

g
I
The gas generator turbine consists of a single high-work stage, with convectively
cooled vanes and blades, and Includes a thermal-response rotor tip shroud to
control tip leakage. The control ring structure that surrounds the turbine blade
row also adds vulnerability protection.
A two-stage, axial power turbine is necessary for acceptable engine performance
with a front drive engine of this size. The rotor is limited in allowable operating
speed due to critical speed of the power turbine shaft; therefore, two stages are
necessary to obtain the desired turbine efficiency.

1. Bearings and Seals

An engine bearing and seal configuration, consistent with the 1977 development
time frame, was established. The system was configured for low heat genera-
tion to minimize the size of vulnerable heat exchanger components. A backup oil
system, with a 6-min emergency operational capability, was required. The
recommended baseline engine bearing and seal configuration is described as
follows:

1. Front Bearing Compartment

Oil-Lubricated, Antifriction Bearings (3)
Hydrodynamlc Lift-off Seals (2)
Carbon Intershaft Seal
Elastomeric Damper

2. Rear Bearing Compartment

Tilting Pad Air Bearings (2)

3. Heat Exchanger

Alr/Oll Heat Exchanger - Integral with front bearing compartment

Oil Temperature - 200 to 250oF

4. Backup Oil System

Oil Mist - 6-mln capability

The hybrid air bearing/oil-lubricated bearing configuration was selected, based

on the requirement that some type of oil-lubricated gearing to the gas generator
shaft in the front compartment would most likely be required. Additionally, the
oil-lubricated bearings have a greater capacity for the radial load that can be im-
posed by misalignment of the power turbine shaft with the alrframe gearbox.

Air bearings are planned for the rear compartment to reduce engine vulnerability
by eliminating oil system components and to reduce the size of the oil heat ex-
changers. A typical air bearing configuration, using tilted pad air bearings, is

28
i
shown in Figure 5. Mechanical Technology, Incorporated, has successfully
tested tilted pad air bearings for stop-start and high-temperature material
evaluation. (!) Other manufacturers have successfully evaluated foil-type
bearings. The final air bearing configuration will not be established as a part
of this program, but the present state of the art Indicates that, with a con-
tinuing research and development effort, air bearings are a reasonable risk
for an engine starting development in 1977.

Air bearings in the rear compartment can operate at higher temperatures than
the oil-lubricated system and, therefore, simplify the overall cooling and the
rear compartment insulation requirements. The airflow requirements for the
bearings are reduced by the higher operating temperatures due to the viscosity
effects. Bleed air requirements are estimated at 0,05 to 0,10 lb/sec.

Hydrodynamic lift-off seals will be used where the available envelope allows.
This will significantly reduce the oil cooling heat load due to elimination of the
rotational friction. The rotating Intershaft seal is proposed as a conventional
carbon face seal because of the limited envelope available.

Hydrodynamic lift-off seals (shrouded Raylelgh step lift pads operated with a
gas film separating the sealing faces) have been designed, fabricated, and tested
at P&WA.(2) Tests have demonstrated the feasibility of operation at gas tempera^
tures to 1200oF, pressure differentials to 300 psi, and sliding speeds to 500 ft/sec.
Conventional contact seals are used in place of labyrinth seals when air and oil
leakage is a problem. Their disadvantage is high wear.

Elastomeric dampers are recommended to replace oil film dampers to improve

engine vulnerability. In-house IR&D testing of elastomeric dampers indicates
adequate life, better tip clearance control, more design freedom with stiffness
and damping coefficients, and better operation with rotor unbalance that might
be caused by combat damage.

Engine heat generation has been reviewed in sufficient detail to establish the
need for an air/oil cooler. A fuel/oil cooler will not have sufficient capacity
at flight idle. Two air/oil cooler configurations were considered: (1) a heat
exchanger located on the inlet OD that uses channeled pressurized oil, and (2)
a heat exchanger located on the Inlet ID that uses oil mist on the front compart-
ment wall. The first approach uses a separate air/oil heat exchanger and trans-
fers the heat to the air that Is bypassed in the IPS. The recommended system
uses the front bearing compartment as an Integral heat exchanger and transfers
the heat Into the main engine alrstream.

^Swenson, K. R., N. M. Hughes, and D. F. Hever, EVALUATION OF GAS

LUBRICATED HYDRODYNAMIC BEARINGS IN A GAS TURBINE ENVIRON-
MENT, Report No. AFAPL-TR-72-41, June 1972.

(2)
Povinelli, V. P., and H. H. McKibbln, DEVELOPMENT OF MAINSHAFT
SEALS FOR ADVANCED AIR BREATHING PROPULSION SYSTEMS, Phase I,
Report No, NASA CR-72737, PWA-3933, and Phase II, NASA CR-72987,
PWA-4263.

29
 .1to

<ü
K

U
a
v

■a
E
•-a
Ja

«f S
2 Cu
z bO
c
o
u •3
Ul
w H
•
m

30
Use of the finned Inlet ID concept results In a performance penalty. A 3.20F
rise in average inlet temperature was calculated at flight idle. A 10F rise in
Tt2 will increase SFC 0.13% and decrease SHP 0.4%. These values are based on
a bulk temperature rise.

An emergency air/oil mist system is recommended to further reduce vulner-

ability. Emergency air/oil mist systems have been tested and proved feasible
for 6-min operation.

The problems with air/oil mist are:

1. Determining the proper oil and air mixture for each

bearing.

2. Cooling the inner races more effectively than the outer to

control thermal distortion,
3. Scavenging air from the compartment.

4. Determining proper oil mist line velocity to avoid problems

of condensation.

A simplified oil system that Is used only In the front compartment is more
amenable to an oil mist backup system than an engine with multiple oil-lubricated
bearing compartments.

A detailed analysis of an air/oil mist lubrication system is presented in

Section VII.

2. Inlet Particle Separator

The IPS is a semlreverse flow separator, which Is Integrated into the compres-
sor Inlet case and uses scavenge airflow from a hot air ejector Incorporated in
the engine exhaust nozzle. A schematic of this system Is shown In Figure 6.
The fixed-geometry tailpipe ejector requires no valves or blower. The use of
preswlrl vanes Is also not required. The design of the ejector must be carefully
coordinated with the power turbine and IR suppressor designs to assure proper
ejector performance over the operating range.

Current development work on erosion-resistant coatings for engine components

Indicates potential application In advanced engines. ™' The optimum engine dust
protection system may Incorporate both a simple IPS and basic erosion-resistant
protection to the engine components.

(3)
McAnally, W. J., Ill, EROSION RESISTANT COATINGS FOR TITANIUM,
Pratt & Whitney Aircraft, AMMRC CTR73-6, January 1973.

31
 DUCT

Figure 6. Semireverse Flow Inlet Particle

Separator

A detailed design of the IPS was not established during this program. The IPS
design was considered flexible to accommodate integration of the C&A com-
ponents. Partial inlet or radial inlet IPS configurations, as shown in Figure 7,
were considered.

3. Starter

The engine starting horsepower requirements were analyzed In detail, since

they have a significant impact on the C&A system design. The engine rotor para-
sitic losses were based on applicable rig data and are shown in Figure 8. The
minimum engine starting speed was based on a'prediction of where stable com-
pressor operation would be expected after ignition and is illustrated on the rep-
resentative compressor map shown in Figure 9. The motoring torque for the
engine for standard and cold day Is shown In Figure 10. The starter was sized
for a 2 ft-lb torque margin above the cold day requirements. The relationship
of starter horsepower to light-off speed Is also shown on Figure 10. For a sea-
level cold-day start, the starter size was established at 16 hp at the light-off
speed of 16,000 rpm. The required torque-speed relationship Is shown In Fig-
ure 11. The starter was sized to accomplish a 30-sec, sea-level, standard-day
start time.
4. Controls and Accessories

The requirements for the C&A components were established, and a preliminary
definition of the component configuration was made to aid In the overall system
selection. The Chandler Evans Corporation (CECO) was subcontracted to
support definition and design of the control system components. The CECO
effort under Contract DAAJ02-70-C-0002, Advanced Engine Control Program

32
for 2 to 5 lb/sec Airflow Engines, was used to guide the control component
designs, where applicable. The following C&A components are required:
Fuel Control and Sensors
Fuel Pump
Inlet Guide Vane Actuator
Compressor Bleed Valve
Anti-Ice Valve
Alternator
Ignition System
Starter
Lubrication System

A detail description of the component requirements and the component imple-

mentation is presented in Section VII,

33
 CD
eo

I
%
o
U

'S
1—4

V
•w

ECD

u
o
+->
a
u
re
a

*
in
P.
■*->

I
u

34
 20 40 60
ROTOR SPEED - rpm (Thousands)

Figure 8. Engine Rotor Parasitic Losses

I
cc

UJ
ff
a.

ff
3
CD

0.2 0.4 0.6

INLET FLOW - lb/sec

Figure 9. Compressor Map During Starting

35
 40

ft
30

ft
20

%
& 10

10

1
1
6
UJ

12 14 16 18 20 22
ROTOR SPEED AT LIGHT-OFF - rpm (Thousands)

Figure 10. Torque and Horsepower vs Ignition Speed

MAXIMUMfl TOROUE

-MINIMUM
■^ ^^ ••
21 ■^^
*** ,
:
IDl E !
■ f
••
• ^"^ ^
1
«1
= •2
f /
•■
■ ■•
v
ui
K
IGNITION
••
SELF
J SUSTAINING \
■
••
/

J
STARTER (:UT-OFF
ft
o 10 20 30 40 50 60
GAS GENERATOR SPEED - 1000 rpm

Figure 11. Starting Torque and Speed Requirement

(Sea Level/Cold Day)

36
 SECTION III

SELECTION CRITERIA FOR CANDIDATE SYSTEMS

The following general performance and design criteria were utilized during the
concept selection and design phases of this program as they applied to the overall
engine C&A system under consideration and are presented in order of their
estimated relative importance.

Reliability - The reliability (safety, mission and unscheduled

maintenance) of the system considering the actual operational
environment of U. S. Army helicopters.

Vulnerability Resistance - Of primary consideration is the vulner-

ability to small-arms fire. In addition to actual direct damage, the
effects of secondary damage, fire and overall survivability were
considered.

Development Risk - This is a judgment that was based on the present

state of the art of the component being considered and its projected
rate of development (assuming a moderate effort is specifically
applied) over the total time frame involved, i.e., formal develop-
ment initiation in 1977.

Cost - Projected production cost assuming a rate of 500 units per

year over a total period of 10 years.

Weight and Volume - Installed weight and volume of the engine C&A
system.

Performance - The ability of an individual component to perform

its required function(s) and its effect on overall system performance.
Maintainability - The relative ability of an individual component or
complete system to be maintained in an operational condition in the
least time, at the least cost and with a minimum expenditure of
support resources.

Installation Flexibility - The characteristics of the C&A system

which provide the least overall design complexity relative to the
engine's installation into a flight vehicle.
These criteria were established as a combined effort of the Army, P&WA, and
CECO and then assigned weighted values on the basis of recommendations of the
personnel associated with the program. The final weighted values for the
criteria are outlined below.

37
 Criteria Weight - Percentage
Reliability 23
Vulnerability 19
Development Risk 14
Cost 12
Weight and Volume 10
Performance 9
Maintainability 7
Installation Flexibility 6
100
The safety-oriented factors, reliability and vulnerability, received weighting
which accounted for 42% of the total. Reliability In a manned aircraft Is, of
course, of prime Importance and received a strong consideration In this analysis.
Vulnerability In a manned aircraft that Is used for a military application that
has a high exposure to ground fire Is another prime consideration, and In this
case was rated approximately 80% of the value assigned to reliability.
The cost-oriented factors, development risk and production cost, received a
total of 26% of the total weighting, with cost rated slightly below development
risk. The development cost was reflected, to some degree. In the development
risk evaluation. While these considerations did not consider all of the factors
relating to life-cycle cost, it Is not anticipated that a complete Iffe-cycle cost
analysis, which was beyond the scope of the program, would have influenced
the overall ratings.

The mission effectiveness oriented parameters, weight and volume, and per-
formance, received 19% of the total and were rated approximately equal. The
operational oriented factors, maintainability and installation flexibility, re-
ceived 13% of the total, with flexibility rated slightly below maintainability.
Even though weight and volume, performance, maintainability, and installation
flexibility received a total weight of only 32%, It Is Important to recognize that,
while these criteria have a small weight in the basic system selection, their
Importance was not overlooked In the detailed design.
These general categories are summarized below:

Rating Percentage

Safety 42
Cost 26
Mission Effectiveness 19
Operational Effectiveness 13

The primary goals of this program were safety and cost oriented, but the per-
formance and operational considerations received more attention as the final
system design was formulated.

To be considered a viable candidate, each system that was considered was re-
quired to exhibit a minimum level of performance in each criteria. For ex-
ample, a system that caused an unrealistic engine performance (TSFC) penalty
was not considered even though it was acceptable in the other areas. This step
was necessary to preclude the possibility of selecting a system, based on es-
tablished rating criteria, which would not be a realistic candidate.
38 ,
 SECTION IV

CONCEPTUAL DESIGN - FRONT DRIVE ENGINE

A. SELECTION OF TEN CANDIDATE SYSTEMS

The C&A drive systems considered for this program were divided into 17 schemes.
The starter drive systems were broken into two basic divisions: mechanical drive
input and integral. A matrix of the candidate systems is shown in Table 1. The
matrix consists of the 17 C&A drive schemes corpbined with the two basic starter
drive systems, including eight types of starter tncive schemes. The total combina-
tion of C&A drive and starters creates 129 possible systems for study.

Each system was reviewed and several were logically eliminated. The remaining
systems required analysis to allow a decision to be made as to whether the con-
cept should have further consideration. The justification used for screening each
system is specified on the matrix and summarized in Table 2. The 10 candidate
systems selected for further evaluation are shown in Table 3.

1. Candidate C&A Drive/Starter Systems

During the conceptual design phase, the method of driving the C&A components
and the method of starting the engine were emphasized. The C&A drive systems
and the starter systems originally considered in the study are described below:

1. Candidate C&A Drives

a. Mechanical

(1) Tower Shaft Drive - Tower shaft drive through a

gearbox with multiple gearing for the various
i
required accessory drives

(2) Cluster Gearbox - Cluster gearbox mounted about

centerline of engine with multiple gearing for
the various required accessory drives

(3) Single-Speed Module - Tower shaft drive through a

gearbox with a single drive shaft. All required
accessory drives run a the same speed from a
comr.on shaft.

b. FlectJ ical

(1) Integral Starter/Generator - Starter and generator

use spme winding, housing. , etc., and are integrated
into the gas generator rotor. The electric generator
can drive one or multiple motors for the required
accessory drives

(2) Integral Generator - Electric generator is integrated

into the gas generator rotor and provides power
for the required accessory drives. Other starting
means must be employed.
I
 ■3
-.IOJIV . diu>vj Q73 a 73
1 1
a ": 1
■M ^ CM N n •r 1

diuivi ionj aoi JOJIU.I 73

■
73

-uao • -1)3913 * VI T

2
i
ax Q73
I I
3 QQ 1
1 1 1
* 1
uo^iv/duiivi XI 73 73 »3 73 ■^ 1
.iiimupXn i VI i 1

jojoiv/Jossaadiuoj
*2
Si 73 73 73 73 73 73 1
.Jijuuinoud i VI 1 - CM <e
X
.loiojv/aojBJouao 73 73 73 73 73 73 73 73 731
S 0(JW3I3 t VI CM (C -!•
JOiKaauao lEoSajui Q 73 a 73 Q 3 73 1
■ 1
+ AI 'O 'III 'I I I
X) W ob M i *1 1
1

w duind
H 3
oiintupXn
2 Q a 73
■ 1
C 73 3 Q 73 1
CO
|i:.ii1<>|U|
A I I
CM CM
I
M M 4

w
> paaia c a 0 Q 3
paxuv A A
373
1 1
W C1!
Q73
CO CM
1
n c!;
"i 1
-f j
Q
paaig 2 Q 0 Q73 Q73 Q 3 QQ ]

5 .luuj.ini.i.ijui A i
in i CM icJ. A 1 1

O i& o
c uana Q Q Q C 73 Qu- a 3 CA 1

W
puc paaia lO A t^ CM .ic!. 1-
1 1
t
1
1

Ü paam a Q EA
i L L 1
SEO Pioo -r

Q CA 1
aoifaauan a■ C 73
I I
Q 73
1 1 1 1 1
jEjSajui CM CM -r 1
g < <
aoiBJauao < C 73 <
I I
/aajjEIS IBuSaiU] CM W is 2

o
ü ainpoiv poadg Q 73 Q 73 a x 1
a|8uis JJEqs aawox 1 1
CM M
1
4-
xoqjBao M

■C aaisni.) f

C
PQ xoqjRao W
IjBqs JaMOL -r

i
2n . 9 « a C
u g a -a c SU a
at
■P 5 5 | «2 5w I, B —
C5
It s^ a §1;
S? 2 S |
.St
O -3 a
M i 9 B be
Q

\* « 4;< &<«
6 a ||
^. O a S 0
feg «i c 2 o
EWOX u
Q w ti CM n un
U i
lEajuBqaaH li;.iX.ilui
siua)|j.<s ■>A!UI JOJUEIS

40
 TABLE 2. IDENTIFICATION NOTES FOR THE STARTER -
C&A DRIVE SYSTEM MATRIX

1. Considered a candidate C&A/starter system

2. Not technically feasible or considered beyond 1977 state of
development
3. Excessive TSFC penalty
4. Logistic requirements not compatible with mission
5. If a mechanical drive is required for start, no advantage to
another drive technique for the accessories
6. For hybrid systems, the starter and accessory drive modes
should be common
7. Excessive reliability penalty
8. An integral generator for control power only offers no advan-
tage if a separate accessory drive device is available
9. Excessive packaging penalty for the available space
10. Excessive vulnerability penalty

S - Starter Drive
D - C&A Drive

Pneumatic

(1) Cold Gas Bleed - Engine airflow gas bleed upstream

of the burner is used for providing power for the
required accessory drives.

(2) Bleed and Burn - Engine airflow gas bleed up-

stream of the main burner Is fed through a separate
combustor, mixed with a fuel supply, and burned.
The combustion products are then used to provide
power for the required accessory drives .

(3) Interturblne Bleed - Engine airflow gas bleed down-

stream of the gas generator turbine is bled and
used to provide power for the required accessory
drives.

(4) Mixed Bleed - Engine airflow gas bleed is bled from

the engine, both upstream of the burner and down-
stream of the gas generator turbine. The hot and
cold gases are mixed to provide power for the
required accessory drives.

41
 d. Hydraulic

Integral Hydraulic Pump - A hydraulic pump is

integrated with the gas generator rotor, and the
hydraulic power from the pump is used to pro-
vide power for the required accessory drives.

e. Hybrid

(1) Any C&A Drive/Integral Generator - All accessories

are driven by mechanical, pneumatic, or hydraulic
power with an electrical generator Integrated
with the gas generator rotor to supply electrical
power for the control and Ignition system only,

(2) Mechanical/Electrical Interface - A tower shaft

is used for driving an electric generator. The
generator Is used for powering the required
accessory drives.
(3) Mechanical/Pneumatic Interface - A tower shaft
is used for driving a separate air compressor.
The compressor Is used for powering the required
accessory drives.

(4) Mechanical/Hydraulic Interface - A tower shaft Is

used to drive a separate hydraulic pump. The
hydraulic pump Is used to power the required
accessory drives.

(5) Auxiliary Power Unit - A self-confined auxiliary

power unit supplies power for the required accessory
drives,
(6) Mechanical Drlve/Electrlcal Fuel Pump Interface
An electric generator, Integrated with the gas
generator rotor, will provide electric power for
driving a variable-speed pump. All other accessories
are driven by tower shaft.
f. Integral

(1) Integral Fuel Pump, Oil Pump and Alternator -

All the accessories are Integrated into the gas
generator rotor and operate at gas generator
speed.

42
 TABLE 3. TEN CANDI

Com

Mechanical Pneumatic

A.
A. Tower Shaft B. Cluster Gearbox A. Cold Gas Bleed
1 3
A. Electric Motor IST
1. APU/Generator
2. Battery

I B. Hydraulic Motor
1. APU Supplied Hydraulics
2. Pressurized Blowdown
®

C. Pneumatic Motor ® ®
1. APU Bleed
2. APU Bleed + Fuel/Com-
bustor/Turbine Starter
D. Auxiliary Power Unit
1. Gas Turbine
®
2. Piston Engine
3. Wankel Engine

«0 2 4 5

i A. External Pneumatic Supply

(Cold Gas Impingement)
w
u
B. External Pneumatic/Fuel Supply/ ® ® ®
Auxiliary Combustor (Hot Gas
Impingement)
C. External Pneumatic/Fuel Supply for ® ® ®
I Engine Ram/Closed IGV's

Legend:
c ' - Basic
® - Derlv

43

. i
 M

lANDIDATE SYSTEMS

Controls and Accessory Drives

Hybrid

D. Mechanical Gearbox
A. Mechanical Gearbox/ B. Mechanical Gearbox/ C. Mechanical Gearbox/ Driving Plus An Electrical
I Electrical Generator Compressor Hydraulic Pump Fuel Pump Drive
9
[±1 "ST

m ®

m ®

(S)

Basic C&A Starter System

Derivative Starter Configuration

*z
2, Candidate Starter Systems

a. Mechanical

(1) Mechanical/Electric - An electric-powered starter

driving a gearbox connected to the gas generator
rotor. Starting power provided by APU/generator
or battery system.

(2) Mechanical/Hydraulic - A hydraulic-powered

starter driving a gearbox connected to the gas gen-
erator rotor. Starting power provided by APU/
hydraulic pump or accumulator blowdown.
(3) Mechanical/Pneumatic - A pneumatic-powered
starter driving a gearbox connected to the gas
generator rotor. Starting power provided by
APU bleed.
(4) Self-Contained Starter - A self-contained gas
turbine, piston, or Wankel engine mechanically
connected to the gas generator rotor.

b. Integral
(1) Integral Electric - An electric starter Integrated
with the gas generator rotor.

(2) Integral Hydraulic - A hydraulic starter Integrated

with the gas generator rotor.

(3) Integral Pneumatic - An external APU supplying

bleed air, which is u^ed In either of the three
following methods: cold gas Impingement, heat
addition In the gas generator burner with closed
engine Inlet, and hot gas Impingement using heat
addition In an external burner,

(4) Cartridge - Solid grain hot gas generating device

used for driving the gas generator rotor directly
or through an Intermediate system.

p
fiEC£DTrc Pi«- nr

45
2. Preliminary Screening Analyses

The logic and analyses used to preliminarily screen the candidate C&A drive/
starter system are summarized in the following section. These analyses reduced
the candidate C&A drive /starter system to 10 basic categories. These 10 cate-
gories included 26 possible C&A drive/starter combinations. Each of the candi-
date C&A drive and starter systems considered is described below.

a. C&A Drives

(1) Mechanical

The candidate mechanical C&A drives considered were:

1. Tower Shaft - Considered a candidate and is shown in Figure 12.

2. Cluster Gearbox - Considered a candidate and is shown in Fig-

ure 13.

3. Tower Shaft/Single-Speed Module - The use of a 20,000-rpm.

single-speed C&A module was discounted on the basis of in-
compatibility with the oil pump system. A - iigle-stage
20,000-rpm oil pump element is not a valid mechanical con-
figuration due to the disproportionate pumping element L/D
(7:1). A two-stage oil pump, using a centrifugal inducer,
was considered to be unduly complex when compared to a lower
rpm single-stage positive-displacement pump. The centrifugal
oil pump inducer is also subject to high blade loading at -650F
conditions.

PTO DRIVE-i
£
v
ACCESSORY/ /]7"
DRIVES /

Figure 12. Candidate C&A Drive (Tower Shaft Drive)

46
 GAS GENERATOR ROTOR

VA_

AIRCRAFT/ENGINE
POWERSHAFT INTERFACE

Figure 13. Candidate C&A Drive (Cluster Gearbox

Drive)
(2) Electrical
The candidate electrical C&A drives considered were:

1. Integral Starter/Generator - The integral starter/generator

illustrated in Figure 14 was discounted for the following reasons:

a. Weight - The combined weight of the starter/generator,

or generator, accessory drive motor, power processor,
and engine cables was estimated at 54 to 58 lb, without
cooling provisions, which is considered to be excessive
for this engine size.

b. Vulnerability - The physical volume of the starter/

generator, motor, and power processor was esti-
mated at approximately 500 in.", which would present
a large exposed critical area in a vulnerability analysis.
c. Tip Velocity - The tip velocity (60,000 ft/min) re-
quirements exceed the current speeds used in aircraft
electrical starter/generators and would require an ad-
vanced development effort for confirmation. Engineer-
ing development in this area is proceeding in the field of
portable power packs usinp smaller diameter genera-
tor rotors operating at tip velocities of 63,500 ft/min.

The development activity for this engine integration

application is considered marginal with respect to the
1977 development time frame goal.

d. Envelope - The envelope requirement dictates that the

starter/generator or generator would have to be lo-
cated upstream of the compressor, which would severely
Impact the inlet and IPS configurations due to the re-
quired diameter of 7.5 In. and the close axial location
relative to the compressor inlet.
47
 e. Maintenance - An integral starter unit would negate
or severely complicate field removal and replacement
of the starter unit.
f. Engine Critical Speed - The incorporation of the genera-
tor rotor would require moving the forward bearing
support for the power turbine approximately 6 in. This
would impose severe critical speed problems on the
power turbine. Similarly, the incorporation of a 6-lb
overhung mass (generator rotor) on the gas generator
imposes critical speed problems on this system. Inte-
gration of a generator IP this horsepower class is not
considered to be technically feasible.

2. Integral Generator - The system was discounted for the same

reasons outlined for the integral starter/generator, since the
electrical component sizes are similar.

ELECTRIC CONTROLS AND

ACCESSORY DRIVE MOTOR

-j-

INTEGRAL ];II
GENERATOR
STATOR ^.TR

INTEGRAL GENERATOR ROTOR

(MOUNTED TO GAS GENERATOR ROTOR)

Figure 14. Integral Electric Starter/Generator

(3) Pneumatic
Pneumatic drive schemes were considered in combination with the integral
starter systems only. The pneumatic drives in combination with a mechanical
starter system were discounted on the premise that a mechanical drive system
sized lor the required starting power would be capable of a C&A drive, with

48
 no compromise in the drive system design. The study of the individual drive
techniques is summarized below:

1. Cold Gas Bleed - Considered a candidate and is illustrated

in Figure 15

2. Bleed and Burn - Discounted on the basis of complexity

3. Interturbine Bleed - Discounted on the basis of vulnerability

4. Mixed Bleed - Discounted on the basis of performance loss.

-TURBINE
CONTROL MANIFOLD
") VALVE
'7 ^tAS'ti
/ X

Figure 15. Candidate C&A Drive (Pneumatic Drive)

A discussion of the analysis follows.

The bleed systems, shown schematically in Figure 16, were analyzed at mini-
mum and maximum engine power levels as follows:

1. Sea level ram/cold day (100% power) - (Maximum engine

pressures, airflow, fuel flow, rotor speed, and turbine
powers)

2. 20,000-ft ram/hot day (flight idle) - (Minimum engine pres-

sures, airflow, fuel flow, rotor speed, and turbine powers).

The engine conditions at these points are summarized below:

Engine Compressor Interturbine Ambient

Engine Power Airflow, Pressure, Temperature, Pressure, Temperature, Pressure,
Condition lb/sec psia "R psla °R psla

SLR/Cold (100%) 4.3 208 1110 63 2200 14.7

20,000-ft/Hot 1.25 21 925 10 1150 7.0
(Flight Idle)

49
 COLD BLEEDS

O/B

BURNER

BLEED AND BURN MIXED BLEED

Figure 16. Candidate Pneumatic Drive Systems

The assumed accessory drive turbine requirements were: 25 hp at maximum

engine conditions and 16 hp at minimum engine conditions. The program re-
quirement of 15 hp for power takeoff at all flight conditions was the dominant
factor In these requirements. Other ground rules for the accessory drive
turbines are shown below:
Turbine Speed - 20,000 rpm at both flight conditions
Minimum AP/P for control valve and lines = 10%
Single-stage turbine
Maximum useful turbine AP/P = 4
Maximum turbine mean diameter = 5 in.
Minimum blade height = 0.30 in.
A turbine for driving the controls/accessories has a small power turndown
(25 to 16 hp) between the two flight points investigated. This is in contrast to the
gas generator engine turbine of 915- to 97-hp turndown or the power turbine of
860 to zero delivered horsepower. (Flight idle is full speed, but no delivered
horsepower for the power turbine.)

Because of this trend, shown in Figure 17, the accessory turbines are sized at
the low engine power point. Therefore, at the maximum power condition, the
accessory turbine will deliver much more power than the required 25 hp. Throt-
tling the accessory turbine system (with a control valve or variable admission
turbine) to produce only the required 25 hp provides an inefficient turbine bleed
system.

50
 10

8
A
I GAS GENERATOR ^^^^ j
TURBINE -% ^^^^ 1
£ 6
POWER TURBINE —K^S^^^ \

o <S^<S CONTROLS AND

I 2
^S'^S' ACCESSORIES —v i

FLIGHT IDLE 100% POWER

(20,000 ft) (SEA LEVEL/RAM)
ENGINE CONDITION

Figure 17. Turbine Horsepower Turndown Ratio

The required bleed flows and engine ASFC are given relative to a conventional
tower shaft/gearbox for a constant power turbine delivered power and are shown
in Table 4.
TABLE 4. TOWER SHAFT GEARBOX TURBINE BLEED PERFORMANCE

Accessory System Turbine Engine

Drive Components Flow, lb/sec ASFC, %

1. Cold Valve and Fixed Turbine 0,644 + 8.3

(Exhaust Overboard)
2. Cold Variable Admission Tur- 0.875 + 5.9
bine
(Exhaust to Inter-
turbine)

3. Interturbine Variable Admission Tur- 0,465 + 8.0

bine

4. Bleed and Burn Valve Fixed Turbine, 0.277 (air) +6.7

Burner, and Fuel System +18,8 Ib/hr
(fuel)

5. Mixed Valve, Fixed Turbine, 0.40 (cold) +19.0

Mixer 0.70 (hot)

51
The two most promising systems of those Investigated were (1) the cold bleed
(exit overboard with a control valve) and (2) hot bleed (exit overboard with a
variable admission turbine). The reasoning behind this selection was:

1, Cold Bleed With a Valve - It has a low performance penalty

(+8,3% SFC) and small engine component changes, except for a
23% larger compressor. It Is a conventional system, which
has been used on other engines. With the üow exit overboard,
no return line to the engine is required,
2, Cold Bleed (Exit to Interturbine) With a Variable Admission
Turbine - Although this system has small engine effects, it is
complicated with a variable admission turbine and has a
vulnerable exit duct back to the engine interturbine,

3, Hot Bleed W'th a Variable Admission Turbine - It had good

performance nd small engine changes. It requires a variable
admission turbine and has a vulnerable supply duct from the
interturbine,

4, Bleed and Burn - This system had good performance and small
engine changes. However, it is a complicated system with a
separate burner, fuel system, and ignition system,

5, Mixed Bleed - This system's performance was bad and required

two ducts.

The use of cold gas for a C&A turbine drive was considered to be more reliable
and less prone to Introducing secondary damage In the event of a hit, as com-
pared to a hot gas system. In the event of failure or damage to the hot gas line,
the 2200oR Interturbine gas products would be discharged Into the rear engine com-
partment and would provide a source of ignition to anything combustible In the area.
The development risk and production cost of the hot gas components would be higher.
The interturbine bleed system would have a slight performance advantage and
might offer some weight advantage. Maintainability and installation flexibility
would be similar for the two systems.

Therefore, the Interturbine bleed was discounted In favor of a cold gas bleed
based on the estimated superiority in reliability, vulnerability, development
risk, and cost of the cold gas system,
(4) Hydraulic

Integral Hydraulic Pump - An Integral hydraulic pump, shown In Figure 18, was
not judged to be technically feasible on the basis of the minimum Inside diameter
imposed by the gas generator shaft. This physical limitation would require
dynamic seal velocities of approximately 500 ft/sec and an excessive pump
inlet pressure to suppress cavitation.

52
 STATOR HOUSING

VANES
POWER TURBINE ROTOR

DYNAMIC SEALS

GAS GENERATOR ROTOR

STATIC SEALS

Figure 18. Integral Pumping Element

(5) Hybrid
A discussion of the hybrid systems considered is provided below:

1. Any C&A Drive - Integral Alternator - An integral alternator

to provide 60 w power for the engine ignition and electronic
control was discounted. An Integral generator was considered
to be practical from an engine Integration standpoint, but
offered no vulnerability, weight, or reliability advantages over
a generator that could be Incorporated as part of a lower speed
accessory or component. Separation of the generator from
the electronic control unit was also not considered to be ad-
vantageous.
2. Mechanical/Electrical Interface - Considered a candidate with
an external electrical starter.

3. Mechanical/Rieumatic Interface - Considered a candidate with

an external pneumatic starter.
4. Mechanical/Hydraulic Interface - Considered a candidate with
a hydraulic starter.

5. Auxiliary Power Unit - An auxiliary power unit for the C&A

drives was discounted. This unit would consist of a JP-fiieled,
internal combustion engine that powered a C&A module. This
system was eliminated on the basis of reliability, where the
successful operation of the engine would depend on the functioning
of two separate powerplants.

6. Mechanical Drive - Electrical Fuel Pump Interface - A

variation of the mechanical drive systems was considered as
a candidate; it used an electric generator/motor coupling for
only the main fuel pump. This would simplify the flow control
fuel bypass and subsequent heat rejection problems, by allowing
an infinite variation of fuel pump speed u. meet the engine
requirements.
53
(6) Integral
Fuel and oll system pumping elements integral with the gas generator rotor were
discounted because of the excessive inlet pressure requirements (up to 1000 psi)
to suppress cavitation. The high inlet pressure requirement is a result of the
65,000-rpm operating speed and the tip speeds, which are established by the gas
generator shaft diameter.
Oil pumps operating in a slinger configuration as an integral part of the engine
rotors were briefly considered and discounted for the following reasons:
(1) charging the center of the shaft(s) would require additional high velocity
dynamic seals, which would add to the heat rejection and system complexity;
(2) additional heat rejection to the oil would result from friction losses due to
the relative shaft velocities; (3) there are technological unknowns with this
pumping scheme, and they were considered beyond the development time frame
for this contract,
b. Starters

(1) Mechanical
This category Includes those starter systems that have a mechanical Interface
with the engine. These candidates, Illustrated In Figure 19, are:

1, Electrical - Considered a candidate; APU/generator and

battery supplies considered
2, Hydraulic - Considered a candidate; APU/hydraulic pump and
accumulator blowdown supplies considered

3, Pneumatic - Considered a candidate; APU bleed and APU/

external combustor supplies considered

4, Self-Contained Starter - Considered a candidate; will not be

considered with the cluster gearbox C&A drive because of
packaging. Gas turbine, piston and Wankel engines considered
as candidates,
(2) Integral

This category Includes those starter systems that do not require a mechanical
Interface with the engine. Three Integral pneumatic systems were considered
for further evaluation and are Illustrated In Figure 20. A discussion of all the
Integral start system candidates Is provided below:
1, Integral Electric - An electrical starter integral with the gas
generator rotor was sized for this application and was approxi-
mately the same size as the integral starter/generator
previously discussed. This system was discounted for the
same reasons outlined for the integral electric C&A drive
techniques.
2. Integral Hydraulic - An integral hydraulic starter was discounted
because of the requirement for high-pressure dynamic seals on

54
 the gas generator rotor. These seals would be required to
direct the high-pressure, externally supplied fluid through a
start turbine and would operate in the nonstarting mode at
rubbing velocities of approximately 500 ft/sec due to the gas
generator shaft diameter. The seals would require a lift-off
feature, or would generate heat during engine operation above
start and would require cooling. This overall system was
not considered to be compatible with the development time frame.

3. Integral Pneumatic - Considered a candidate. Three integral

pneumatic systems were established: cold gas impingement,
cold gas supply using heat addition in the gas generator burner
with a closed engine inlet, and hot gas impingement using heat
addition in an external burner.
4, Cartridge - Cartridge starters for the primary starter mode
were discounted because of the logistics involved in supplying
starter grain assemblies. In addition, manual replacement of
starter grains between starts or the development of a multistart
cartridge grain feed system would be required. The potential
hazard of carrying grain assemblies onboard a military aircraft
was also considered as a disadvantage,
AIRFRAME/ENGINE INTERFACE
! k

BATTERY
üUÜ STARTER MOTOR

HYDRAULIC
ACCUMULATOR STARTER MOTOR

STARTER MOTOR

APU
1. HYDRAULIC
2. PNEUMATIC
3. ELECTRIC
SELF-CONTAINED STARTER

FUEL

Figure 19, External Starter Systems

55
 AIRFRAME/ENGINE INTERFACE

APU

(a) COLD GAS IMPINGEMENT

(b) ENGINE RAM/CLOSED IGV'S

APU

HOT GAS
IMPINGEMENT

Figure 20. Integral Starter Systems

3. Selection of Five Candidate Systems

Of the 10 basic systems selected for further evaluation, there were 26 Individual
C&A/starter concepts. It was required to reduce the candidate configurations
to five basic systems, using the established performance and rating criteria.
The logic network used to screen the candidates Is Illustrated in Figure 21 and
Is summarized below:

1. Starters

a. External
• Select best airframe-powered starter (electric,
hydraulic, or pneumatic)
• Select best self-contained starter (gas turbine,
piston or Wankel)
• Compare and select best airframe-powered starter
vs best self-contained starter.

b. Integral
• Select best integral start mode (cold gas impinge-
ment, hot gas Impingement, engine ram/closed
IGV's).

56

• \
 u
< «

oc
GO
> =
z ** < i O
J oc z o
M < O O S

:;3

u<
i
.<-
zo
ui
S
w <
3 Siz
K
o <
z
□
9 CO

<
u
i
■t->
en
z
.<.
I zu
Ui
>
S
< _i
Ü Ui

|£
zo
|
Ü

ü I
U
a
.ah
re

—*
g3 &£ 15 i
3 "ö £«• I* be
o Ui Ui 2
Ui a.
eg
1V3INVH33W 1VU931NI

SU31UV1S
57
,
 2. C&A Drives

• Select best of mechanical drive with electric fuel pump

drive vs an all-mechanical drive.

• Select best of mechanical drive vs pneumatic drive.

• Select best hybrid drive (electrical, hydraulic, or

pneumatic).

Based on the analysis, the five candidate systems outlined below were selected
for further study;

Tower Shaft Drive/Gas Turbine Starter

Tower Shaft Drive/Cold Gas Impingement Starter
Cluster Gearbox Drive/Air Turbine Starter
Cluster Gearbox Drive/Cold Gas Impingement Starter
Hybrid Hydraulic Drive/Hydraulic Starter.

The analysis is summarized in the following paragraphs.

a. External Starters

(1) Airframe-Supplied Power

A study was conducted to evaluate six different engine starter concepts for a twin-
engine helicopter application based on the weighted assessment criteria previously
established. The results of the study are summarized in Table 5 and show that
the APU-pneumatlc had the highest rating.
As indicated in Table 5, three of the starter concepts are powered by an auxiliary
powerplant and use hydraulic, pneumatic, and electric starter motors, respectively.
The fourth system uses a battery-powered electric starter and the fifth, an accum-
ulator blowdown hydraulic system. The last system is a combustion/turbine starter,
which requires a pressurized supply of air and fuel for the combustor, which has
an integral turbine starter.

Table 6 summarizes the data obtained from the references and used in this
evaluation. An explanation of how the assessment weightings were determined
follows:
1. Reliability - MTBF's were computed for each starter concept
by averaging the MTBF's given in the references for each com-
ponent when more than one data point was available.
The weightings were computed on the basis of giving the lowest
MTBF system 23% (maximum weighting for reliability) and
the other systems a lesser percentage proportional to their
MTBF's.

58
 CO 05 o CO o 5£> o

I!
to •- •
• • • • • •
00 lO CO t> lO w Ol o
CO

in o o t> t0 00
. • •
lO -^ (N •*

^1
CO r-t o o (N o to o N
• • • • • • • • •
00 00 lO in to
50

§
p
Ü
3< a»
w
o m o o
• • • •
00
•
o
•
CD
i
05

4-» I m o o •* in a> 't

> U p (C
B a.
en <
w .a
Ü
13
§ O O o o o o
05 o O 05 m
iM 00
i
pi
<

w
PQ
<
1 CO
iH
•
in
00
•
o o in
o
• • 05'
0
OJ
•
m
•
CO
0
•
co
00
00
to
•

0 0 O O O O O O
•a o • • • • « • • •
o
CO 05 O 05 to o
r-t

-a
o
> 0 3
£ 1 "9
o 3
* 3 i
Is ■ e .e "5132
5
•a si^ -ö
u 0 I to
o
II 1•aS 3«
is«
5S
o
H
> Ü

59

j
 S
II
a5 S!2
3 x
iJ 111
3 5 u
• g-a « E
as £8^ -s

SC

S1

11
So"
181 iifl
ii
a £ o E
23
II
o E
III «3•f £2
O M

<
h ■B r-.
e
•*■ n ö t- > ©l-r
55 ij o o o ©|c ' © ©' d ©loj
Q BC-:
'S a ■*• a> © t- ©I©
£
o
—*
H I
<
P 1! 32 M
3 5«- 5
E

w
1
u 1 It i
u Si2
z a
o
ü
«
H
H >i

tf
< 'S

S
5^ E Ef
— 9
0 Q ,„
> C

PQ
<
5 fi -r © M|(-
• • I . • .
M O ^ ©|« M « ©'|io ©' V ©lirt

«■s

u 2.
lI sg£a
ao ■ ' £
P 15_
fill
re II IS u
u 0 i i "i -c
as a.
3 u is c Gil l!-
4 ►. 5 • I
III
B s
i l||
«wSl E
UJ « d S rt i
S<äi! U, U U
<a-
IISS n a t; 85

60
i

2. Vulnerability - Vulnerability was computed to be proportional

to the volume of the starter. This was done because the
probability of small-arms fire hitting any section of a helicopter
is proportional to its exposed area. Nothing was done in this
evaluation in consideration of locating components on the less
vulnerable areas of the helicopter or secondary damage (fire)
that could result from a hit. Those concepts that require fuel
for their operation will be more vulnerable when secondary
damage is considered.

3. Development Risk - The weights given each system for develop-

ment risk were based on engineering judgment. All of the
concepts being considered are state of the art, and the de-
gree of development risk will be related to the weight and
size reduction and performance design goals of each sys-
tem. The starter concepts that use an APU were judged
to be a little higher development risk primarily because
the APU size and weight and performance are based on
1975 technology. However, none of the concepts are con-
sidered a high development riak.

4. Cost - Some cost data were available in the referenced re-

ports; however, the weights given in Table 6 represent estima-
tions of the relative costs. The cost of the other concepts was
estimated in proportion to their relative complexity as com-
pared with the APU systems.
5. Weight and Volume - The assigned weightings were determined
by separately assessing weight and volume for each system based
on the lowest weight, with the lowest volume getting 10 points
and higher weights and larger volumes a proportionally lower
number of points. The system with the highest combined point
total for weight and volume was weighted 10, which is the
maximum for this design parameter.
6. Performance - The ability of the APU-powered starter concepts
to perform the starting function was judged to be equal. The
battery electric system was evaluated the lowest because of
Its cold-day limitations. The accumulator-hydraulic system
and the combustion starter system require pressurized oil,
pressurized fuel, and air supplies, respectively, which will
also have cold-day limitations. These limitations can be
taken care of by cold weather pressure topping using hand
pumps, but the performance was judged to be lower for this
reason.
7. Maintainability - The maintainability of the starter concepts
was evaluated based on multiplying the reliability by the
average maintenance hours per maintenance action as de-
termined from the references.

8. Installation Flexibility - Battery-electric starter was con-

sidered best since the starter has good flexibility In mounting
to the engine and requires only a single electrical connection to

61
 the airframe. All other systems required additional and
more complex connections for airframe-to-engine interface
and were downgraded accordingly.
(2) Self-Contained Starters

A survey was made to evaluate the potential of using an engine-mounted, self-

contained, JP-fueled starter. Three basic engine types, gas turbine, piston
engine, and the Wankel, were considered.
A starter specification was generated and several manufacturers were surveyed.
The specification required the use of JP fuel and a provision for a manual secondary
starting system.

Assuming that cost and development interest were not overwhelming disadvantages,
the three candidates were compared as shown in Table 7 and rated as shown in
Table 8.
The study showed that a gas turbine starter (GTS) would provide the lightest, least
vulnerable, and most reliable overall starter package, but would require an expen-
sive development effort. Packaging of a GTS may also be a limitation in certain
installations.

TABLE 7. SELF-CONTAINED STARTER EVALUATION RATING

Weicht Starter
Criteria Factor Gas Turbine Piston Engine Wankel

Reliability 23 23 12.7 17.0

Vulnerability 19 19 9.8 4.5
Development Risk 14 14 10.0 12.0
Cost 12 4 12.0 8.0
Weight and Volume 10 8 10.0 9.0
Performance 9 9 6.0 6.0
Maintainability 7 7 3.0 5.0
Installation Flexibility 6 6 5.0 5.0

Total 100 90 68.5 66.5

62
 I
o 5
is
c
c B
E C L
si
II
av
IE
B 2

ur
l|
us
H
2 S

is

t)
o
e
a«
£
Pi fl|
i
u
a.
o £ lit
m us
c

111
o o o
SSIS O t-< iH

o
Si0- o
US
Ü C 1-4
Li '/>
o

e
E

> 3) 0
IH B
m 0
^ B

o
>

s
u.
-i n f IM rj in
3 o t-* e t~ in o in
«_ u

II

9) X x
u
Clutc

3
5
nc

o
c •a
c -a — "2
1 u as
«» tfl ^ S M s
5 Jaü 8B -g8
■
Geai
Engi
;ary-

a V B S)
«O iwu
H ■-» E ?
Sa H w^. it 2i«2,

63 I
 The results of the survey and discussions with vendors are summarized in the
following paragraphs.

a. Gas Turbine Starters

Gas turbine starters (GTS) In the 16-hp class are not being produced, but could
be developed for a specific application. Development, qualification, and pro-
duction costs would be relatively high, considering the requirement for an altitude
relight. For most applications, an APU is used for engine starting and also
provides other functions such as power generation or environmental conditioning.
The development and production cost Is, therefore, spread over several functional
requirements.
The GTS configuration considered would Incorporate a planetary reduction gear
drive and an overrunning clutch. A hydraulic accumulator blowdown start system
would be used. This configuration also allows the added flexibility of cold
cranking the main engines to check fuel and electrical systems.

b. Piston Engine
Several commercial piston engine manufacturers were contacted but showed no
Interest In a starter development effort. The primary reason given was limited
production quantities. No major technical difficulties In using a plston-englned
starter were uncovered.

Starting at -650F with JP fuel would require fuel Injection or the use of a fuel
preheater. A dual-clutch arrangement would be used. A centrifugal clutch
would allow a smooth acceleration of the engine, and an overrunning clutch
would be used to decouple the starter.
j;. Wankel Engine

Curtiss-Wrlght Corporation was visited to review the use of a Wankel engine for
a starter application. No technical difficulties were uncovered, during the dis-
cussions, which would limit the application of a Wankel. Little enthusiasm was
shown for participation In a development effort for this application. Again, the
reasoning used was the limited production quantities. Weight and volume
estimates were provided.

An air-cooled Wankel was considered that dictated the use of large cooling fins
and influenced the size and weight estimate. Starting at -650F would dictate
the use of a fuel injection system. The operating speed considered was 12,000
to 15,000 rpm and would require development of a 3000-psl fuel injection system.
The reliability estimates for the Wankel were based on the relative complexity
as compared to a conventional piston engine.
(3) Airframe-Supplied Power vs Self-Contained Starters

Based on the previous study information, the best engine-mounted, self-con-

tained starter was then compared with the best engine-mounted starter that Is
supplied with power from an external source. In this study, the GTS was com-
pared with the APU/pneumatlc. The relative comparison Is summarized in
Table 9 and shows the GTS with a slightly higher rating.

64
 TABLE 9. STARTER CONCEPT EVALUATION

Starter Concept - Rating

Weight APU
Criteria Factor Pneumatic Gas Turbine

Reliability 23 21.0 23.0

Vulnerability 19 19.0 19.0

Development Risk 14 14.0 14.0

Cost 12 12.0 12.0

Weight and Volume 10 10.0 7.4

Performance 9 7.0 9.0

Maintainability 7 6.6 7.0

Installation Flexibility 6 6.0 5.0

Total 100 95.6 96.4

These systems were rated closely, primarily because of the similarity of the
components. Both concepts use a gas generator and a turbine/gearbox for trans-
fer of energy to the engine. The GTS configuration uses a dedicated system for
each engine, while the APU/pneumatlc uses a common gas generator and separate
air turbine starters. It should be noted that the cost of the APU and GTS systems
were rated similarly, assuming a specialized development effort for each system.
This will not be a valid assumption If the APU Is used for other aircraft functions
and the cost Is shared.
For this program, the GTS was used where the envelope allowed. For envelope-
limited applications, or in installations where an APU is required for other air-
frame functions, the APU/pneumatic system is recommended.

b. Integral Starters

Three candidate systems were evaluated that provided means of starting the main
engine without requiring a mechanical connection to the engine rotor. These
systems use bleed air supplied by a separate APU and are Illustrated In Figure 22.

65
 M
>
o
^

8J
2
^ Wi
2
M ^
oo oc
s
J
_i

V)
*%
Ui
Z
o

I
i
z f :

a |
5X 1
ji <

P 1
cx
<
h
tt
^- t
z
Ui
2
09
> 0)
•
UI
o | •
z be
c
s Si

1
«3
Q
Q
<J • «
(N
e« <N
(U
C
3)
PM

66
Air supplied by a pressurized bottle system was briefly considered but was
discounted on the basis of size, weight, and the potential hazard involved with
high-pressure pneumatic bottles subject to a military environment. A comparison
was made of a pressurized bottle system and an APU. The following APU was
chosen for comparison:

Bleed Flow = 0.63 Ib/s ■

P/P = 6.8

Volume = 0.47 ft3

Weight s 42 lb
Torque at Ignition = 5.2 ft-lb

Since the total f.as mass (flowrate x time of starter operation) must be carried
In the bottle, tTie starter torque was varied to reduce total mass of air, as shown
in Figure 23. Even though flowrate Increases with starter torque, the time to
starter cutoff and total mass flow decreases.
Increasing tank pressure reduces the required tank volume, as shown in Fig-
ure 24. Very high pressures are necessary to obtain a volume competitive
with an APU.

u.
u.
g
Ü

cc
s
O
t-
UJ
5

4.8 5.2 5.6 6.0 6.4 6.8

STARTING TORQUE AT IGNITION -

Figure 23. Impingement Starting Characterl^lcs

67
„1
 UJ
S
3

Z
<

<
Z
CC

2 4 6
TANK PRESSURE - psi (Thousands)

Figure 24. Tank Volume Requirements for Im-

pingement Starting Gas Required for
Two Starts

(1) Cold Gas Impingement

This system provides Input energy to the gas generator rotor by impinging on
the rotor surface with a high-velocity gas supplied by an external APU. For
this application, several locations for the impingement area were considered.
A location at the rear of the gas generator turbine rotor was selected and is
shown in Figure 25. The addition of air at this location has a minimal effect on
normal engine operation, as opposed to an injection point on the compressor,
where the externally supplied air might Interfere with normal compressor
operation.

6S
 TURBINE BUCKETS
TURBINE STARTER NOZZLE

GAS GENERATOR
TURBINE BLADE

GAS GENERATOR
EXIT GAS

STARTER INLET
SUPPLY GAS
SECTION A-A

Figure 25. Impingement Starter

The characteristics of the selected coid gas impingement system are outlined
below:

Gas T = 600oR
Impingement Surface
No z 3 fa» Diameter = 5.4 i&.

Mach Number 1.0

Pressure Ratio 1.893

Flowrate 1,226 lb/sec

Total Orifice Area 1.928 in.
(2) Hot Gas Impingement

This system Inputs energy to the gas generator rotor by impinging on the rotor
surface with a hot gas, which Is supplied by an auxiliary combustor. The com-
bustor is engine-mounted close to the injection point and Is provided air by an
external APU. A schematic of the system is shown In Figure 22. The system
requires the additional complexity of a separate ignition and fuel control system.

69
The characteristics of the hot gas impingement system are outlined below:

Mach Number 1.0

Gas Temperature 246011
Pressure Ratio 1.893
Flowrate 0.58 lb/sec
Total Orifice Area 1.348 in?
t
Mean Diameter of
Turbine Buckets 5.4 in.
(3) Engine Ram/Closed IGV's

This starting method uses bleed air from an APU for supplying air to the engine
burner during start. Air introduced Into the engine burner Is forced through
the engine turbine by closing the compressor Inlet area. The compressor IGV's
are used to close the compressor inlet, as shown In Figure 26. With the com-
pressor inlet closed, work required to pump Inlet air Is eliminated and, there-
fore, reduces the horsepower required to rotate the engine rotor to any specific
speed. The air Introduced into the engine combustor is capable of rotating the
engine up to minimum fuel pumping speed, at which time fuel from the engine
fuel system Is also Introduced Into the combustor and Ignited. Additional
energy available from the combustion gases will now accelerate the rotor above
a self-sustaining (34,000 rpm) rotational speed. The relationship of torque
available vs torque required for the rotor parasitic losses Is shown In Fig-
ure 27. The auxiliary air supply Is now removed from the engine combustor,
the IGV's are scheduled to their normal run condition, and the engine is accelerated
to ground Idle. This transient Is Illustrated on the compressor map In Figure 28.
Since the transfer of air supply from the APU to the engine compressor requires
some transient time, the Initial speed at which the transfer Is .nltlated must be
high enough to allow for "coast-down1' during the transient and retain an above-
self-sustalned speed at the completion of the transfer. The expected engine
coast-down characteristics are shown In Figure 29. An APU flowrate of 0,42 pps
at a pressure ratio of 1,9 will be required,
(4) Rating of Integral Starters

The numerical ratings for the three systems are shown in Table 10. The cold
air impingement system received the highest rating (82,3), primarily because
of the simplicity and proven capability. The engine ram with closed IGV's re-
ceived the next highest rating (81,6) and was downgraded primarily because of
development risk. The hot air impingement system received the lowest overall
rating (72,3),

70
UJ UJ Ik lU Ui

piss
a>

I
U
0)
tn

U
&

w
P
00
(D
c
is
>
ID

a»

£
71
UJ

§
o

10 20 30 40 50
GAS GENERATOR ROTOR SPEED - rpm (Thou$ands)

Figure 27. Torque Characteristics of Ram Start/

Closed IGV's

0.5 1.0 1.5

COMPRESSOR CORRECTED FLOW - lb/tec

Figure 28. Compressor Map of Ram Start/Closed

IGV's

72
 60
-
50

1
> 40 y
30 \
I
20 \

10
V^
10 20 30 40 50 60
TIME - tec

Figure 29. Engine Coast-Down Characteristic of

IGV's Open
It is worthwhile to note that if the engine ram with closed IGV's was experimentally
proven, the overs>il rating would be better than the cold air impingement system.
The system would'require slightly more engine complexity, but would reduce the
airframe APU size by 66%, as compared to the cold gas impingement system.
Experimental work in this area is justifiable, A discussion of the ratings is
provided as follows:

1. Reliability - The cold air impingement system was considered

to be the simplest and most reliable. The hot air impingement
system was downgraded because It requires a separate igniter
and fuel system for the small burner. The third system (with
closed IGV's) requires sequencing of the IGV's and APU shut-
off valve.
2. Vulnerability - The criteria used were volume and number of
components (Table 11). The engine ram with closed IGV's was
considered as the baseline (volume = 0.25 ft3). The cold air
impingement system had a large volume (1,08 ft3) but a fewer
number of components. The hot air impingement system re-
quired a burner plus a fuel and ignition system for both the
engine and the APU and liad a volume of 0,58 ft3,
3. Development Risk - The cold impingement was considered the
least risk. This system will start the engine; the only problem
is the interface with the engine/rotor for an optimum engine
system. The hot air impingement will also start the engine,
but, in addition to its interface with the engine/rotor, the
burner must be developed. The engine ram system with
closed IGV's will require experimental evaluation to define
the compressor recovery characteristics for the transient
when the IGV's are opened. The compressor must recover
from an unstable or stalled condition with the inlet closed
and establish a normal compressor-engine operating line
during the coast-down period.
73
a
g
h
p I3 00
o
<3J
«5

a»
o
o
.H
•
00
CO

O
•
«D
to

00

s
2
<
20 S
5
Uffl
II 00

o
co 00 Oi
•
o
•

W
K
"*• w «> O o o <D N n
• • • • • • • •
"5
I
ai
fa

o
I oo in

tf

i c
I
w
II 00 C5 m o 00
s c
o o o
•
ft o
•

a I
B a
& o in o o O 00 n
• •
•a 8,-ß n OS
• • •
05

I I 1-1 r-l oo

eg
II
CO

52;
w moo CO O O 00

0 O rH rH O iH iH O
3

<: o> (M OS (0 o
o

£
Q
w

PQ
0) i
4J
s 1>
3PQ I ^ "S Ü
0)

3
1 i1 1 a c
1 o
s
§
ao 1 1
I
0)
"a 1 i
> I s
74
 CO

,9 5 <D
ca»
fa IN
C
•a i
c c 04
w
li
t
in w w U5•+
in
rH
0 05
«)
0 IM
5
UPQ

s
O u
a
w
< S.S e x:
u 9 -P w ü
ed
2o ^^ 00 W
0
S5 • a
O
H
S
11 \
O 00
w m
t>
05
m
0

I
to Go ■t <n 0 0 0 LO
< c • • • • • •
SM O M
M
0 iH
w
w 0 t»
^
00
IN
O 00
t>
'S Ö
<
>
c
o !§0
■s
w s
a a
<
•a S.S
to
CO

2O ^
as IN

w cn 00 00
(0 IN o m o in
•
u M ^ lH Z Z ^
w m in m
2 a T3
0)
3
w 'S "O
.S "a

8 g
o
c
I (1)

be
fc to c
I 5 < < 5
,.Q a ja
'C
0

0 *■>
tt
0
to
'S -ft i _f U i -M to P
QJ OH

^II 73
Q
«
<
e
Q
s a
a h 2 f3 - & 0)
ex .3
ti 73 W < 1-1
rH N

75
 4. Production Cost - The cold impingement has a larger API),
but everything else is standard. The hot impingement has a
smaller APU, but requires a separate burner, ignition system,
and a fuel pump. Tho engine ram requires the smallest APU,
but more expensive IGV's and actuation system.
5. Weight and Volume - The sizes Include the APU's accessories,
except for their starter systems. For the hot air impingement
system, the burner was considered as being mounted on the
engine; therefore, two would be required for a twin-engine In-
stallation.
6. Performance - The engine ram system with closed IGV's was
downrated because of the potential of a slight performance
penalty due to the IGV design,
7. Maintainability - The hot air Impingement system was down-
graded because of the large number of components (burner,
fuel nozzle, fuel pump, motor, and Ignition system). The
engine ram system with closed IGV's was downgraded slightly
due to the special IGV's.

8. Installation Flexibility - Volume and number of components were

used for the criteria. The engine ram had the smallest volume
and thus was the baseline. The cold air Impingement was
downgraded because of Its larger APU volume. The hot air
Impingement system was graded lowest because of volume and
the requirement that the burner (10 In. length and 5 In. diam-
eter) be mounted on the engine to prevent long, hot lines.

c. Pneumatic vs Mechanical C&A Drive

The cold gas bleed pneumatic turbine accessory drive system previously selected
as the best pneumatic drive was compared with a tower shaft accessory drive
system using the selected weighting criteria previously established.

The tower shaft system received a rating of 92.1% vs a rating of 69.1% for the
cold gas bleed system, as shown In Table 12. Even though the cold gas bleed
system was compared only to a tower shaft system In this study, a comparison
with any other direct mechanical drive system such as a cluster gearbox arrange-
ment would yield similar results because mechanical drive systems. In general,
have similar advantages and disadvantages. Based on this analysis, the cold
gas bleed system was eliminated as a candidate accessory drive system.

7b
 TABLE 12. CONTROLS AND ACCESSORY DRIVE
SYSTEM EVALUATION RATING
Pneumatic vs Mechanical C&A Drive

Tower Shaft Pneumatic

Weight Rating Rating

Criteria Factor Factor Rating Factor Rating

Reliability 23 1.0 23.0 0.60 13.8

Vulnerability 19 1.0 19.0 0.80 15.2

Development Risk 14 1.0 14.0 0.75 10.5

Cost 12 1.0 12.0 0.40 4.8

Weight and Volume 10 1.0 10.0 1,00 10.0

Performance 9 1.0 9.0 0.20 1.8

Maintainability 7 0.3 2.1 1.00 7.0

Installation Flexibility 6 0.5 3.0 1.00 6.0

Total Rating 100 92.1 69.1

The better of the two systems in each of the selection criteria is given a factor
of 1.0, and the other system is given a factor that indicates Its ranking in that
category relative to the other. Figure 30 illustrates tower shaft and pneumatic
turbine drive systems used for this study. The tower shaft system showed ad-
vantages in all areas except maintainability and installation flexibility. The
reliability of the pneumatic system was downgraded due to the requirement for
a pneumatic control valve and speed control loop.
d. Electrical vs Mechanical Fuel Pump Drive

Consideration was given to an electrically driven variable-speed pump as opposed

to a conventional mechanically driven pump. The potential advantage of the
electrically driven pump would be in simplification of the control system, In that
the pump speed would be infinitely varied to set the desired main engine flow
Instead of an in-line throttle valve or a bypass valve. The electrical drive system
will reduce the fuel pumping inefficiency and the heat rejection to the fuel. The
drive systems were evaluated as shown in Table 13.

The mechanical drive was considered superior in all areas except performance,
maintainability, and installation flexibility. The electric motor drive has the
disadvantage of size and weight, which also impact the vulnerability assessment.
The mechanical drive was selected for further consideration.

77
I I

cc
Q

01

I
i
g
u.
<
I
ec
IU

i
o
00
IU

I
I*

78
 TABLE 13. FUEL PUMP EVALUATION RATING

Mechanical vs Electrical Driven Pumps

Mechanical Drive Electrical Drive

Selection Weight Rating Rating

Criteria Factor Factor Rating Factor Rating

Reliability 23 1.0 23.0 0.9 20.7

Vulnerability 19 1.0 19.0 0.8 15.2

Development Risk 14 1.0 14.0 0.9 12.6

Cost 12 1.0 12.0 0.6 7.2

Weight and Volume 10 1.0 10.0 0.5 5.0

Performance 9 0.5 4.5 1.0 9.0

Maintainability 7 0.8 5.6 1.0 7.0

Installation Flexibility 6 0.5 3.0 1.0 3.0

Total 100 91.1 79.7

e. Hybrid C&A Pump Drive Systems

Three hybrid drive systems, shown in Figure 31, which used either a hydraulic,
pneumatic or electrical interface between a tower shaft drive and the accessory
package, were considered.
INTERFACE
HYDRAULIC, PNEUMATIC, OR ELECTRICAL

MECHANICAL DRIVEN
(HYDRAULIC PUMP,
r ACCESSORY
PNEUMATIC
\ DRIVE MOTOR
COMPRESSOR, OR
ELECTRICAL GENERATOR)

Figure 31. Candidate C&A Drive

79
The hybrids offered flexibility in the location of the C&A components in that only
the drive motor must be located in the vicinity of the tower shaft. In all three
cases, the drive motor also was used as a starter for the main engine.

The electrical system was discounted on the basis of the starter system studies,
which showed that an electrical system had disadvantages in overall size,
performance, and weight. In addition, the size and weight of the engine-mounted
starter/generator and drive motor were not competitive with those of other sys-
tems In this horsepower class.

The hydraulic and pneumatic hybrid systems were studied in greater detail,
since the hydraulic and pneumatic start systems were rated closely. The
hydraulic starter/pump and drive motor and the pneumatic starter/compressor and
drive turbine systems are shown schematically in Figures 32 and 3?, respectively.
The mechanical/hydraulic hybrid system was selected for further evaluation.
A relative ranking of the two systems Is shown In Table 14.
A discussion of the relative rankings of the two systems Is shown below:

1. Reliability - The relative complexities of the two systems are

considered to be similar. The pneumatic system Is rated
higher due to the Insensltlvlty to contamination and the demon-
strated Improved reliability of pneumatic drive components.
2. Vulnerability - The pneumatic system was downrated because
of the large-diameter (approximately 7 in.) starter/compressor
and drive turbine required.

3. Development Risk - The pneumatic system was downrated

due to the development risk associated with the required con-
vertible starter/compressor.
4. Cost - The relative costs of the two systems were judged to be
similar.

5. Weight and Volume - The hydraulic system has the largest

weight penalty, but the pneumatic system requires the largest
volume. The hydraulic system was rated slightly higher on
an overall weight and volume basis,
6. Performance - The hydraulic drive system (pump and motor
combined) had the best overall efficiency, 73%, as compared
to the pneumatic, 50%.
7. Maintainability - The pneumatic system was rated easier to
/
maintain.

8. Installation Flexibility - The hydraulic system has the best

Installation flexibility based on smaller components.

80
■

oc
<.

NNNN»» ■ -NN^NWN^

(1)
—-

tn
0)
u

1
.my^w
^ff CO
s
35
1^

CM

I
81
 tu
s

0»
>

i
CO
co
0)
ü

CO

82
y
 TABLE 14. HYBRID CONTROL AND ACCESSORY DRIVE SYSTEM
Hydraulic vs Pneuiratic Evaluation Rating

Hydraulic Pneumatic

Weight Rating Rating

Criteria Factor Factor Rating Factor Rating -

Reliability 23 0.8 18.4 1.00 23.0

Vulnerability 19 1.0 19.0 0.60 11.4

Development Risk 14 1.0 14.0 0.80 11.2

Cost 12 1.0 12,0 1.00 12.0

Weight and Volume 10 1.0 10.0 0.90 9.0

Performance 9 1.0 9.0 0.80 7.2

Maintainability 7 0.8 5.6 1.00 7.0

Installation Flexibility 6 1.0 6.0 0.18 4.8

Total 100 94.0 85.6

4. Selection of Two Candidate Systems

Five control and accessory drive/starter systems were evaluated during this
phase of the program. Preliminary design layouts of the five systems were made,
and the systems were evaluated using the performance and rating criteria estab-
lished In the Requirements Definition Phase. The control, fuel pump, PTO, and
oil system components were considered to be common for this study. The five
candidate control and accessory drive/starter configurations evaluated are out-
lined below.

a. C&A Component Description

(1) Tower Shaft C&A Drive/Gas Turbine Starter (TS/GTS)

A tower shaft-driven gearbox is used to drive a 65,000-rpm combination fuel pump

and alternator, a 20,000-rpm PTO, and a 10,000-rpin oil pump. A self-contained
gas turbine starter is mounted on a 20,000-rpm engine gearbox pad for engine
starting. The starter consists of a gas generator, a hot gas power turbine, and a
gear reduction system for the power turbine. An overrunning clutch Is used to
decouple the starter. This configuration is shown in Figure 34.
(2) Tower Shaft C&A Drive/Impingement Starter (TS/IS)

A tower shaft drive, as described in system No. 1, is used for the C&A com-
ponents. Starting is accomplished by a cold gas impingement starter integral
with the gas generator turbine. The starter is powered by an airframe-mounted
APU, with a compressor discharge (cold gas) bleed. This configuration is shown
in Figure 35.
83
 CO

XI
o u
z
<
a en
CO
< Ü
z
<
I-
I
UJ , p-

o I
E

§ I
8 CO

s,

84
 e
a
Ba>
cbe
a
B
v.
X
su

s
2

85
(3) Cluster Gearbox C&A Drive/Air Turbine Starter (CGB/ATS)
The drives for the C&A components are provided by gears that are clustered
around the gas generator drive shaft. The C&A component drive speeds are
the same as described in system No. 1. An air turbine starter is used and is
mechanically connected to the gas generator rotor through the cluster gearing.
The air turbine starter consists of a high-speed turbine and a gear reduction
system to couple with a 20,000-rpm pad. An overrunning clutch is used to
decouple the starter system. This configuration is shown in Figure 36,

(4) Cluster Gearbox C&A Drive/Impingement Starter (CGB/IS)

The C&A drive is the same as described in system No. 3, and the starter system
is the same as described in system No. 2. This configuration is shown in
Figure 37.
(5) Hybrid Mechanical, Hydraulic C&A Drive/Hydraulic Starter (HD/HS)

The hybrid C&A drive used consists of a hydraulic starter/pump that is connected
to the gas generator shaft by a tower shaft, and a hydraulic motor, which is used
to power the C&A components. An accumulator blowdown system is used during
the start transient to drive the hydraulic starter/pump and for powering the C&A
drive motor. After the engine is self-sustaining, the hydraulic starter is switched
to a pumping mode and is used to drive the C&A components. The configuration
is shown in Figure 38.
b. Analysis

The five systems were assessed for each of the established rating criteria. The
results are summarized in Table 15 and show the CGB/ATS starter and the TS/
GTS as the two systems selected for detailed design. The ratings are discussed
below:

1. Reliability - A reliability assessment was made for both the

starter system and the C&A drive. The starter system reliability
was given a 20% weighting and the C&A drive an 80% weighting.
The results of the analysis are summarized In Table 16.
The cluster gearbox drives were rated as being more reliable
than the tower shaft drive because of fewer components In the
drive train. The CGB/IS and CGB/ATS systems were rated
closely together and ranked first and second, respectively.
The TS/IS and TS/GTS systems were also rated closely to-
gether and ranked third and fourth, respectively. The HD/HS
had the lowest reliability because of the hydraulic pump and
drive motor In the drive train,

2. Vulnerability - The vulnerability of the total system was

determined by an assessment of both the starter system and
C&A drive system components. Since the protection afforded
by the airframe Is not known, the total volume of the alrframe-
mounted starter system components was used as the vulnerability
criterion. The starter vulnerability was given a 25% weighting.
The exposed vulnerable area of the engine-mounted C&A drive
components was assessed and used as the vulnerability criterion.
The C&A drive vulnerability was given a 75% weighting.
86
 ^1

5
c
3
3
H
lU

I
t4
cc
oc
<
U
V
■4->
< CO
oc 3
H
Ui
>
E
o CO

I
87
 Pff u
(co 0)
<♦- ■e
iS
t< CO

^
£M
c
0)
/
f Q)
bo
d
•3
a
6

I
u
to
3

fa

88
 u
B

2
a
v.
>
Q
UJ
2
ui
O
z
<
K
C
<
<
ec

>
I 89
•

P3 N ^- ift

If-
pa *

? z
CO

It
3- Ü

H
t 11-
«s

I 0
£&

its
N
ffl «o
<
fa- W
K«
PQ
<
■P S ^
PI
|§|
9 I

>
oil

Is t* t- o
5 a
B a •^na saso
0 o

u
CO 0) Cl]

JJ 2 i 8 1
rt N rf ^ US

90
 The vulnerability assessments are summarized in Table 17.
. ' The TS/GTS system received the highest rating, primarily
because the relatively large GTS afforded protection to the
C&A drive components. The CGB/ATS was rated second and
reflects a smaller engine-mounted starter. The TS/IS and
CGB/IS system ratings reflect no protection from an engine- *
mounted starter and a larger alrframe-mounted starter APU.
The HD/HS system received the lowest vulnerability rating
because of the large exposed vulnerable area of the hydraulic
pump and drive motor.
3. Development Risk - The C&A drive system/starter configurations
for all systems except the HD/HS were essentially state of the
art and were judged to have a similar development risk. The
HD/HS system requires a convertible starter motor/pump and
a C&A drive motor that must be developed to be extremely
reliable. The hybrid system was therefore the only system
downrated for development risk. *
4. Cost - The cost of the systems was based on estimated production
pricing. The development cost of the components was not re-
flected in the cost evaluation, with the assumption that all sys-
tems would require the same relaüve amount of development
effort. The HD/HS system had the lowest cost because of the
lack of an APU or GTS In the start system.

The cost of the other candidate systems was estimated to be

essentially the same. It should be noted, however, that If the
APU used in systems Nos. 2, 3, and 4 was used for other air-
frame services in addition to starting, then the cost could be
shared and a higher rating assessed.

5. Weight and Volume - The weight and volume of the total starter/
C&A drive system was assessed on the basis of preliminary
component designs. The weight rating was given 50% of the i
total, and the volume rating was given the remaining 50%. The '
weight and volume tabulations are shown in Table 18. The
tabulation shows that the CGB/ATS system has the lowest overall
weight and volume and received the highest rating. The TS/GTS
system was second, and the rating reflected a volume penalty
for the two GTS systems. The HD/HS system was rated third and
had the highest weight but the second lowest volume. The CGB/IS
and TS/IS systems were rated fourth and fifth, respectively, pri-
marily because of the large APU required for starting,
i
6. Performance - The performance of the systems was evaluated
for both starting and during normal operation. The starter
efficiency was evaluated In terms of the required starter fuel
flow and was given a 10% weighting In the evaluation. The C&A
drive performance was weighted at 90% and evaluated on the
basis of drive efficiency. A tabulation of the performance
factors Is shown In Table 19. The mechanical drives were
all rated at the same relative efficiency. The HD/HS efficiency
was lower and reflected the product of efficiencies of the tower
shaft drive, hydraulic pump, and hydraulic motor.

91
 Ia
B
1
S
o
o
o

§
K

« 0)
o
1*
e
in o in

'S «

a a
SI in o in o
o f o
3O V
PH
in o m
o
—£
is
0 0
H^

>tt *-•• 0
in
in
•t
0
0
in
S
in
0
n hct^ n (N in CM tt
tu »g 'S 58
— in
■a
K *

0/
=5^
| g
■- .S 05 o w
IS ■t O in
JS ed
OJ cd
K
|«
1 n
u 8 «4
M .S
so < ^ s«
1« ^H
o>
0
t-
CO
« (M
in S
0 • • • •
> MMt^ > CM

« V■! ^i% v •
1> > "a
||g in
0» g 8 8 in
CO (N m CM
0)
> M 0
• 0
• 0
• 0
«
3 11
n — 00
"g B 1-1
pq S«
J3
Ml o ao 05
Ü*

si fä § 8
«>
GO »> O 1
ll o in CO

^ a IS
0 -c ill
s fel s
«-. 2 S cc 09
a ■s.
i
«1ß f3 C5 o o 00
u u CM CM
<
u 0

12
23
CQ
Ü Q
B 12 o

92
 ¥
j
c
s
K

M
•i
C
aa e^ to
00
K

o
Ä
•*->
b£
w »4
J3
s 05
00 to
Ol 00
rS V
is
4-»
X
s0 'SM s
—1
IN

s
O
g 05
H £

§ 4)
>
i«
■*-»

f„
be 6^
o
s 9)
•
aS
•
9$
•
<o
•
HJ 0 a> o o o o
K^
a>
> M
^I •u
C3
.s-«-) S S g 8
< CS
PS K
,
§ >i

5)
« >
■-4 g
s SÖ eS oo
o>
00
o»
QO
05
00
Oi
(N
r>
w
OS
f-H >V *-»
"fl
*-* J3bo,^ s § o § s
"3 « o
n
OJ
> M o
e o o
« ■ 3
T* cd
* 1 S

II
<u
S II in
M1

ffl PQ \
ü Ü Q
12 ß O O S

93

, 1
 7. Maintainability - The systems were evaluated on a comparative
basis for maintainability. The rosults of the evaluation are
summarized in Table 20 and show the CGB/ATS system
as the easiest to maintain. Comments on the five systems th.t
guided this rating are summarized In Table 21.

8. Installation Flexibility - Installation flexibility of the systems

was evaluated, including both system volume and Installation
complexity. The volume was weighted at 50%, and the in-
stallation complexity was weighted the remaining 50%. The
results are summarized in Table 22 and show the CGB/ATS
with the highest rating. The GTS/TS system was downgraded
because of the potential interference with the airframe gear-
box, and the TS/IS and CGB/IS systems were downgraded be-
cause of the large APU required for start.
TABLE 20. MAINTAINABILITY (RATING = 7%)

C&A Drive/Starter Relative Rating,

Configuration Rating % Ranking

1. TS/GTS 0.75 5.25 2

2. IS/IS 0.69 4.83 4

3. CGB/ATS 1.00 7.00 1

4. CGB/IS 0.72 5.94 3

5. HD/HS 0.53 3.70 5

TABLE 21. MAINTAINABILITY COMMENTS

1. TS/GTS

a. Two GTS units are required.

b. Tower shaft is a more complicated design, with more bearings
and gears.

c. There is less flexibility In mounting than remote APU.

d. There are fewer components than APU-driven ATS.

e. Fuel filters and GTS controls add maintenance.

f. GTS mounting has good accessibility.

94
 TABLE 21. MAINTAINABILITY COMMENTS (Contiaued)

2. TSAS
a. Requirement for ATS is eliminated.

b. Remote APU results in good mounting flexibility.

c. Air lines, nozzles, and filters add maintenance problems.

d. APU and APU starter are very large and heavy for a pneu-
matic system.

e. Pneumatic supply lines are not a Are hazard and subject to

leakage monitoring.

f. Air nozzles may have an accessibility problem.

3. CGB/ATS

a. Pneumatic supply lines are not a fire hazard and subject

to leakage monitoring.
b. Remote APU results in mounting flexibility.

c. Configuration Is smallest and lightest considered,

d. Separate APU and ATS add an additional component.

e. Integral gearbox is less complicated than tower shaft configura-

tion, but must be disassembled to gain access to oil pump.

f. Pneumatic system has fewer contamination problems than

hydraulic system.

4. CGB/IS
a. Same considerations as TS/IS, except for gearbox.

b. Integral gearbox is less complicated than tower shaft configura-

tion, but must be disassembled to gain access to oil pump.

95
 TABLE 21. MAINTAINABILITY COMMENTS (Continued)

5. HD/HS

a. Hydraulic system is same as initial starter; therefore, it is

relatively compact.

b. Accessories are driven hydraulically, can be remotely

mounted, and are regulated independently of engine speed,

c. Hydraulic pumps are run full time.

d. Hydraulic systems are subject to frequent leakage problems,

and they are also a fire hazard.

e. Hydraulic systems are sensitive to contamination

f. System is the heaviest of all considered.

B. FRONT-DRIVE STUDY CONCLUSIONS

The analysis conducted for a front drive engine configuration with a 15-hp PTO
indicates that the controls and accessory drive are best accomplished by a mechanical
drive. Either a cluster gearbox or a tower shaft arrangement can be considered.

Starter system studies have indicated that the best overall engine/airframe starter
choice is the gas turbine starter or the air turbine starter. In installations where
the development cost of a gas turbine starter or an APU for an air turbine starter
is prohibitive, then a battery/electric or accumulator/hydraulic motor start system
can be considered.

Certain observations from the initial design studies were considered in the recom-
mendations for later program phases and are summarized below.

• The requirement for provision of a PTO is a major factor in

configuring the gearbox. This relatively large, low-speed
drive contributes significantly to gearbox size,

• Control and accessory packaging is compromised by a front-

drive power turbine. Integral or direct-drive components
are not possible.

• The basic engine design is not optimized for a front-drive

power turbine, as compared to a rear drive, because of the
impact of the power turbine shaft on the gas generator bore
diameter.

96
 b£
—*C
c w in F-t T N
rt
K
,
| t- i- o X K
• • • • •
»
C3 t^
'', 07 t: C7 T

a
.t£bl
■M
0»
00
01
iH e
o
c-.
no
BS
co ■- i^ (C o• <c l^
T H • • •
,8 >
K l>
o o i-i O e
to
4-1 w in in in m
— x: i- K 1- « m
c- in eb in i^
o '3 • • • • •
o o o o o

c
«> -^.
o in o in in ej

•£3
o
iff o
n
•
o
T
•
-t-
-r
•
f-4
f
•
o
i^
•
o o O O o o
c as ts
a U
a
o
01
73 > be
w C »H o C5 M Ö
rt 3 .5 •
x>
•
X
•
X
• •
c "71 rt o o O o ^H
o K
ao «
o •a
«0 o in ,; o
• • • rt• •
© o -i" o
h-l
n •t <* -r in

<
H
s - m in o o in
-t 00 o X m
xf i-H w rt C^l
• • • • •
o o o o O
M
w
1-1 o
PQ o
-' bo
< E S c OJ I~ o tr fH

■31
00 o CO in
"o • • • • •
o o O o
> OS K

O
* l
SC^J»
c
in
n
to l^ X e
o 00 X N
S W (M rt 1M en
o ■"' i-H « rt n CVl
>

[2
E S < 'Ü cq
v. s
Systei

gurati«

Ü
tarter

a
ü Q
u
%"%\
u
<s • •
n
• •
m
•

97
 SECTION V

CONCEPTUAL DESIGN - REAR-DRIVE ENGINE

The Jtudy was continued to define the optimum C&A system for a rear-drive en-
gi le. The requirements of the baseline engine and C&A components were estab-
lished. Engine definition studies were conducted and led to definition of the
be seline engine configuration. The candidate C&A drives and starter techniques
wire defined, and a matrix of possible C&A drive/starter systems was established,
''he candidate systems were analyzed, and two systems were recommended for
additional study.

A. ENGINE DEFINITION STUDIES

Prior to establishing the baseline rear-drive engine configuration, engine defini-

tion studies were conducted in several areas to evaluate feasibility of selected
candidate configurations.

1. Air Bearing Application Study

The front-drive engine studies showed that the engine oil system represented a
major portion of the engine's vulnerable area. To evaluate complete elimination
of the oil system, further analyses were conducted to determine if an all-air-
bearing engine was within the 1977 development time frame.

A preliminary engine layout, shown in Figure 39, was established, employing all
air bearings. From this drawing, shaft critical speeds and air bearing loads
were evaluated. The gas generator configuration has a rotor bending mode at
47, 800 rpm, which is less than the expected idle speed. For the engine to pass
through this mode without intolerable bearing loads and/or deflections, multi-
plane balancing (which is not currently state of the art) would be required. The
power turbine was not analyzed, but its critical speed was estimated to be
acceptable.

This analysis showed that the thrust bearing size and location (more distance
between bearings) created critical speed problems. Improvements in the gas
generator critical speed could be obtained by reducing the diameter of the thrust
bearing piston and the distance between bearings.

Bearing thrust and radial loads were established as shown below:

1. Engine Unbalanced Loads

Gas Generator Thrust = 370 lb

Gas Generator Radial = 77 lb
Power Turbine Thrust = 1135 lb

2. Engine Maneuver Loads

Gas Generator Thrust = 28 lb x 10g = 280 lb

Gas Generator Radial = 14 lb x 10g - 140 1b
Power Turbine Thrust = 15 lb x 10g = 150 lb

98
 c
o
■>->

II
o
Ü
a>
c
•a
w

<

a*
n

99
Based on the loads established, consideration was given to the feasibility of an
all-air-bearing engine which would be consistent with a Uff? development time In
frame. This was accomplished by reviewing experimental data and through
conversations with Mechanical Technology, Incorporated (MTI). The literature
survey Indicated that the necessary analytical tools needed to deteruvine opera-
ting characteristics of air bearings exist today, but the test data necessary to
substantiate these theoretical predictions for air thrust bearings do not. Based
on discussions with MTI, the bearing load/area cannot exceed 18 to 20 lb/ln.2 f0r
thrust bearings and 40 to 50 lb/ln.2 for radial bearings. From thrust and radial
loads calculated, radial bearings for the gas generator offer no problem at
12.3 lb/In.2. The power turbine radial load/area was not calculated, but It Is ex-
pected to be In the same order of magnitude. Thrust bearing load/area was
much larger than allowable, with the gas generator at 32.5 lb/in.2 and the power
turbine at 64 lb/ln,2.

Based on present technology, thermal and dynamic Instabilities are problems

associated with the thrust balance pistons of the size used In the conceptual lay-
out. Increasing the piston diameter, as would be required to lower the load per
unit area, would Increase Instability problems. The effect or severity of bearing
Instability could be evaluated only by appropriate testing.

The adverse Impact on engine critical speed of a larger thrust balance piston also
Is not desirable. Thrust balancing techniques are not expected to change by 1977.
The load area ratio for the gas generator could be reduced by baselining an engine
using a shrouded centrifugal compressor in combination with a radial Inflow tur-
bine. The thrust balance piston on the power turbine could be eliminated by
allowing the alrframe gearbox to support the power turbine thrust load. A con-
ceptual airframe-gearbox-supported power turbine Is shown in Figure 40.
In summary, for the selected baseline engine configuration, radial air bearings
appear to be within the 1977 development time frame, while thrust bearings do
not. The power turbine load distribution requirements greatly exceed the demon-
strated thrust bearing capabilities. The gas generator thrust load distribution
requirements are much closer to the demonstrated levels, but are sufficiently in
excess of these levels to also be considered beyond the 1977 development time
frame.

An engine without an oil system is still a most desirable configuration and should
be further pursued. For this application, complete elimination of the oil system
would require the following efforts: (1) further experimental development of air
thrust bearings for the gas generator to Increase their load-carrying capability
by a factor of 2 to 3, and (2) use of an engine/alrframe Interface, where the power
turbine is supported by the alrframe gearbox.

2. Bearing Configuration/Arrangement Study

The bearing configuration and arrangement for the baseline engine were established
from an optimization trade-off study. The study Indicated that at least one oll-
lubrlcated bearing would be required on each rotor to support the unbalanced
thrust loads.

100
 8

I
OH
T3

I
<
o

s,
fa

mi
For the gas generator rotor, the forward bearing location was chosen for the thrust
bearing. The forward location allowed close control of the centrifugal compressor
clearances, which would not be possible with a thrust bearing location in the
rear, since dimensional tolerances and thermal growth differences of the rotor
assembly, relative to the compressor case, would require excessive compressor
clearances. A radial air bearing was selected for the rear support of the gas
generator, based on the adequate load-carrying capability of air bearings and
the benefit of reducing the overall oil system volume and weight.

The heat generation summary for the selected configuration is shown in Table 23.

TABLE 23. BASELINE ENGINE HEAT GENERATION SUMMARY

Heat Generation ~ Btu/min
Compartment Location SLS Ground Idle

Front Compartment Ambient -47.8 -41.1

Middle Compartment Ambient Air Cooled Air Cooled

Rear Compartment Ambient + 17.7 +3.1

Compartment Seal Friction All Self-Acting All Self-Acting

ä; 0 .^0

Seal Leakage Negligible Negligible

Gas Generator Thrust Bearing 66 42.5

(Tandem)

Power Turbine Thrust Bearing 24.1 24.1

Gas Generator Journal Air Cooled Air Cooled

Power Turbine Journal Air Cooled Air Cooled

Gearbox ^58.0 <58.0

Total +118 <86.6

The bearing configuration for the power turbine was selected by comparison and
rating of several candidate configurations. Study layouts of the configurations
were made and relative comparison of the candidates accomplished. The candi-
date comparisons are outlined in Table 24.

The No. 4 candidate (rear overhung turuine with a forward air bearing, and a rear
oil-lubricated bearing) was selected based on the shortest coupling length and
superior vulnerability resistance. This configuration also had equal or superior
ratings in all other areas when compared to the other candidates.

102 kL
 <0
M
"2 CO
0)
£3
<U
Ö
•a J «
9< a >
S< a ea e
s M CO
to
co "03
en
3 3 Ü
Meg W ja
u co » V « 01 H 9
Xi | 5 «.2 n U 2
o o
<v 'S u c CO V
u
^3 ».a E w 3
m to
S a
v u a>
dec
tu < O M PQ ooo o
at
bfi
T3
a)
.2 S« a «
ü 'SCO 3 3 u
s n v ca $
CO
0)
W ^ a)
m X) 5 H.2 •P4 .PH — «.2
0)
9)% «'S <oo o> 'S
■*-►

U
c
S)
< H
OJ
o g CO
II la
g CO o
3 I co a»

J
CO
4)
CO
CS
n
v v a>
e ä e
OOO o
Is
0)

-a

o
0)
a
IM
It!
5 s o
en tn
E he
1 J2 J co v
o M
I 3 0)
0)
o
«.2 «.2
«'S to % (0 "o
sag
<;oo «.2
u u E 6
S 6
S CO CO to CO O 0) « ||
o S a ^ c c
o < S 0) rt v
HOO O la
I M
■a T3
U
4-»

W
0) •a
E
re
a
o
X!
I 5-
TO

J3
3 5«3
<u 01
c
0)
c 3
3 g o
Meg M
I- -H ^
■ M •»< "M 3

01
^ Cfl
OJ 0)
<oo 0/
o o (0 (0 CO CO V 0) 0) CO
a > » e e c
< 0
tu O O H H s s ea
PQ OOO o
3
0)
a 2
H H ea
| 6
^ be
c OJ

3 C3
£
E
w to be i^
4J <
80) u«to 5* 'S
■2 e
ü

01
3
•f* *->
0)
PQ •a0> 3
"'S
-4^
CO
>>
bpS
§2
13 2iS
0) CO
X> 3 h 09 a CO " co -a *a C> 3 •«
E 0
3 2 s b —. 3
O
0
>
i CO
3 0)
H fe K Ü o O^ ^c^ Ü

103
3. Integral Start Turbine

The Integral start technique, previously selected and described In Section IV, used
directional jets that Impinged on buckets machined on the gas generator turbine.
This approach, while simple In concept, required a relatively large APU (1.26 lb/
sec at a pressure ratio of 2. 36). The C&A systems using this Integral start tech-
nique were, therefore, penalized due to the weight, volume, and vulnerability con-
siderations of the APU.

In an attempt to make the integral start system a more viable candidate, con-
sideration was given to means of improving system efficiency. A design evalua-
tion was made of a separate, dedicated starter turbine Integral with the gas gene-
rator rotor. The configuration evaluated Is shown In Figure 41. This system
used a 5. 02-ln. diameter turbine and a 30% admission nozzle, and required an
A PU flowrate of 0. 37 lb/sec at a pressure ratio of 4. 0.

The disadvantages of this design prevented Its application. The most significant
was the Impact on engine critical speed. The starter turbine mass and Increased
engine rotor length reduced the critical speed margin below the desired 30%,
considered minimum for the engine design. The second most significant impact
was on overall engine performance. A 10-hp parasite drag was predicted for
the starter turbine at 100% engine operating speed, and 1% of engine airflow leak-
age was predicted due to additional dynamic seal requirements. Pressure losses
In the turbine exhaust case because of the large struts required for the start tur-
bine air Inlet and exhaust passages were anticipated, but were not evaluated. The
turbine added 2. 1 In. to the engine length and 8. 1 lb to the overall weight. There-
fore, this Integral starter turbine was not recommended, and the Impingement
starter system was retained as the prime Integral starter candidate.

B. BASELINE ENGINE DEFINITION

The engine requirements for the rear drive configuration were identical to those
described In Section IT, except that this engine (1) has a single-stage rear drive
power turbine with a design speed of 36,000 rpm and (2) does not have a 15-hp,
20,000-rpm PTO.
The axial power turbine was sized to operate at 36,000 rpm to obtain the desired
operating efficiency with a single stage. Since the rear-drive engine facilitates
use of a short power turbine shaft, critical speed problems are not encountered
at this higher rotor speed, as compared to the long shaft required for the front-
drive engine configuration. A lower power turbine speed could be provided at
the expense of a second turbine stage.

104
 e
3
H
t

105
The baseline engine bearing and seal configuration is described as follows:

1. Front Bearing Compartment

Oil-Lubricated, Antifriction Bearings

Hydrodynamic Lift-off Seals
Elastomeric Damper

2. Middle Bearing Compartment

Tilting Pad Radial Air Bearings

3. Rear Bearing Compartment

Oil-Lubricated, Antifriction Bearings

Hydrodynamic Lift-off Seals

4. Heat Exchanger

Air/Oil Heat Exchanger - Integral With Front Bearing Compartment

Oil Temperature - 200 to 250oF
5. Backup Oil System

Oil Mist - 6-min Capability

A cross-section? I view of the baseline engine configuration is shown in Figure 42.

C. CANDIDA TE C&A DRIVE/STARTER DESCRIPTIONS

During this program phase, the method of driving the C&A components and the
method of starting the baseline engine were considered. Systems that were
evaluated as not technically applicable or beyond the state of the art for a 1977
development time frame during the study of a front-drive engine were not con-
sidered if the design requirements for application to a rear-drive engine were
essentially the same. The C&A drive systems and the starter systems initially
considered in the study are outlined below.

1. Candidate C&A Drives

a. Mechanical
(1) Tower Shaft Drive - Tower shaft drive through a
gearbox with multiple gearing for the various re-
quired accessory drives

(2) Cluster Gearbox - Cluster gearbox mounted about

centerline of engine with multiple gearing for the
various required accessory drives
(3) Single-Speed Drive - Tower shaft or cluster gear
drives a single-speed drive shaft. All required
accessory drives would run in a tandem arrange-
ment at the same speed from this shaft.
106
,
 c
o
E

o
Ü
Oi
c

w
i)
c
0)
to

xf

£
S
fa

107

i i
b. Pneumatic

(1) Cold Gas Bleed - Engine airflow gas bleed upstream

of the burner is used for providing power for the
required accessory drives.

(2) Bleed and Burn - Engine airflow gas bleed upstream

of the main burner is fed through a separate com-
bustor, mixed with a fuel supply, and burned. The
combustion products are then used to provide power
for the required accessory drives.

(3) Interturbine Bleed - Engine airflow gas bleed down-

stream of the gas generator turbine is bled and used
to provide power for the required accessory drives.

(4) Mixed Bleed - Engine airflow gas bleed is bled from

the engine, both upstream of the burner and down-
stream of the gas generator turbine. The hot and
cold gases are mixed to provide power for the re-
quired accessory drives.

c. Hybrid

(1) Any C&A Driv^/Integral Generator - All accessories

are driven by mechanical or pneumatic power with an
electrical generator integrated with the gas generator
rotor to supply electrical power for the control and
ignition system.

(2) Mechanical/Pneumatic Interface - A tower shaft is

used to drive a separate compressor. The com-
pressor is used to power the required accessory
drives.
(3) Mechanical/Hydraulic Interface - A tower shaft is
used to drive a separate hydraulic pump. The hy-
draulic pump is used to power the required acces-
sory drives.

(4) Mechanical Drive/Electrical Fuel Pump Interface -

An electric generator, integrated with the gas gen-
erator rotor, will provide electric power for driving
a variable-speed fuel pump. All other accessories
are driven mechanically.

(5) Mechanical Drive/Electrical or Pneumatic Oil Pump

Interface - An electric generator, integrated with
the gas generator rotor, or a pneumatic bleed from
the engine compressor will provide power for a
variable-speed oil pump. All other accessories
are driven mechanically.

108
 2. Candidate Starter Systems

a. Mechanical
(1) Mechanical/Electric - An electrically powered
starter driving through a gearbox connected to the
gas generator rotor. Starting power is provided
by APU/generator or battery system.

(2) Mechanical/Hydraulic - A hydraulic-powered

starter driving through a gearbox connected to the
gas generator rotor. Starting power is provided
by APU/hydraulic pump or accumulator blowdown.

(3) Mechanical/Pneumatic - A pneumatic-powered

starter driving through a gearbox connected to the
gas generator rotor. Starting power is provided
by APU bleed.

(4) Self-Contained Starter - A self-contained gas tur-

bine, piston, or Wankel engine mechanically con-
nected to the gas generator rotor.

b. Integral

(1) Integral Pneumatic - An external APU supplying

bleed air, which is used in either of the three fol-
lowing methods: (1) cold gas, (2) heat addition
through combustion in the gas generator burner
with closed engine inlet, and (3) hot gas using heat
addition through combustion in an external burner.

1. Selection of Six Candidate Systems

The C&A drive systems considered for this phase were divided into 12 schemes.
The starter systems were broken into two basic divisions: mechanical drive
input and integral starters. A matrix of the candidate systems is shown in
Table 25. The matrix consists of the 12 C&A drive schemes, combined with the
2 basic starter drive systems, including 7 types of starter drive schemes. The
total combinations of C&A drives and starters created 84 possible systems for
study.

Each candidate system was reviewed in light of the trade-off studies and analyses
conducted during the analysis of a front-drive engine. If the analysis was
directly applicable, the same conclusions were used. If the requirements had
changed for the rear-drive engine, the previous results were reviewed in light
of the new requirements, and additional analyses were conducted where required.

109
 dumd no
jliuuinoud ao (BJUIOOI^ M a (A Q a in a a 73 a 73 a

/OAI-UI lu.iiuuqjoiv
JJJ. AJ, A AA A AA AA

duina
lanj lujuj.wix in Q M M Q u) a »a a 73 a 73 a

OAian |EJ!ueqj3|M J>J> n in in AA AA A AA AA

n
OOCJ u
a 0
Ü
,i!|ni;.i|)A||
(0 M Q IA Q <Aa »0 73 a 73 a
i -JOIU|
/|Cj|ucqjo|v n Js A IAA n A « A A AA AA i1 V'
0
o
03 0
n
aoej
w -.Dili] .iiiL'iunou,!
IA »} Q M M Q a in a a 73 a 73 a
H , ]L'.)iui:q.).ii\i
co in in cUJ> A CO A A AA AA
M 0
^ 01
05 .KniMOU.W) |UJ^3)U] tn in en 73 73 73 3

w /»Aua VJO ^"v A A A A A A

>
N-4
1
a
Q f paoia w a M Q a 73 a a 73 a 73 a
5j

b
> .
^H
-
> psxiw ih w m M M 4A A AA AA t 5
« i IB
— b paaia tn Q (ß Q a 73 a a 73 a 73 a c
w u
o
B ouiqjnjjoiuj AA in M A ^A A AA AA •£
C o
w 4) tting wQ M a a 73 a a 7} a 73 a
(9 Ja
i i
b c Ae). i
A AA AA
pue paaiQ
H i
U 0, in N CM ^C CM
<
<
Q
paaia coQ 73 a a cn a a 73 a 73 a
sBo Pioo in eg Ac!. i
CM
i i
A AA AA
<
51
2S
»
•
ajnpoiv M tn 73 73 73 73
A A 1 Ml
A A 2-a «5
w paadS-aiSuis _ L, 9) M
0 C 0) -c si

i c
re
b
o
xoqjeao
Jajsnio
«9
1
in
«3
A
73
4-
73
A
73
1
in

§
u
S
(0 «3 73 73 73 !; « n n-S 4>
««MS J3M0X A A ~ 1 i—t
A in S t T!
c
•" K D
"£ 5" >
-1 it •« S S
< ■fe.t S^ s £
Q 4-*
■o C
6 it
Q
s B ü ^ a; S 0
a
&
<
U 1 i
c

|l|*lfli
a
pa
t n
a &
o
•3 a
to
<N
w
la
n
3 <u
B 1 « i • I«a s 1
1
M

1
3
hJ
PQ
u
v.
1 1 c
o
U
i
c
H

<
1
o
c
a
1
U
o
<3
n
o
CO
o
3
a

1 X 1 1 IT.

8
s I 1 Si
S 1 '
73a — Ncotincc

IBojuBqjaw ! IBaSajui

SJ31JB1S

110
1 i
Six candidate systems were selected for evaluation and are outlined below:

1. Tower Shaft Drive/Air Turbine Starter

2. Tower Shaft Drive/Impingement Starter
3. Cluster Gearbox Drive/Air Turbine Starter
4. Cluster Gearbox Drive/Impingement Starter
5. Single-Speed Drive/Impingement Starter
ß. Single-Speed Drive/Impingement Starter/Integral Alternator

An outline/discussion of the rationale and analyses required to obtain the selected

systems follows:

1. Candidate C&A Drives

a. Mechanical

(1) Tower Shaft - Considered a candidate

(2) Cluster Gearbox - Considered a candidate

(3) Single-Speed Drive - Considered a candidate with

the module operating at 15,000 rpm.
b. Pneumatic

The pneumatic drive system selected during the

front-drive engine study was the cold gas bleed.
This selection was also considered applicable to
the rear-drive engine.
During the front-drive analysis, the best pneumatic
drive was compared with a mechanical drive, and
the mechanical drive was selected, based on ad-
vantages in all areas except maintainability and
installation llexibility with overall ratings of 92.1
(mecha"icäl) vs 69.1 (pneumatic). The reduced
performance penalty for a pneumatic drive for this
application (less bleed air required due to deletion
of PTO) would not significantly alter this evaluation.
Therefore, pneumatic drive systems were dis-
counted in favor of mechanical drives.

c. Hybrid

(1) Any C&A Drive/Integral Generator - ■ -

An integral generator was discussed and discounted
for the front-drive engine, based piimarlly on the
fact that several component drive pads were avail-
able that offered more favorable opportunities for
integration of the generator as compared to the basic
engine. For the rear-drive engine application, the
integral generator was considered a candidate only
for the single-speed module drive operating at
15,000 rpm.
ill
(2) Mechanical/Pneumatic Interface
The front-drive analysis showed that the mechanical/
hydraulic interface was preferred over the
mechanical/pneumatic system, based on superior
or equal ratings in all areas except reliability and
maintainability. The overall ratings were 94. 0
(hydraulic) vs 85.6 (pneumatic). This system
selection was also considered applicable to a rear-
drive engine without PTO, since the drive system
components are sized by the starter requirements.

(3) Mechanical/Hydraulic Interface

Front-drive analyses of a mechanical/hydraulic
drive showed that the mechanical/hydraulic drive
had lower overall ratings in the areas of reliability,
vulnerability, development risk, performance, and
maintainability as compared to any mechanical
drives. This analysis was also considered valid for
a rear-drive engine without PTO since the drive
system components were sized by the starter re-
quirements.

(4) Mechanical Drive/Electrical Fuel Pump Interface

During the front-drive engine study, consideration

was given to electrically driving the fuel pump from
an integral generator. This system was discounted
in favor of a mechüiical drive based on higher
ratings of the mechanical drive in all areas except
performance, maintainability, and installation
flexibility with an overall rating of 91.1 (mechanical)
vs 79.6 (mechanical/electric). The same con-
clusions are applicable to a rear-drive engine with-
out PTO.

(5) Mechanical Drive/Electrical or Pneumatic Oil

Pump Interface

During this phase, the study was expanded to

cover the possibility of an electrically or pneu-
matically driven oil pump. Since the oil system
flow requirements are essentially a direct func-
tion of engine speed, no engine performance im-
provements were anticipated when compared to
a mechanical drive. The advantage of an elec-
trically or pneumatically driven oil pump would
be eliminating or simplifying the gearbox. Since
the baseline engine definition studies defined an
engine with an oil-lubricated, gas generator for-
ward bearing, no significant simplification of the
overall oil system would be realized for this

12
 application. Based on the previous study of an
electrically driven fuel pump, the only advantage
to an electrically or pneumatically driven oil
pump as compared to a mechanical drive would
be in maintainability and installation flexibility
with disadvantages in all other areas. Therefore,
the electrical or pneumatic oil pump drive was
discounted.
2. Candidate Starter System

The starter requirements for a rear-drive engine are the same

as for a front-drive engine. The starter system studies conducted
during the previous analysis were considered applicable. The candi-
date starter systems are outlined below.

a. Mechanical

(1) M-- >>anical/Electric - This system was discounted

o t.ne basis of the front-drive studies, which rated
an APU/pneumatic starter superior to a battery/
electric starter. The overall ratings were 69.2
(electric) vs 85. 0 (pneumatic).

(2) Mechanical/Hydraulic - This system was discounted

on the basis of the previous studies, which rated an
APU/pneumatic starter superior to an accumulator/
hydraulic starter. The overall ratings were 71.8
(hydraulic) vs 85.0 (pneumatic).

(3) Mechanical/Pneumatic - An air turbine starter sys-

tem was considered a candidate.
(4) Self-C ontf.ined - A self-contained gas turbine
starter (GTS) rated overall In the front-drive study
as slightly better than an air turbine starter (96.4
GTS vs 95.6 ATS). The GTS would be used only
where the envelope would allow. If an APU is re-
quired for any other alrframe purpose, then an ATS
would be preferred.

For this application, the starter packaging require-

ments are more critical because of the reduced
gearbox size without PTO. For this study, no ad-
vantage in using a GTS as compared to an ATS was
indicated.

113
 b. lotegral-Pneumatic

(1) Cold Gas Bleed - This system was considered a

candidate.

(2) Engine Ram/Closed IGV's - This system was dis-

counted on the basis of the front-drive studies, which
rated a cold gas bleed system over an engine ram
closed IGV's system based primarily on reduced
development risk with an overall rating of 82. 3
(Cold Gas Bleed) vs 81.6 (Engine Ram/Closed IGV's).

(3) Hot Gas Impingement - This system was also dis-

counted on the basis of the previous studies, which
rated a cold gas bleed system superior or equal to
a hot gas impingement system in all areas except
weight and volume with an overall rating of 82. 3
(Cold Gas Bleed) vs 72.3 (Hot Gas Impingement).
2. Selection of Two Candidate Systems

The six C&A drive/starter systems were evaluated during this task. Preliminary
design layouts of these systems were made, and each system was evaluated based
on the selection criteria previously defined. From these candidates, the two C&A
systems with the highest overall rating were se'ected for the detail design phase
of the program. These systems were:

1. Tower Shaft Drive/Air Turbine Starter

2. Cluster Gearbox Drive/Air Turbine Starter

A discussion relating to the selection of the two candidates is presented in the

following paragraphs.
a. Candidate C&A Drive/Starter Description

(1) Tower Shaft Drive/Air Turbine Starter (TS/ATS)

A tower shaft drive is used to drive a ^&,000-rpm fuel pump/alternator package

and a 15,000-rpm oil pump. An air turbine starter is coupled to the gas genera-
tor shaft through a 1:1 ratio bevel gear train. The air turbine starter consists
of a high-speed air turbine with a cluster gear reduction. An overrunning clutch
decouples the start system when the engine is self-sustaining. The TS/ATS con-
figuration is shown in Figure 43.

(2) Tower Shaft Drive/Impingement Starter (TS/1S)

A tower shaft drive, as described for the TS/ATS configuration, is used for the
C&A components. Starting is accomplished by cold gas impingement on buckets
machined on the back side of the gas generator turbine. The bleed air is pro-
vided by an airframe-mounted APU. The TS/IS configuration is shown in Fig-
ure 44.

114
 '

I
•p"

SI

115
 I-

tJ

c
i

{
i

I
5
s

116
.1
(3) Cluster Gearbox Drive/Air Turbine Starter (CGB/ATS)

The drives for the C&A components are provided by gears clustered around the
gas generator rotor. The C&A components and starter system are identical to
those described for the TS/ATS. The CGB/ATS configuration is shown in Fig-
ure 45.

(4) Cluster Gearbox Drive/Impingement Starter (CGB/IS)

The C&A drive train is the same basic design as the CGB/ATS configuration
described above. Starting is accomplished by the impingement start system
described for TS/IS system. The CGB/IS configuration is shown in Figure 46.
(5) Single-Speed Drive/Impingement Starter (SSD/IS)

A single-speed shaft (15,000 rpm) is used to drive all C&A components. A

15,000-rpm integral fuel pump/alternator and a 15,000-rpm oil pump are co-
axial with the drive shaft. The drive shaft is coupled to the gas generator rotor
through a reduction ring-gear/pinion arrangement. Starting is accomplished by
the impingement start system. The SSD/IS configuration is shown in Figure 47.
(6) Single-Speed Drive/Impingement Starter/Integral Alternator (SSD/IS/IA)

This drive configuration is similar to the SSD/IS, but has the alternator coupled
directly to the gas generator rotor. The 65,000-rpm alternator design is used,
and other components operate at 15,000 rpm. Starting is accomplished by the
impingement start system. The SSD/IS/IA configuration is shown in Figure 48.

b. Analysis

Assessment of the six systems was made considering each of the rating criteria
defined previously. The results are summarized in Table 26 and show the TS/
ATS and the CGB/ATS as the two systems with the highest ratings. Discussion
of the ratings is outlined below.

(1) Reliability

A reliability assessment was made for both the starter system and the C&A drive.
The start system reliabilitv was given a 20% weighting and the C&A drive an 80%
weighting. The results of the analysis are summarized In Table 27. The air tur-
bine starter system was rated as less reliable than the impingement starter due to
the addition of the ATS to the pneumatic scheme. Gearbox reliability was based
on total number of bearings, number of floating splines, and number of gear en-
gagements contained in each gearbox. The SSD was downrated, primarily due to
the number of bearings and gear engagements Involved In the rlng-gear/plnlon
coupling at the gas generator shaft.

117
.

u
u
a

t4

I
en
3
 a
a
09

a
i
a
I

to
3

119
 0»

MM

I
5

i!c
in

8
g.
• i-H

120
HW~

IM
0

9)

00

121
 o ^t

S U.
^, o o
s: tf
i:
M br o a
o o
HS
n
CM

•
ftl
o O n — o
c
« o I- — «• is «c o >
I—t
u 01 0) u>
,_o
X OD 00 X
H

•a
"u
s
0. I > bC
'S .c

1$ . «
H
o e
2 si^
4>
o V •
PQ 41 (S O
n

<;
* < Igi O

cn •r <n
«t«
O N
T
• t-

•-«
O -" ** ^-
i-i
^
t-t
w-4 ^^
Cl
PH

o —
w c
O OB
S JSSS
4. >
ffl §J<.
§■«
IT (0 <c tc c
• W
w
O .2 t--r M" M w
< T«- «M
C4
IM •-X
f
.-(
M

>
1 < <C O (C o o o

Z: eg
§ S S 8 § §
$
-•* c5 — o -J -^ —
^ 05
^ ■W tn
■s« • •
O

h s o! -H 0) o o
<C
— ^-
1 M P4
lure Ra
000 Sta
mponen

9
> r- t- 1- I-
. 00 w OD M N Tl S U
o ■-" (C C M c t£
& 1 ^
•a ÜÄjfc

Ifi
Mi OJ
u
e
r<
o

M
o

N
P4
CM
o>
ni
C4

^
ts
S o o

Ä
5|
5'
crt <
B < £2
>v
in

if in
h a 2| S S s 35 8
aa
< a
t < 22
s 8u

122
A
(2) Vulnerability

The vulnerability of the total system was determined by an assessment of both

the starter system and C&A drive system components. Since the protection
afforded by the airframe is not known, the total volume of the airframe-mounted
starter system components was used as the vulnerability criterion. The starter
vulnerability was given a 257o weighting. The exposed vulnerable area of the
engine-mounted C&A drive components was assessed and used as the vulnerability
criterion. The C&A drive vulnerability was given a 75% weighting. The vulner-
ability assessments are summarized in Table 28. All of the candidate systems
that use the impingement starter were penalized due to the requirement for a
larger APU to provide the increased bleed air as compared to the ATS. The TS/
ATS and the CGB/ATS received the highest ratings because of fuel pump shielding
(in the bottom view) afforded by the ATS.

(3) Development Risk

The C&A drive system/starter configurations for all systems were essentially
state of the art and were judged to have a similar development risk. Shaft seals
will be required for the GS.OOO-rpm alternator and were considered an added de-
velopment risk for those C&A drive configurations using this component, and the
systems were penalized accordingly.

(4) Cost

The cost of the systems was based on estimated production pricing. The develop-
ment cost of the components was not reflected in the cost evaluation, with the
assumption that all systems would require the same relative amount of develop-
ment effort.

Therefore, the cost of all six candidate systems was based on the number of com-
ponents in the particular C&A system and was estimated to be essentially the
same.
(5) Weight and Volume

The weight and volume of the total C&A drive/starter system were assessed on
the basis of preliminary component designs. System weight was SO'/? of the total
rating, and volume was 50% of the total rating. The weight and volume assess-
ments are shown in Table 29. The tabulation shows that the TS/ATS system has
the lowest overall weight and volume. The CGB/ATS system was second, and
the rating reflected a slight volume penalty due to the larger gearbox. All sys-
tems using the impingement starter reflected a high volume due to the larger APU
requirement and were downrated accordingly.

123
 ■H n N

m in
• •
o o

«I

if o in oo
o
U5 in in

o M C5 tc
^
&5 ao Su on tc in
us § in XI m in

O
o on tc to
Alfas in
m <o
N r- t~
V f f
«I
t.
< 2 c
.2"-3
8 00 in s o
to

pa A» « o o
<

In
p o
o n
> w^ .S oo
•H © pH
< «C UJ t-
00 •a
u
W
01 •
oo o oo 00 00
^ m i»
o M e o o
a
v
OS » e
-ac
w
4>

CO «>»
o
o
iH
co
P
o
t-l
<o
rt
co
»<
co
eo I
t
eg
"-' CO
»-( O i-l o

o
M
o
**
o
fi
o
*t
11
15
O MOM CM IN
«0 .H CO eo CO

o o

1 58
41 V
ca «
< "öl "öl
CO (. H K K
ä < 22 en w n n
«^
« <*- ffl n
4
U u
3 to
H
Ü
U
o
u
Q (

124
 c
c

o in n
e

<C « M
Ä 'T T
SÄ
fi *
05 »

_ £
n be 8 T T
O V o o o

o
in
o
05>

S c in n
«33 CO n n

c
>

C-l 0» O) f wt
(Jl n n OD o
n M i- 'T m
o ■"' n •^ CO n CD
>

4) •
in
s• 1-
•
i-

•
is
•
US
c
•S « in
1^ O o •ac
u

4)
> be
I3 ■3.S
o

2w
n en in c
| n
s 00
8 00
<u
6
o

S -2 «-•
S
U5.
ES as
< CQ CO n
o Q
•a § tn Ü
u
C
H H u

125
(6) Performance

The performance of the systems was evaluated for both starting and normal
engine operation. The overall starter performance was evaluated in terms of
the required APU fuel flow and given a 10(/( weighting in the evaluation. The
C&A drive performance was weighted at 90% and evaluated on the basis of drive
efficiency. The gearbox efficiency was s^t at 98% and then adjusted lower,
based on the total number of floating splines, gear engagements, and bearings
for each C&A/Drive candidate. A tabulation of the performance ratings is
shown in Table 30. The performance of all six candidates was rated essentially
equal, primarily due to the common use of a mechanical accessory drive.

(7) Maintainability

The systems were evaluated on a comparative basis for maintainability. The re-
sults of the evaluation are summarized in Table 31 and show the CGB/ATS sys-
tem as the easiest to maintain. Comments on the six systems that guided this
rating are summarized below:

1. TS/ATS

Alignment of bevel gearing is critical.

Oil pump is externally accessible without gearbox dis-

assembly.

Maintainability is considered a function of gearbox

components: bearings, gears, and shafts (total of eleven).

The ATS is an additional gearbox-mounted component.

2. TS/IS

Alignment of bevel gearing is critical.

Removal of fuel pump/alternator assembly i!5 required

prior to oil pump removal.

Maintainability is considered a function of gearbox

components: bearings, gears, and shafts (total of eight).

Maintenance or detailed inspection of engine starter

cannot be performed without separation of engine case.

126
 r /

s? o ■

£
e
01
%
§
o I
a1*
«3
o PS
s "is Ss
9J O 4»

C-5
en 00 00 «
- S2
I, >
o
c
4)

i i
CM o (N O

u
<
I I*

fiil*\ 00 t-
0> 01

o
CO
S § 2 o o o
IK
4) > " o- o o

s -» o
§ IC o
— o

in 5 5
■za\
I'
§1
H
M
en
< i2 a a
<
01 oi 8 Qi O3?
H H

127
I i
 TABLE 31. MAINTAINABILITY (RATING - 7%)

Relative Rating Ratin Ranking

TS/ATS 0.97 6.8 2

TS/IS 0.89 6.2 4

CGB/ATS 1.00 7.0 1

CGB/IS 0.91 6.4 3

SSD/IS 0.89 6.2 5

SSD/IS/IA 0.80 5.6 6

CGB/ATS

Gear alignment is less complex than TS arrangement.

Accessibility to oil pump requires partial disassembly

of gearbox.

Maintainability is considered a function of gearbox com-

ponents: bearings, gears, shafts (total of eight).

The ATS is an additional gearbox-mounted component.

CGB/IS

Gear alignment is less critical than TS arrangement.

Accessibility to oil pump requires partial disassembly

of gearbox.

Maintainability is considered a function of gearbox com-

ponents: bearings, gears, shafts (total of eight).

Maintenance or detailed inspection of engine starter

cannot be performed without separation of engine case.

5. SSD/IS
Alignment of components i.s complex since all components
are coaxial and, therefore, dependent.

Accessibility to oil pump requires removal of fuel pump/

alternator assembly.

Maintainability is considered a function of gearbox com-

ponents: bearings, gears, shafts (total of eight).

128
 • Maintenance or detailed inspection of engine starter can-
not be performed without separation of engine case.

• All gears lie in same plane, making access and disassem-

bly less complex.

6. SSD/IS/IA

• Alignment of components is complex since all components

are coaxial and, therefore, dependent.

• Accessibility to oil pump requires removal of fuel pump

assembly.

• Accessibility to alternator requires complete disassem-

bly of gearbox.
• Maintainability is considered a function of gearbox com-
ponents: bearings, gears, shafts (total of eight).

• Maintenance or detailed inspection of engine starter

cannot be performed without separation of engine case.

• All gears lie in same plane, making access and dis-

assembly less complex.
(8) Installation Flexibility

Installation flexibility of the candidates was evaluated, including both the system
volume and installation complexity. The volume was weighted at 50%, and the
installation complexity was weighted the remaining 50%. The results are sum-
marized in Table 32. The systems with the impingement starter were downrated
due to the larger APU requirement. Installation complexity for all candidates
was considered equal except that candidates with the ATS were slightly penalized
because of this added component.

TABLE 32. INSTALLATION FLEXIBILITY (RATING - 6%)

Relative Rating Rating Ranking

TS/ATS 1.00 6.0 1

TS/IS 0.76 4.6 3

CGB/ATS 0.98 5.9 2

CGB/IS 0.75 4.5 4

SSD/IS 0.75 4.5 5

SSD/IS/IA 0.74 4.4 6

129
D. REAR-DRIVE STUDY CONCLUSIONS

The analysis of the starter/C&A system for an engine with a rear-drive power
turbine and no PTO requirement has shown that the C&A drive is best accom-
plished mechanically, using a tower shaft or cluster gearbox configuration. The
gearbox provides a 65,000-rpm fuel pump/alternator drive, a 15,000-rpm oil
pump drive, and an interface for the starter.

The recommended starter is an air turbine unit. For those installations where
an APU is not desirable, the air turbine starter can be replaced with an elec-
trical or hydraulic unit.

The starter/C&A study for a rear-drive engine has confirmed the observations
outlined below, which were originally made as a result of the front-drive study.

Elimination of the PTO greatly simplifies the engine gearbox

design.

C&A packaging is improved with a rear-drive engine.

Several basic advantages for a rear-drive power turbine engine configuration as

compared to a front drive were also identified.

Bearing DN and rotating seal velocities are lower, which

should contribute to a more reliable engine.

The gas generator turbine disk will be lighter due to a smaller

bore.

The power turbine critical speed will be higher and will allow
the use of a single-stage power turbine.
The hub/tip ratio of the gas generator compressor will
be improved.

Maintainability of the power turbine module is improved.

Based on these observations, it was recommended that further detailed design

of a starter/C&A drive system be based on a rear-drive engine configuration.

A rear-drive engine configuration may not fit all installations, but It does repre-
sent an optimized powerplant package. For those airframe Installations requir-
ing a front-drive engine, it was not anticipated that the basic starter/C&A system
components would change. The primary impact on the starter/C&A drive system
would be nonoptimum packaging. The controls and accessory packaging for a
front-drive power turbine engine configuration with an electric starter pad Is
shown In Figure 49.

130
 I

I
5

1
i
«

£
ä
E
131 ^ ..
.Ä. ■— - ------ -—-———
 SECTION VI
SELECTION OF SYSTEM FOR DETAILED DESIGN

A. CANDIDATE SYSTEMS
The conceptual design studies for a rear-drive power turbine described in Sec-
tion V concluded that the C&A drive is best accomplished by a tower shaft gear-
box or cluster gearbox arrangement. Both configurations use a gearbox-mounted
air turbine starter. A description of the TS/ATS and CGB/ATS systems is given
in the following paragraphs.
1. Tower Shaft Drive/Air Turbine Starter

Figure 50 details the TS/ATS configuration and shows the locations of the control
and accessory components and the critical control system sensor interfaces.
The tower shaft is coupled to the gas generator rotor through a 0.462:1 ratio
spiral bevel gear set and is driven at 30,000 rpm. The alternator is incorporated
on the tower shaft assembly. The alternator is a cartridge-type unit that runs
in an oil-mist environment in a double-piloted cylindrical cavity above the engine
centerline.

At the upper end of the tower shaft is a combination spiral bevel and spur gear.
The spiral bevel gear drives a centrifugal fuel pump at 65,000 rpm through a
2.167 gear ratio. The fuel pump drive shaft also contains the air-oil separator
and is mounted in an oval-shaped cover, which pilots on the upper end of the
alternator and bolts to the inlet case. The cover, which also protects the oil
pump gear, is sealed by a rubber O-ring.
A spur gear drives the oil pump at 15,000 rpm through a 0.5:1 gear ratio. The
oil pump is mounted parallel to the alternator in another cylindrical cavity with
cored passages for oil pump supply, scavenge, and discharge. The oil pump
configuration is a high-displacement vane pump using a through-vane concept and
includes a tandem scavenge-pump element.
The air tuibine starter drives the engine through a quill shaft, supported radially
and retained axially to facilitate starter installation. The quill shaft is retained
by a cone-shaped cover when the starter is removed. An overrunning clutch
within the starter disengages the starter turbine after the engine becomes self-
sustained. The starter will be required to provide 5.2 ft-lb output at 16,000-rpm
engine rotor speed, where ignition will occur and will operate up to 43,000-rpm
engine self-sustaining speed.
The centrifugal fuel pump and inlet guide vane (IGV) actuator are integrated with
the fluid controller located on top of the engine above the IGV unison ring. The
electronic computer is located adjacent to the fluid controller. A finned heat
exchanger section protrudes through the inlet duct to provide a heat transfer path
to the inlet air to cool the electronic components.

132
.i
 2

u
€
B
53
s
s
3
H

133
■ ■ ■ -■- -■ ■- - -^^-
..■.^. ._. „_ . _.
**
_
The torque sensor is a shaft twist-type indicator, designed into the power turbine
shaft by means of three gear-toothed wheels adjacent to each other. One gear
is attached to the output shaft and the other two are on the end of a torque tube
that is concentric and splined to the output shaft. The three-gear system is a
modification of the conventional phase displacement torque measuring system,
where the third (position) gear compensates for shaft misalignment not accounted
for by the torque and reference gears.

The power turbine overspeed sensor is a ring of cantilevers, splined to the for-
ward end of the power turbine shaft. Centrifugal deflection of the cantilevers
is used to generate a pneumatic overspeed pressure signal to the fluid controller.
The optical pyrometer Is incorporated on top of the engine case with the aperture
assembly sighting down and tangential to the gas generator blades. Installation
in the engine requires a straight aperture assembly, with the detector assembly
located within the electronics package.

The 65,000-rpm fuel pump (Figure 51) Is contained within the hydromechanical
fluid controller and results In the packaging arrangement shown In Figure 52.
Weights and volumes of the alternator, pump, and fluid controller are listed In
Table 33. The ATS and electronics package, common to both the TS/ATS and
CGB/ATS systems. Is also Included in the table.
2. Cluster Gearbox/Air Turbine Starter

Figure 53 details the CGB/ATS configuration and shows the location of the control
and accessory components. The control system sensor Interfaces are the same
as In the TS/ATS design. The CGB/ATS features a primary shaft driven by the
gas generator rotor at 20,600 rpm through a 0.317:1 spur gear ratio. An internal
gear is used to obtain a large speed reduction while keeping the two shaft center-
lines as close together as possible.

ONIVI VIINC

Figure 51. 65,000-rpm Centrifugal Pump

134
4 t
 ion Muutr-
Mll AOJUIT ■

ONIVi VIINE

MOLDED HARNESS TO
ELECTRONIC CONTROL

Figure 52. Tower Shaft Drive Fluid Controller

and Fuel Pump

TABLE 33. COMPONENT WEIGHTS AND VOLUMES

TS/ATS SYSTEM

Component Weight, lb Volume, in.

Alternator 2.6 9.9

Fuel Pump 1.8 16
Fluid Controller 7.4 74
Electronic Computer 3.8 76
Air Turbine Starter 9 140

The air turbine starter, common to both candidate configurations, meshes with a
spur gear on the forward end of the primary shaft. An overrunning clutch within
tho starrer disengages the starter turbine after the engine becomes self-sustaining.

The oil pump gear, located in the middle of the primary shaft, drives the oil
pump at 15,000 rpm through a 1.728:1 gear ratio. The oil pump, also common
to the TS/ATS design, is mounted in a cylindrical cavity in the inlet case and
is connected to cored passages for oil pur ip supply, scavenge, and discharge.

The gear driving the oil pump is also used as an idler gear to drive the fuel pump
drive shaft at 65,000 rp^i through a 4.34:1 gear ratio. An air-oil separator is
located on the fuel pump drive shaft.

135
 a
Ü
0)

c
S
3
H

2u
S
a
CO
3

n
tsi

136
The alternator in the CGB system is incorporated into the fuel pump package as
shown in Figure 54 and will run dry at a maximum speed of 65,000 rpm. The
hydromechanical fuel-metering package and the electronic computer are mounted
the same as in the tower shaft configuration.

CENTKIFUGAL
INOOCtH

«LTEKNATOR

Figure 54. 65, 000-rpm Centrifugal Pump/

Alternator

The fluid controller size and weight are detailed in Table 34, and the packaging
arrangement is shown in Figure 55.

B. EVALUATION, RATING, AND SELECTION OF ONE SYSTEM

1. Summary

The tower shaft/air turbine starter (TS/ATS) control and accessory drive system
was recommended for the detailed design phase, based on a detailed comparison
with the cluster gearbox/air turbine starter (CGB/ATS) system. Design layouts
of the two candidate systems were completed and evaluated using the weighted
selection criteria previously established for the program. A summary of the
evaluation is included in Table 35.

The two candidate systems used a common gas generator and power turbine.
The common controls and accessories components included the starter, elec-
tronic computer, sensors, ignition system, compressor bleed valve, fuel pump,
IGV actuator, and hydromechanical metering system. Components that were
not common Included the control and accessory drive (gearbox), starter drive,
alternator, and hydromechanical system packaging. These analyses concentrated
on evaluating the differences in the basic components, and the differences brought
about by the variations in the overall component arrangement.

137
 TABLE 34. COMPONENT WEIGHTS AND VOLUMES
CGB/ATS SYSTEM

Component Weight, lb Volume, in.

Fuel Pump/Alternator 4.8 40

Fluid Controller 10.8 94

MOLDED HARNESS TO
ELECTRONIC CONTROL

Figure 55. Cluster Gearbox Drive Fluid

Controller and Fuel Pump/
Alternator

TABLE 35. OVERALL SYSTEM EVALUATION

Systems Relative
Criteria Rating
Criteria Weight, % TS/ATS CJB/ATS

Reliability 23 23 20.8
Vulnerability 19 19 14.8
Development Risk 14 14 13.5
Cost 12 12 12
Weight and Volume 10 10 7.9
Performance 9 9 7.9
Maintainability 7 7 6.5
Installation Flexibility 6 6 5.7
100 100 89.1

138
The two systems were closely rated in all areas except vulnerability, where the
TS/ATS system showed a distinct advantage. This was primarily due to the
integration of the alternator with the tower shaft drive and the compactness of
the tower shaft gearbox as compared to the cluster gearbox.

The TS/ATS system showed advantages in all other categories except cost, where
the two systems were rated equally, with the resultant overall rating of 100 for the
TS/ATS and 89.1 for the CGB/ATS system. These analyses reflected a clear choice
for the TS/ATS system.

2. Analyses

The completed preliminary design layouts were reviewed and graded using the
selection criteria described herein. The results of these analyses are sum-
marized in the following paragraphs.

a. Reliability

The component failure rate estimates reflected inherent design features affecting
reliability. These included susceptibility to dry starts, shock loading, potential
of misalignment, and the design operating levels relative to the state of the art.
The approach used "base" failure rates, i. e., similar component failure history
from other P&WA turboshaft engines. Each component base rate was adjusted
for differences in design and duty requirement, e.g., bearing DN's, environ-
mental conditions, startup frequency, etc. The adjusted rates were then com-
piled to determine an overall failure rate for each scheme. The analyses
showed a preference for the TS/ATS system as compared to the CGB/ATS system.
The CGB/ATS failure rate was greater due primarily to the larger number of
bearings and seals required for the alternator. A point in favor of the TS/ATS
scheme was that the alternator did not require additional bearings (eliminated
one bushing), did not require shaft seals (eliminated two seals), and eliminated
one spline coupling. Alternator bearing failures are among the more dominant
alternator failure modes in field engines. Operating the alternator in the gear-
box environment as compared to an ambient environment was not seen as a penalty
to the TS/ATS system.

The relative reliability evaluation of the two candidate systems is summarized

in Table 36.

Based on a total weight of 23 for reliability, a rating of 23 was given to the

TS/ATS and a rating of 20. 8 was given to the CGB/ATS system.

b. Vulnerability

The vulnerable areas were tabulated for both the start and in-flight modes. The
start mode considered the engine-mounted starter and fuel system components,
while the in-flight mode considered only the fuel system components. The results
of the vulnerability analyses are shown in Table 37.

139
The TS/ATS had a vulnerability advantage due to the greater degree of protection
provided for the alternator and the compactness of the tower shaft gearbox. Both
schemes had good frontal protection. The greater tooth contact area of the spiral
bevel gears enhanced survivability in the event of tooth damage to the TS/ATS
system.

Vulnerable area estimates were made for four views of each scheme (front,
bottom, left, and right). The vulnerable area was taken as the product of a
component's projected area and the probability of a kill (given a hit). Estimates
of kill probability were governed by:

1. Degree of protection afforded by neighboring components

2. Degree of protection provided by the engine structure

3. Survivability of the damaged component as a function of hit

location, size, and velocity of missile.

Components analyzed included the alternator gears (web and teeth), bearings
(races and rolling elements), shafts, and splines. The oil pump was included
in the final numbers; however, it is felt that a direct hit would not cause an
immediate kill.

The projectile used was a 30-caliber (0. 303 in. diameter) bullet impacting with
a 2500-ft/sec strike velocity at 0-deg obliquity.

TABLE 36. RELIABILITY SUMMARY

Ab = Basic Failure Rate/106 Flight Hours

Shock Shock Loading Factor
Dry Start Dry Start Factor
Xt = Total Failure Rate/106 Flight Hours
1. Tower Shaft/Air Turbine Starter

Dry
Component x
b Shock Start H Remarks

Gears
Spur 1 2 1 2 1 spur/2 gear meshes
Spiral Bevel 4 1.5 1 6 2 spiral bevel meshes
Splines 4 1.5 1 6 4 spline couplings
Shafts 6 2 112 1 at 30,000 rpm
1 at 15,000 rpm
4 at 65,000 rpm
Bearings
Rolling Element 5 1 2.5 12.5 3 at 65,000 rpm
2 at 30,000 rpm
Bushings 5 1 15 3 at 15,000 rpm
2 at 65,000 rpm
(DN's within state of the
art)
140
 TABLE 36. RELIABILITY SUMMARY (CONTINUED)

Tower Shaft/Air Turbine Starter

Dry
Component Xb Shock Start XT Remarks

Seals 1 25 3 at 65,000 rpm (oil)

2 at 65,000 rpm (fuel)
Oil Pump Element 1 1 Dual vane element (3
bearing supports)
Alternator 20 20 30,000 rpm (oil mist
environment)
Total 89.5

Cluster Gearbox/Air Turbine Starter

Dry
Component Xb Shock Start XT Remarks

Gears
Spur 1 2 1 8 4 spur/gear meshes
3 spline couplings
Splines 3 1.5 1 4.5
Shafts 6 2 1 12 4 at 65,000 rpm
1 at 20,300 rpm
1 at 15,000 rpm
Bearings
Rolling Element 5 2.5 12.5 2 at 20,300 rpm
3 at 65,000 rpm
Bushings 6 1 6 3 at 15,000 rpm
3 at 65,000 rpm
Seals 7 5 35 1 it 41,000 rpm
2 at 65,000 rpm (oil)
4 at 65, 000 rpm (fuel)
Oil Pump Element 1 I 1 Dual vane element
(3 bearing supports)
Alternator 20 1 20 65,000 rpm (ambient
environment)
Total 99.0

141
 TABLE 37. VULNERABILITY ANALYSES

Absolute
Vulnerable Area, Relative Ratine Total
in? Rating Start Inflight Absolute
Start Inflight Start Inflight 25^ 759i Rating

TS/ATS 20.2 18.7 1 1 0.25 0.75 1.0

CGB/ATS 25.2 24.2 0.8 0.77 0.20 0.58 0.78

Based on a total value of 19 points for vulnerability, the TS/ATS system was
ra,Led at 19, and the CGB/ATS was rated at 14. 8.

c. Development Risk
The development risk of the two candidate systems was rated essentially the
same, with a small advantage evident with the TS/ATS system. The gear trains
and the alternator configurations were primarily considered in this evaluation,
since the other C/A components were common.

The gearbox bearing DN's and gear pitch-line velocities were evaluated for both
schemes and are tabulated in Figure 56 for the TS/ATS and in Figure 57 for the
CGB/ATS. These levels of DN and pitch-line velocity were within the state of
the art and were not considered as high-risk items. The development risk of
using spiral bevel gears (TS/ATS) as compared to spur gears (CGB/ATS) was
considered to be essentially the same.

The development risk of the overall CGB/ATS drive system was considered to be
slightly higher due to the additional bushing and dynamic seals required in the
alternator drive.

The alternator configurations for each system are summarized in Table 38.

The development risk of the ,, 'temator electrical components was rated the same,
based on no large operating temperature differential. The penalty for developing
the additional bearings, seals, and splines for the CGB/ATS alternator is re-
flected in the gearbox development risk rating.

The development risk ratings of the drive system and components are summarized
in Table 39.
Based on a total value of 14 points for development risk, a rating of 14 was given
to the TS/ATS and a rating of 13. 5 was given to the CGB/ATS.

d. Cost

Estimated production pricing for the candidate C&A drive systems and the C&A
components was determined. The gearbox component costs for both drive
schemes were assessed to be so close as to be comparable. The estimated
absolute cost of either drive system was $6,000. The control and accessory
components for each drive system are compared in Table 40.

142
 BEARING ON OR GEAR
1 LOCATION DIAMETER, mm rprn PITCH-LINE VELOCITY

B1 19 65,000 1.235 x 106

1 B2 12 30,000 0.36 x 106

B3 13 30,000 0.39 x 106 |

D4 12 65,000 0.78 x 106

B5 12 65,000 0.78 x 106 |

G1 25.4 65,000
17,017 ft/min
55.1 30,000

G2 36.2 30,000
11,250 ft/min
73.3 15,000

G3 42.7 30,000
13,191 ft/min
19.7 65,000

Figure 56. Tower Shaft Drive System Bearing

DN's and Gear Pitch-Line Velocity

143
 ,
BEARING DN OR GEAR
LOCATION DIAMETER, mm rpm PITCH-LINE VELOCITY |
B1
i 19 65,000 1.235 x 106 j
B2 6
I 16 20,287 0.324 x 10 j
6
I B3 16 20,287 0.324 x 10 i
B4 10 65,000 0.650 x 106 \
6
B5 10 65,000 0.650 x 10 |
26.2 65,000
G1 17,527 ft/min |
83.8 20,287
26.2 20,287
G2 5,470 ft/min
13 40,973
61.2 20,287
G3 12,800 ft/min i
82 15,137
82 15,137
G4 12,800 ft/min |
19.1 65,000

Figure 57. Cluster Gearbox Drive System Bearing

DN and Gear Pitch-Line Velocity

s
 TABLE 38. ALTERNATOR DESIGN REQUIREMENTS

TS/ATS CGB/ATS

Speed, rpm 30,000 65,000

Ambient Temperature, °C 121 80
Ambient Environment Oil Mist Atmospheric
Additional Drive Spline Required - 1
Additional Seals Required - 2
Additional Bushings Required - 1

TABLE 39. DRIVE SYSTEM AND COMPONENT DEVELOPMENT

RISK RATINGS

Relative Relative Ratings Absolute Rating, %

Component Weight, % TS/ATS CGB/ATS TS/ATS CGB/ATS

Oil Pump 20 1.0 1.0 20 20

Fuel Pump 20 1.0 1.0 20 20
Alternator 20 1.0 1.0 20 20
(Electrical
Components)
Gear Train 40 1.0 0.9 40 36
100 100 96

TABLE 40. COMPARISON OF C&A DRIVE SYSTEM COMPONENTS

TS/ATS COB/ATS

Hydromechanical Package/Pump/Altemator $2,550

Hydromechanical Package/Pump $2, 280
Alternator 225
Electronic Computer and Sensors 3 400 3,400
Torque Sensor (Speed Pickups) 45 45
Power Turbine Overspeed Sensor 165 165
Compressor Bleed Valve ^_ 275 275
Total $6, 390 $6,435

A tabulation of the cost ratings is shown in Table 41. Based on a total weight
of 12 points and no significant differences in the total cost, both systems were
given a rating of 12.

145
 TABLE 41. COST

Rating = 12

C&A/Starter Absolute
Configuration Costs, $ Rating

TS/ATS 12,390 1.0

CGB/ATS 12,435 0.996

e. Weight and Volume

The weight and volume of the two candidate systems were assessed on the basis
of preliminary engine system designs and the control and accessory system
vendor analysis of the C&A components. Since the air turbine starter is common
to both candidates, the tower shaft and cluster gearbox drives were evaluated on
the basis of gears, gear shafts, bearings, and supporting members to determine
the weight of each system. The volumes were determined directly from the pre-
liminary design layouts.

The weight of the CGB drive is 3. 6 lb. The weight of the TS drive is 3. 8 lb.
The number of bearings in each drive system is the same. The alternator/oil
pump gear shaft and the cylindrical cavities for the oil pump and alternator
account for a major portion of the TS drive system weight penalty.

The volumes of the CGB and TS gear trains were determined to be 110.5 and
86. 6 In.3, respectively. The major Impact on the CGB volume Is the axial oil
pump cavity and length of the primary gear shaft, which displaces the air turbine
starter axlally forward 1 in.

The C&A components for each candidate system are common, with the exception of
the alternator and hydromechanlcal package. The CGB arrangement, which incor-
porates the fuel pump, alternator, and fuel metering package into one unit, results
In a combined package weight and volume of 15. 6 lb and 134 ln.3, respectively.
The tower shaft arrangement, which has a combined fuel metering package and
fuel pump and a separate alternator, weighs 11. 8 lb and has a total volume of
99.9 ln.3.

The combined weights and volumes for each candidate configuration are shown
in Table 42.

TABLE 42. COMBINED WEIGHTS AND VOLUMES FOR

CANDIDATE CONFIGURATIONS

TS/ATS CGB/ATS
Weight, Volume, Weight, Volume,
lb ln.3 lb ln.3

C&A Drive 3.8 86.6 3.6 110.5

C&A Components 11.8 99.9 15.6 134.0
Total 15.6 186.5 19.2 244.5

The overall weight and volume rating of each candidate C&A drive/starter
system is shown In Table 43. System weight was 50% of the total rating, and
 volume was 50% of the total rating. Based on a total rating of 10 points for
weight and volume, the TS/ATS system was given a rating of 10 and the CGB/ATS
system was given a rating of 7.9.

TABLE 43. WEIGHT AND VOLUME

Weight (50T) Volume (5090

C&A Starter Weight, Relative Absolute Volume, Relative Absolute Total Absolute
Configuration lb Rating Rating in? Rating Rating Rating

TS/ATS 15.6 1.0 0.5 186.5 1.0 0.5 1.0

CGB/ATS 19.2 0.81 0.41 244.5 0.76 0.38 0.79

f. Performance

Each C&A drive/starter cc*ifiguration was evaluated on the basis of engine inlet
blockage due to the gearbox configuration and gearbox drive efficiency to de-
termine the effect on system performance. Inlet blockage was 60% of the total
i-ating, and drive efficiency was 40% of the total rating.

The 'ower shaft configuration was determined to have a 38-deg inlet blockage
(Figure 58), while the cluster gearbox configuration has a 48-deg inlet blockage.

A method of rapidly assessing the sensitivity of a compressor to circumferential

distortion has been developed by P&WA, East Hartford. This technique indicates
that a 20% degradation in distortion tolerance of the engine would result with the
COB arrangement.
The C&A drive efficiency for each configuration was assessed to be equal because
of the similarity In the number of bearings, shafts, and gears and the similarity
in the shaft speeds. A tabulation of the performance ratings Is shown In Table 44.
Based on a total weight of 9.0 for performance, the TSA.TS was given a rating
of 9. 0 and the CGB/ATS was given a rating of 7. 9.

0 - INLET BLOCKAGE

Figure 58. Tower Shaft Contiguratlon With n

38-deg Inlet Blockage
 TABLE 44. PERFORMANCE

Drive Efficiency (40<3?) Inlet Blockage (60^)

Total
Relative Relative Absolute Inlet Relative Absolute Absolute
Configuration Drive q Rating Rating Blockage, deg Rating Rating Rating

TS/ATS 1 1 0.4 38 1.0 0.6 1.0

CGB/ATS 1 1 0.4 48 0,79 0.48 0.88

g. Maintainability

The maintainabilitj of the two candidate systems was evaluated, with the result
that a slight preferei 3e for the TS/ATS system was seen as compared to the
CGB/ATS system. The control and accessory component accessibility and gear-
box complexity were also considered. This preference was primarily due to the
better accessibility of the oil pump and alternator for the TS/ATS system. The
maintainability evaluation is summarized in Table 45.

Based on a total weight of 7 for maintainability, the TS/ATS was given a rating
of 7 and the CGB/ATS was given a rating of 6. 5.
h. Installation Flexibility

The installation flexibility of the two candidate systems was evaluated conside "ing
the applicability of the configurations for alternate engine installations. The
engine/airframe Interfaces and basic component accessibility were also con-
sidered.

The engine/airframe interfaces for the starter, fuel supply, and electrical con-
nections are the same for both systems. Accessibility to the C&A components
is good in both configurations, provided that there is free access to the top of
the engine. The major C&A components (starter, electronic computer, hydro-
mechanical metering system, and fuel pump) are similarly located for both
systems. The oil pump is considered to be less accessible In the CGB/ATS sys-
tem and would require partial gearbox assembly. The evaluation is summarized
In Table 46.
Based on a total weight of 6.0 for installation flexibility, a rating of 6.0 was
given to the TS/ATS and a rating of 5.7 was given to the CGB/ATS.

148
 o

is I-

— d d

73
h
Z
w
O
o M
u
L3 B1
o X A
I«
pS^Co p E 5 ** « S -9
h r-
< H
<
(-
V BOS L
i
bh O S S K .
Ü
< E o
cs
r r.
i- h «5 c ü 5 E " '■
S£.üEo«f 3 ■£ .5 E - ">

>
<r,
a
0
V
Cd
« 0J« £ « 3^»•=«
it •
• i«iT
t* S ^ 2 F
Ci g ff-1
C &

II
1
C '
w E
ri S
D £ « S 5^ = 2 -3
r r. (-aESncrwa M U t. -C E ^ H
h

< n i
- c o
o o
£lo
<
H
|o 8o >- c

P I I ö o
o »
I
Z 5 o

<
e 4
I
S5hr Eo u
■v
«
£ I t £ E
• o a s E
> o
: ju
■>*• 3 ta
pi ^CÜ o

a ill
g o
|E
■> o-
vS
a. K Q> tn 0

s £ a
< 3 n E3
e C O u -^
o o .g g| o So «-5
« aPt-
c ^ ~ 73 n 3 "5 U n « a.

I
^§
tl
=o i 1 i
boxC
s and

fü S« o
togs

i» e o
u
£o Ie 51 E
ty

c
E t Art 19 *
I
E
o
*w S Ü §
s
o
o
3 » r
u. E < C I 3 a
h 51
-r 5 5

149
 TABLE 46. ENGINE/AIRFRAME INTERFACE AND
COMPONENT ACCESSIBILITY

Relative Rating Absolute Rating

Weight TS/ATS CGB/ATS" TS/ATS CGB/ATS

Engine/Airframe Interface 50 1 1 50 50
Accessibility
Component Accessibility 50 1 0.9 50 45
100 95

ISO
 SECTION VII

DETAILED DESIGN OF SELECTED SYSTEM

Approval of the recommended tower shaft/air turbine starter control and acces-
sory drive system was received from the Army, and further optimization of the
C&A systems was initiated. To decrease vulnerability of the electronic com-
puter, the control was moved to the top of the engine, adjacent to the fluid con-
troller. Air cooling of the electronics housing would be provided by compressor
inlet airflow.

In an effoi to reduce engine length, starter configuration studies were accom-

plished to guide the definition of the available starter envelope.

Control and accessory component preliminary requirement specifications and

final envelope drawings were prepared and submitted to the component vendors.
Control system sensor interfaces and component interconnects and interfaces
were defined. Consideration was also given to environmental conditions, cool-
ing, routing, and accessibility.

Meetings were held with control and accessory system vendors to review the
component specification requirements and to discuss the potential interfaces.

Garrett/AiResearch was contacted to provide input for the starter specification.

The need for engine-supplied lubrication of the starter bearings and gears was
established due to the high operating speed of the overrunning clutch. Imple-
mentation of an emergency clutch disengagement feature was also discussed and
included as part of the starter requirements.

The oil pump mounting configuration was reviewed with Sundstrand Corporation
to confirm the approach of using a cartridge-type pump that would pilot in^o a
cylindrical cavity within the accessory gearbox. For installation flexibility, the
spur gear drive will be an engine-supplied part. It was determined that an O-ring
seal should be provided between the pump inlet and discharge cored passages and
that Sundstrand would provide a shear section or equivalent mechanical feature as
part of the pump drive section. Design analysis determined acceptable piloted
cavity clearances on the basis of cavity material thermal growth and required gear
meshes.
The Chandler Evans Corporation began final design of the control and fuel system
components applicable to the tower shaft drive engine configuration. The following
paragraphs detail the final C&A system components and the study of an emergency
lubrication system.
Control and accessory component layouts were coordinated, and the final layout
drawings were received from the component vendors and incorporated in the
engine layout drawing. The C&A system components are described in the fol-
lowing paragraphs.

151
1. Electronic Control and Sensors

a. Electronic Computer

The electronic fuel control technology selected for this study is being developed
by Chandler Evans for the U. S. Army Air Mobility Research and Development
Laboratory, Eustis Directorate. This selection was made because the control
was originally selected for development, based on design trade-off studies that
are consistent with the goals of the subject program. The control is being devel-
oped for small (2- to 5-lb/sec airflow), advanced turboshaft engines for which
formal awelopment would begin in 1975.

The electronic computer is a hard-wired, special-purpose, hybrid computer,

which provides all of the computation, control, scheduling, and logic function
requirements for controlling the engine. It incorporates a 256 eight-bit "read
only" memory for start and acceleration scheduling and provides IGV closed-
loop scheduling and compressor bleed valve control. Both are controlled as a
function of corrected gas generator speed.

The computer module houses the radiation pyrometer silicon chip for the turbine
blade temperature limiter and the variable capacitance compressor discharge
pressure sensor. These sensors are located within the computer housing to take
advantage of the cooled environment. Cooling is required to enhance the relia-
bility of the electronics and to isolate them from the engine temperature environ-
ment. The electric power dissipated in the computer is only about 6w, so
the conventional self-heating problems are obviated; therefore, air or fuel cooling
can be used.

The computer controls the IGV actuator and the position of the main metering
valve through stepper motors. IGV and metering valve positions are fed back
to the computer through resolvers.

The control/accuracy and sensing requirements are listed in Tables 47 and 48.
The fly-by-wire approach has been incorporated into the electronic control de-
sign. Electrical PLA and collective pitch input from the airframe have been
provided. The main fuel shutoff valve is solenoid operated. Provision for a
redundant shutoff solenoid, using airframe power, can be added for independent
pilot control. External adjustments of maximum and idle gas generator turbine
speed are located on the package exterior. The computer can also accept cock-
pit trim of power turbine set speed. The starter cut-out signal is also provided
to operate the starter inlet air solenoid valve.

The electronic computer layout is shown in Figure 59. Heat dissipation is

accomplished by means of the front mounting flange that channels internal heat
into the engine inlet case. All the C&A components are designed to endure a
50 g vibration environment radially from the engine centerline. The electronics
integrity is maintained by means of vertically mounting the printed circuit boards,
vibration absorbing grommets at the board mounting points, and viscoelastic
material sandwiched into the printed circuit board multilayer construction.

152
 TABLE 47. SENSOR REQUIREMENTS

Sensor Control Engine Condition Monitoring

Gas Generator Rotor Speed X X

Power Turbine Rotor Speed X X
Output Shaft Torque X X
Gas Generator Blade Temperature X X
Compressor Inlet Temperature X X
Burner Pressure X X
Engine Oil Temperature X
Engine Oil Pressure X
Magnetic Particle Detector(s) X
Oil Level X
Fuel Filter Bypass X
Oil Filter Bypass X
Engine Start (No.) and Time X
Counted

TABLE 48. FUEL CONTROL REQUIREMENTS

Requirement Accuracy

Semiautomatic Start Sequencing Pilot Initiates Fuel Flow

Wf/P Start, Acceleration, and Deceleration ± 2 Ib/hr, ± 4%, ± 6%
Proportional Gas Generator Governor ± 1%
Isochronous Power Turbine Governor ± 3% Transient, ± 0.25%
Steady-State
Inlet Guide Vane Scheduling ± 1% Stroke
Compressor Bleed Scheduling Two-Position
Torque: Limiting 115 ± 5%
Display See AV-E-8593
Loading Sharing ±5%
Turbine Blade Temperature: Limiting ± 16 0C
Display ± 50C
Power Turbine Speed Trim 85 to 115%
Power Turbine - Backup Overspeed Limiter 50-ms Response

153
 I
4i| > sA C
o
u
o
■a
\
o
0)
W

- I w

OS
in

154
b. Torque Sensor

The torque sensor measures the twist of the power turbine shaft by means of
three adjacent gear-toothed wheels, as indicated in Figure 60. One gear is
attached to the output shaft, and the other two are attached to the end of a torque
tube that is concentric and splined to the output shaft. The three-gear system is
a modification of the conventional phase displacement torque system, in that the
third (position) gear allows for compensation of shaft misalignment not accounted
for by the torque and reference gears. Three magnetic pickups monitor the toothed
wheels and provide an approximate sinusoidal signal to the electronic computer.
When torque is applied to the shaft, the resultant twist causes circumferential
displacement of the toothed wheels, which, in turn, causes a phase shift to occur
between the signals generated by the magnetic pickups. This phase shift is de-
modulated and is presented as a measure of shaft torque. A thermistor is in-
cluded with the pickups to enable the electronic unit to modify the phase shift
signal to compensate for shaft temperature change. A more detailed study could
indicate that the measurement of turbine blade temperature may be used to pre-
dict torque tube operating temperature, thus eliminating the requirement for a
thermistor. The electronic module that transduces phase shift and modulates
the output as a function of shaft temperatureis located in the electronic control.
Installation of the torque sensor assembly is shown in Figure 61. Full-scale
angular shaft twist has been calculated to be 5,6 deg and, coupled with the system
accuracy requirements, has required the relative movement between the pick-
ups to be held to a minimum. The three sensors and the temperature probe are,
therefore, formed into one module. The module itself is supported radially, as
indicated in Figure 60, to further minimize movement due to vibration and thermal
growth. The electrical cable and magnetic pickups are air-cooled to maintain a
maximum operational temperature of 450oF.

c. Radiation Pyrometer

Gas generator turbine blade temperatureis measured by means of a lens-type

radiation pyrometer. Installation of the probe in the turbine case allows a
straight sighting tube to be used, thus allowing consideration of either a lens
or aperture-type system. The aperture assembly requires the use of a glass
rod to transmit the light back to the fiber-optic cable and is limited by the maxi-
mum usage temperature (1000oF) of that rod. Metal temperatures of the turbine
shroud can reach 1900oF, and temperatures at the probe tip, 1200' to 1400°F. It
is doubtful that the 1000oF glass rod temperature limit could be maintained unless
cooled purge air is used.

The aperture system could also suffer from an erroneous maximum temperature
signal due to a buildup in light intensity as the blade pass JS by the probe tip. The
lens system has a more precise target area and would not experience this problem.
Based on the above criteria, the lens approach was chosen for the C&A application
and is shown in Figure 62.

The probe assembly Is mounted in the engine combustor case and is surrounded
by compressor discharge air at 750° to 800°F. Compressor air passes through
small purge holes drilled Into the sighting tube upstream of the lens to keep the
sighting area purged. The maximum operating temperature (1400oF) of the lens
will not be exceeded using the uncooled compressor air for purge,

155
 1
t8

i
■73

U
o
H

o
to
£
fe

156
.
 3

<
u
0
2

I
o
H

CD

157
 FIBER-OPTIC
CABLE

0.03 In. DIAMETER

SEAL
THROUGH FOUR PURGING AIR
HOLES
-3.879 in.

SECT. B-B
SECT. A-A

Figure 62, Optical Pyrometer

The lens focuses the light sigaal back onto the fiber-optic bundle, which then
transmits the signal to a silicon detector, located in the electronic control.
Maximum air temperature surrounding the fiber-optic bundle should not exceed
1650F, well below the 800oF capability of the cable.

Analytical estimates have been made to determine maximum gas generator

turbine blade metal temperature. These estimates are based on the blade cool-
ing scheme shown in Figure 63, where cooling air is supplied to the leading-edge
root cavity and makes three radial passes through pedestal arrays before dis-
charging at the trailing edge. The trailing-edge root cavity is cooled by a small
amount of air circulating from the base root to the platform plane. Maximum
metal temperatures are projected to occur along the pressure side leading edge
of the airfoil and between 50 to 80% of the blade height. Design of the probe
installation in the turbine case, illustrated in Figure 64, is such that the lens
target diameter (0.100 in.) will sweep along all but the initial 25% of the blade
length (nearest the root) on the pressure side leading edge, as the blade passes
by the probe tip. Although the point of minimum life on the blade is expected to
occur on the pressure side, trailing edge at approximately 50% of blade height,
time at temperature on the leading edge is directly related to blade minimum life.
The expected range of gas generator turbine blade metal temperature is 1200° to
1800oF,

d. T _ Sensor

The engine inlet air temperature sensor comprises a thermistor network housed
in a probe (Figure 65). Around this is an outer perforated tube that allows ade-
quate airflow to the probe and also provides a degree of mechanical protection.
The sensor is mounted in the compressor inlet annulus. The range of engine In-
let temperature is -650F to +1650F.

e. P Sensor
Co

Burner pressure is sensed by means of a varlable-capacltance-type transducer

housed within the electronic control to take advantage of the close proximity of
the transducer to the signal amplifier. This system results in a high degree of
accuracy over the large turndown in sensed pressure, 7 to 220 psia.

158
 •TR Al LING-EDGE
PEDESTAL ARRAY

COOLING AIR IN

Figure 63. Turbine Blade Cooling

PATH OF X
TARGET—/^

POWER TURBINE ROTOR

Figure 64. Optical Pyrometer Installation

ICU
 PROTECTIVE
OUTER SHELL

PROBE

Figure 65. Engine Inlet Air Temperature

Sensor

f. Fluid Controller/Overspeed Sensor

The fluid controller, shown functionally in Figure 66, meters fuel flow over the
range of 20 to 500 pph to the engine and varies the compressor inlet guide vanes
upon command from the electronic control. The package also houses a pneumatic
regulator and logic functions for the power turbine overspeed sensor. Package
implementation and cross-sectional views of the fluid controller are shown in
Figures 67 and 68.

The fuel metering system incorporates a diaphragm-actuated, throttling-type

metering head regulator, which regulates the pressure drop across the metering
valve to approximately 9 psid. The relatively low pressure drop is used to maxi-
mize the flow area of the main metering valve. The valve is positioned by the
electronic computer-controlled stepper motor. Fuel flow feedback to the com-
puter is provided by a resolver, which indicates metering-valve position. A push-
pull, solenoid-operated shutoff valve is included. A latching-type solenoid is used.
Redundant solenoid coils can be included if it is necessary to operate the shutoff
valve from the cockpit using airframe power.

160
 cc
o
O 3^
s
< o
> tu
o
^UJ i" „ -I»- O . 2 2 o
<UJ > Z <2 _i * u

UJ
cc
c
oc g
UJ
o
^>
s*
UJ
CC
a. ■*-*

C
OC
a. <
Z
ö-S3
oc
•a
u
0»

Sis: c
o
Uj UJ uj Ü
OC Op 2
T3
o- ^ s
S A m
S O OC
O UJ ^ tu,

§ *s«
5 = D UJ to
to
UJ ZUig IV
oc UJ oc a. I-
^ "8 w o?S
a oro. I
I
>
T3

3
1
U

1
3

8
i
w

«I

E
s
fa
 >
$
3
■v
%
u
i>

2
c
o
u
a

oo

163
The hydromechanical schematic also indicates a concept for providing power
turbine overspeed protection using the centrifugal deflection of a cylinder of
cantilevers for the overspeed signal. This deflection causes a regulated pneu-
matic pressure signal to be ported to a bellows actuator, which resets the
metering head regulator spring load, causing the regulator to close and thereby
reduce fuel flow.

The primary part of the power turbine overspeed sensor is a ring of cantilevers
splined to the forward end of the power turbine shaft (Figure 69). Centrifugal
deflection of the cantilever fingers changes the pressure recovery between sup-
ply and receiver ports, thus generating a pneumatic signal. The design airflow
required by the sensor is 0. 006 lb/sec. The range of power turbine overspeed
governor operation is above the normal maximum speed range of the main elec-
tvonic power turbine governor.

POWER TURBINE SHAFT-

CANTILEVER
FINGERS

DISCHARGE PORT

RECEIVER PORT

Figure 69. Power Turbine Overspeed Sensor

Implementation

The IGV actuator will be incorporated as part of the hydromechanical flow con-
trol package, and inlet guide vane position will be scheduled by the electronic
control as a function of corrected gas generator speed. The actuator require-
ment includes a maximum torque of 20 in. -lb, a 50-deg rotary movement, a
100-deg/sec response time, and a ± 1% accuracy. To eliminate vulnerable hy-
draulic lines, the IGV actuator is an integral part of the hydromechanical pack-
age. Output rotary motion is provided by a bellcrank. A resolver is used to
feed back IGV position to the computer. Once the fluid controller has been in-
stalled on the engine, the IGV system can be indexed via the adjustment mech-
anism shown in Figure 70. The differential set screw is rotated to provide proper
alignment of the IGV actuator output yoke with the engine IGV bellcrank.

164
 tt*

c
0)

<

AJ «tt i
^-.J o

165
g. Fuel Pump

The fuel pump must operate with suction fuel inlet conditions equivalent to an in-
let pressure of 1. 0 psi above true vapor pressure and a vapor/liquid ratio of 1. 0.
No dry-lift requirements apply. A maximum pressure rise of 650 psi and a
maximum flow of 500 Ib/hr are anticipated.

The 65,000-rpm-maximum-speed centrifugal pump, shown in Figure 71, will

provide the required starting flow of 10 Ib/hr and deliver the required flow and
pressure over the normal operating range. A separate high-speed inducer stage
is used, and a jet pump is used to charge the inducer.

INLET

MAGNETIC
PICKUP

DISCHARGE REF
TO INLET

Figure 71. 65, 000-rpm Centrifugal Pump

Figure 72 shows the head flow characteristics of the pump. The pump is required
to provide 23 Ib/hr at 21 psi rise at 10,000 rpm to meet engine starting require-
ments. Figures 73 and 74 illustrate temperature rise and overall efficiency
characteristics, respectively. The pumping system is located in the fluid con-
troller package and is detailed in Figures 67 and 68.

A magnetic pickup is also part of the pump housing to sense gas generator speed.
The unit measures rotational speed of the centrifugal pump and provides the sig-
nal to the electronic control.

166
 <1>
co _
E
a

ii
z

(u g 6

.5 «o -H

sis
CO

0)

&
•^
fc

n I 9)
s
13 9
to
w "-s
Z a.
m
o u
(M £
z
I
O

II Ui a
Z8 a«
in
3 'S o
*J o e
—
g m ^
.g «D ^H

oieo^tsin^mN CO

(tpsjpunH) ltd - 30UVH3Sia dWOd

fa

167
 6 ^-^- 80% rum J
100% rpm
^
5 60% ^S^*
IPIII^^>^^

>
o A /.
z
/

/y
UJ
5
üu. 3

J/
UJ FUEL: JP-4

f
TYPE: CENTRIFUGAL
< 2
oc
UJ

r
i

0
1 2
FUEL FLOW • Ib/hr (Hundreds)

Figure 74. Estimated Efficiency for 65, 0%0-rpm

Centrifugal Fuel Pump

h. Compressor Bleed Valve

The compressor bleed valve is a two-position, solenoid-piloted, air-operated

valve. The position of the valve is scheduled by the electronic control as a
function of corrected gas generator speed. The valve requirements include a
maximum operating pressure of 175 psi, an effective area of 0. 15 in.2, a
0. 200-ms response time, and a maximum operating temperature of 800oF.

The layout drawing and cross-sectional views are shown in Figures 67 and 68,
respectively. The solenoid for the control of the compressor bleed valve is
included in the hydromechanical package, instead of mounted to the valve. This
was done to keep the solenoid away from high-temperature bleed air.

i. Alternator
The alternator is a three-phase machine with a rotating permanent magnetic field.
The stator assembly comprises a three-phase output winding and a single-phase
control winding. Regulation is achieved by pulsing direct current through the
control winding in one direction to increase the alternator output and in the re-
verse direction to reduce the output. Provision was made in the design for two
ignition exciter power windings. Regulation for the alternator is accomplished
in the electronic control. The rectified dc output voltage from the alternator is
compared to a voltage proportional to the desired alternator output. The re-
sultant error is passed through a dynamic compensation section; it then causes
a switching current regulator to either "boost" or "buck" the alternator output,
depending on whether the alternator output is lower or higher than required. The
alternator is sized at 60w to provide electrical power for ignition and for the
electronic control.

168
Installation of the alternator on the tower shaft is shown in Figure 75. The inter-
face cabling for the alternator is brought out via a flange. The cable is passed
through a hole in a flange extension to the alternator support housing. The hole
will then be sealed with a grommet or an epoxy compound. The alternator in the
tower shaft configuration will require an individual connector for its interface to
the electronic control to facilitate assembly/disassembly.

OIL ^Eiuiri
PUMP DRIVE v-L x
FUEL PUMP DRIVE
GEAR MESH
GEAR MESH

k
TOWER ^
SHAFT

STATOR » CONNECTOR TO
COILS ELECTRONIC
CONTROL

OIL-MIST DRAIN TUBES

GAS GENERATOR
ROTOR GEAR MESH

Figure 75. Engine Alternator Tower Shaft

Assembly

The alternator/tower shaft assembly is designed so that all the major components
are replaceable. However, the entire assembly is considered as a line replace-
able unit. A review of the alternator/tower shaft concept indicated that the re-
quired spiral bevel gear alignment tolerances are attainable with a replaceable
unit. Close tolerances on critical dimensions will assure the replaceability of the
unit. Any shimming required would be internal to the alternator and not part of
the alternator installation or engine assembly procedures.

The alternator will operate in an oil-mist environment with cooling passages in

the support housing to allow draining of oil through the assembly.

j. Ignition System

The ignition system consiSre oi dual-circuit exciters operating dual-redundant

igniter plugs. The exciter is a hermetically sealed solid-state device that has
two independent circuits, both of which are capable of operating the two redundant

169
igniter plugs that are sequenced for alternate firing. The minimum exciter volt-
age available for ionization under a maximum demand condition is 5 kv, and the
maximum exciter voltage is limited to 5.5 kv. Energy available to the igniter is
1 joule.

The ignition system is powered by a regulated three-phase synchronous alternator

with a rotating permanent magnetic field. I he initiation of ignition is selectable
(automatic or manual), and the "off" condition is accomplished by shorting the
ignition winding outputs. Ignition system operation is required throughout 15 to
100% of the rated alternator speed and has an allowable duty cycle of 10 min "on,"
20 min "off."

Location of the exciter is defined on the engine cross-sectional layout drawing

(Figure 95).

k. Starter System

Engine starting requirements were estimated during the front drive engine studies.
As shown in Figure 76, the engine starter system is estimated to require a 5.2 ft-
1b output at 16,000-rpm engine rotor speed, where ignition will occur, and operate
up to 43,000-rpm engine self-sustaining speed. An air turbine starter configura-
tion was established as a baseline starter system during the front and rear drive
analysis.
SEA LEVEL/COLD DAY
12
MAXIMUIV1 TORQUE
10
m R

6 -MINIMUM TORQU L
■
A
""U~ •

2
^•^ r^ ' ^. 10*1■ .E
■
D *
O ' *^ •
cc n
O
Ulm
2
|\
i /
o^
a A ^ ■' /
S SELF-
oo
H-UJ
s
Q
IGNI' ION
•■
SUSTAINING
••
■
K/
STARTER C UTOFF
-8
10 20 30 40 50 60
GAS GENERATOR SPEED • rpm (Thousands)

Figure 76. Estimated Starting Torque and

Speed Requirements

The Garrett/AiResearch air turbine starter uses a single-stage, axial-flow

turbine in a parallel spur gearing arrangement (3-to-l ratio) with the sprag-type
overrunning clutch (Figure 77). The starter is lubricated with pressurized oil
from the engine. Starter output is a splined shaft with integral decoupler.

170
 u
t
n
CO

8
u
5

t-
t^

2
ä
•^
fc

171
i ■
The starter unit pilots into the engine inlet case and is bolted in place around the
outside housing. The air inlet shutoff valve is incorporated in the ATS package
and is actuated by a close-coupled solenoid. The solenoid is actuated by a com-
mand from the electronic control when the starter cutout speed is reached. A
weight and cost advantage results from having control of the shutoff valve sole-
noid in the electronic control, where sensed speed is available, rather than by
means of a control box and sensor mounted on the starter package. The cutoff
switch actuation speed is 43, 000 to 45, 000 rpm. Estimated performance of the
ATS is shown in Figure 78.

"RiüTR: 1 \ 1 1 (
1. STARTER DUCT INLET CONDITIONS:
PRESSURE - 45 psia
TEMPERATURE - 140oF
2. AMBIENT PRESSURE - 15 ptia
3. STARTER AIRFLOW - 35.5 lb/min
-4. INLET DUCT PRESSURE DROP ■ 0.5 pti
EXHAUST DUCT PRESSURE DROP > 0.5 psi

0 10 20 30 40 50 60 70
STARTER OUTPUT SHAFT SPEED - rpm (Thousands)

Figure 78. ATS Performance

The starter has a design life of 5000-hr overrunning clutch operation, or 10,000
start cycles.
For a single-engine installation, where an electric motor starter might better
fit the overall system requirements, the impact on the study engine configuration
will be minimal. An overrunning clutch arrangement would be used, and the
engine oil system could potentially be used for bearing and overrunning clutch
lubrication. Additionally, engine oil might be considered for cooling the field
coils of the starter motor. An electric starter would be located similar to the ATS,
except that a larger envelope and heavier support structure would be required.

172
 1. Lubrication System

The engine lubrication system is a dual-compartment, pressure-fed oil system,

The front bearing compartment is integral with the oil tank. A 15,000-rpm oil
pump/scavenge pump was sized in coordination with Sundstrand.

(1) Lubrication System Description

The major components of the engine lubrication system are: main reservoir
tank/heat exchanger, lubrication-scavenge pump, oil filter, bypass valve,
scavenge flow strainer, and relief valve. These components are shown sche-
matically in Figure 79. All components will be grouped within the gearbox in
the forward part of the engine.

MAR
MABING
COMFARTMtNT

Figure 79. Engine Lubricating System Schematic

Oil from the reservoir enters the pressure element of the pump. It leaves the
outlet of the pump and flows through an internal passage to the oil filter. The
system uses two valves that are parallel with the filter. One is a filter bypass
valve that allows oil to flow directly from the intake to the outlet side of the filter
should the filter become clogged; the other is a relief valve that acts as a safety
valve for the entire system, bypassing oil directly from the pump to the oil
reservoir should excessive system pressure develop.

Oil flows forward and aft from the core of the filter. The forward flow exits from
the filter and enters the front compartment to service the gas generator rotor
bearing (front of compressor), accessory drive gears, splines, bearings, and
the starter gears. After servicing these parts, the oÜ drains downward through
the gearbox components to the primary oil supply reservoir in the accessory drive
gearbox housing.
The aft flow of oil exits from the filter to the aft part of the engine to service the
power turbine bearings In the rear compartment. A sump at the bottom of the
rear compartment collects the scavenge oil, which Is, In turn, serviced by the
lubrication-pump scavenge element and returns the oil from the sump to the main
reservoir.
173
k
The oil filter recommended is a 40p screen-type filter, located in the accessory
drive housing in series with the pressure pump. A conical strainer screen will
be incorporated in the scavenge flow return line to prevent foreign particles from
entering the scavenge pump.
The filter bypass valve is a spring-loaded poppet valve installed in series with
the pressure element and parallel with the filter screen. If the filter clogs,
a differential pressure of 18.3 psi will open the poppet. The relief valve is
located in the accessory gearbox parallel with the oil filter. It is a spring-loaded,
poppet-type valve that bypasses oil to the oil reservoir to protect the engine
against excessively high pressures.

The main oil reservoir will serve as the oil cooler and will use the inlet air as
the coolant. It is designed to limit oil temperature at the lubrication-pump inlet
to a maximum of 250°F (l20oC).
Magnetic drain plugs are installed on the lubrication system components to be
used as system drains and to keep the system free of foreign materials.
Engine lubrication system design data are shown in Table 49.

(2) Oil Pump Description

The main lubrication pump is a two-element vane pump, mounted on the rear
face of the accessory drive gearbox. One element supplies oil to all lubrication
points within the engine, and the other element is used to scavenge oil from the
rear bearing compartment sump. The elements are driven by a common drive
shaft in one tandem cartridge arrangement, located in the accessory drive gear-
box.
Lubrication oil from the service element will enter through an internal port on
the side of the pump and exit through an internal port to the filter screen. An
elastomer O-ring, incorporated between the supply and scavenge ports on the
pump housing, minimizes leakage.
The oil pump configuration is based on a fixed-displacement vane pump developed
by the Sundstrand Corporation. The pump element design is illustrated in Fig-
ure 80, and the final pump layout, based on the lubrication system design require-
ments, is shown in Figure 81.

(3) Impact of Oil Sump Heat Load on Engine Inlet Air

Design of the engine inlet particle separator forces clean air along the periphery
of the oil tank and into the compressor inlet annulus. Oil in the tank can reach
250°F and thus can cause an increase in engine inlet air temperature above ambient
conditions.

The engine heat loads, applicable to the lubrication system, were determined at
both SLS maximum and idle speed conditions, and are summarized in Table 50.
Although an idle heat load of 115.5 Btu/min was calculated, it was recommended
that a 150 Btu/min heat sink capacity be designed.

174
A typical helicopter mission cycle includes a cruise condition at 5000 ft, 100 knots,
and 609r power. The engine heat load at this operating point was obtained by scaling
the SLS maximum power heat load with the cruise condition gas generator speed.
The resulting increase in compressor inlet air temperature, at all three points,
was determined. Engine influence coefficients relating shaft horsepower (shp)
and specific fuel consumption (sfc) to changes in compressor inlet air tempera-
ture were determined from performance deck data and established the perform-
ance penalties shown in Table 51.

TABLE 49. ENGINE LUBRICATION SYSTEM DESIGN DATA

Engine Lubrication System Data

Type Pump Recirculating, Positive-Displacement

Oil Specification MIL-L-7808D
MIL-L-23699
Oil Temperature at Pump Inlet -65 to 250° F
Valve Settings:
Filter Bypass Valve 15 to 20 psi
Relief Valve TBD

Oil Pump Design Requirements

Power Setting
Design Criteria Military Idle

Pump Speed, rpm 15,000 11,500

Maximum Oil Temperature, 0F 250 250
Minimum Oil Temperature, 0F -65 -65
Maximum Pump Environmental
Temperature, 0F (15 sec/continuous) 302/176 302/176
Pressure Pump
♦Oil Flowrate, lb/min 43 33
Pump Pressure Rise, psid 70 43
Minimum Pump Inlet Pressure, psia 6.75 6.75
Scavenge Pump
**Oil Flowrate, lb/min 44 34
Pump Pressure Rise, psid 20 12

♦Supply pump oil flowrate value includes 20% overcapacity for growth and
aeration.
♦*Scavenge pump flowrate includes four times capacity for aeraiion plus a
20% growth factor.

175
 ^
ROTOR

ROTOR
^
L HOUSING-i ^XZ
VANE
PAIR

PORT
^
t^l

"
ROTOR
M VANE
VANE CORE
ENLARGED
INSTALLATION
EXPLODED VIEW

Figure 80. Lubricant/Scavenge Pump Element

176
. ;
 etu

00

177
TABLE 50. SMALL GAS TURBINE ENGINE HEAT GENERATION* SUMMARY

Heat Generation - Btu/min

Compartment Location SLS Ground Idle

Front Compartment Ambient -47,8 -41.1

Middle Compartment Ambient Air Cooled Air Cooled
Rear Compartment Ambient +17.7 +ll
Compartment Seal Friction All Self- All Self-Acting
Acting
»0 %o
Seal Leakage Negligible Negligible
Gas Generator Thrust Bearing (Duplex) 132.0 85.0
Power Turbine Thrust Bearing 10.5 10.5
Gas Generator Journal Air Cooled Air Cooled
Power ^urbine Journal Air Cooled Air Cooled
Gear^ix «58.0** <58.0
Toti +170.4 <115.5

*A jplicable to lubrication system

**Q
[ rs - 13 Btu/min, Qbearings 45 Btu/min

TABLE 51. PERFORMANCE PENALTY HEATING OF ENGINE

INLET AIR DUE TO ENGINE HEAT GENERATION

Increase in Decrease in Maximum

Engine Condition ^Tt2 sfc. % Available shp, %
SLS Maximum Power 3.52 0.47 1.4
SLS Idle Power 5.51 0.73 NA
5000 ft/100 kt 4.33 0.58 NA
60% Power

The performance penalties are not considered excessive but, to some degree,
can be compensated for. As shown in Figure b2, the oil sump/heat exchanger
and IPS could be configured to remove heat from the oil by the IPS scavenge
flow only. Entrance to the compressor at the bottom of the engine would be
blocked, and fins would be added to the bottom of the oil sump. In addition,
the upper portion of the accessory drive cavity cr.ul'j be insulated to minimize
the possibility of adding heat to the compressor inlet airstream. Some in
crease in size of the airflow ducting for the modified IPS configuration is antici-
pated, but the Impact is considered minimal.

178
179
m. Emergency Lubrication System
The original intent of the program was to evaluate a backup emergency lubrication
system with a 6-min capability. Discussions with Army program management and
vulnerability personnel indicated that a 30-min backup system capability was more
consistent with the airframe gearbox design goals. The study was, therefore,
revised to include consideration of both a 6- and 30-min capability system. The
most promising configuration for a backup oil system was identified as an air/
oil-mist system using a separate oil supply, integral with the main oil tank.

Two approaches to air/oil-mist lubrication systems have been pursued. Oil-

mist systems have been tested for full-time operation that use significant air-
flows and very small oil flowrates. Limited tests have also been conducted on
emergency systems that used much smaller air flowrates at the expense of higher
oil flowrates. A conservative approach to a system design for a 30-min capa-
bility would consider the steady-state design criteria. For the purpose of this
evaluation, reliable operation over a 30-mln time span was considered a steady-
state condition. The engine performance penalties for the bearing air supply
using the steady-state design criteria were acceptable during backup system
operation, and the small oil flowrates were compatible with a goal of a minimum
emergency oil reservoir size. During the course of the evaluation of a system
using the full-time system criteria, it was determined that the bearing coollng-
alr/englne-lnlet-alr heat exchanger size requirements were unacceptable from
an engine installation standpoint. This evaluation assumed a reasonable Increase
In demonstrated bearing Inlet temperatures.

A system design using the limited test data on air and oil flowrates for an emer-
gency system was then evaluated. The study showed that the bearing cooling-
alr/englne-inlet-air heat exchanger could be Integrated acceptably as a part of
the engine inlet particle separator bypass duct. The required emergency oil
reservoir capabilities could also be incorporated without significant design
compromises.

A system design that Is optimized for reliability and minimum impact on the
engine design will probably be somewhere between the two cases evaluated. Ex-
perimental verification of a detailed system design will be required to confirm
the exact applicable design criteria.

(1) Design Requirements

The design requirements for the air/oll-mist system were established by review-
ing documented test programs, by referring to design guides for commercial
air/oil-mist installations, and through informal discussions with personnel now
involved with experimental efturts. The most extensive testing of air/oll-mlst-
lubrlcated, high-speed, small-bore bearings Is being performed by SKF Industries,
Inc., under NASA contracts, taformal discussions with NASA and SKF personnel
helped to establish the guidelines for this application study.

The integrated accessory systems engine bearings do not exceed size and DN
values of 19 mm and 1.24 x 106, respectively. Maximum bearing speed would
be 65,000 rpm. The engine bearing configuration is Illustrated In Figure 83.

180
 OiNfRATOn Tl'IMIINC
ROTOR ROTOR

MARINO ON OR OiAR
lOCATION OlMHtri«, mm tm mCH-UNC VILOCITY

■1 » H.ooe 12» a 10*

■I 12 »goo 0.» a 10»

■a 13 10,000 ON a 10*

M 12 16.000 0 70 a 10*

M 12 «.000 0.70 a 10*

»000
01 17,017 nimm
H.I 30.000

M.2 »,000
02 11.2K H/mui
712 11,000

42.7 »000
01 11.101 Hlmm
1t.7 «.000

M If 30.000 0 0» a 10*

Figure 83. Tower Shaft Drive System Bearing

DN's and Gear Pitch-Line Velocity

(2) A System Design Based on Full-Time Design Criteria

Results of preliminary testing, based primarily on a 46-mm ball bearing running
at 38,000 rpm and a 25-mm bearing running at 44,000 rpm, indicated that a 6-min
or 30-min air/oil-mist lubrication system was possible for the engine size and
speed range studied. The 46- and 25-mm bearing-air/oil-mist tests used 15
and 10 scfm of airflow, respectively. These data points were plotted as a function
of DN and linearly extrapolated to determine air flowrates for the bearing sizes
and operating speeds studied. A precise extrapolation of these data would involve
determination of the relative bearing heat loads. Since prediction techniques
for bearing heat loads in an oil-mist environment were not established, a linear
extrapolation was considered to be a conservative approach, since the maximum
system bearing DN was between the data points considered. The requirements
are detailed in Table 52.
No significant distinction between the requirements for a system with a 30-min
and a system with a 6-min capability was made in terms of air or oil flowrates.
High-speed bearings usually fail rapidly after the lubrication or heat exchange
medium is removed, or reduced to a level where the bearing fits and clearances
are mismatched due to thermal distortion. Some reduction in air/oil flowrates
for the lighter loaded gearbox bearings and the gearbox drive gears may be pos-
sible for a 6-min system, but would not significantly impact the total system
design requirements.

181 w. 1
 TABLE 52. AIR AND OIL FLOW REQUIREMENTS

Bearing Position Airflow, iscfm Oil Flow, cc/min

Tandem Ball Pair 28 0.28
Tower Shaft Drive Ball 11 0.11
♦Power Turbine Ball 23 0.23
Tower Shaft Ball 4 0.04
Tower Shaft Roller 4 0.04
\
Fuel Pump Ball 7 0.07
Fuel Pump Roller 7 0.07
Gears (3) 15 0.15
Total 99 0.99

General Notes:

(a) Air in at 170°F

(b) The above schedule uses 0.01 cc/min oil flow for each (1) scfm
a rflow.
(c) Resulting bearing outer ring temperatures ~350oF

♦Air/oil flow doubled because of hot environment

(a) System Implementation

The air/oil-mist distribution system implementation is shown in Figure 84.

Static pressure is bled from the 10:1 centrifugal compressor at a selected
location along the shroud. Because the airflow required (4.2% Wae) would cause
an unacceptable performance penalty during continuous operation, the airflow is
bled only upon loss of primary oil system pressure by action of an on-off, oil-
pressure-operated valve. A significant part of the successful testing accom-
plished to date used low-temperature bearing cooling air (170°F). As a first
attempt at a system design, this limit was considered. To maintain this limit,
a heat exchanger would be required. The system pressure requirement was
sized for an altitude (5000 ft) idle power operating condition, but must also
function at a sea level maximum power point. To compensate for the variation
in static pressure, a pressure regulator is required.

As mentioned previously, the original oil reservoir configuration assumed an

oil supply integral with the main oil tank. However, both SKF and commercial
suppliers (Alemite) of oil-mist systems have determined that to keep the oil
particles from wetting out on the distribution piping, the mist velocity should
not exceed 24 ft/sec. This requirement imposes large line sizes and the design
difficulty of getting the mist to the various bearings.
Considering the very small total backup oil quantity required (30 cc), it was
determined that a better approach would result with three oil reservoirs at the
major bearing locations: the accessory gearbox, the front bearing compartment,
and the rear bearing compartment. This approach will allow high supply air
velocities and small tube diameters. It also results in minimum pressure loss,
which will keep the system supply pressure requirement at a minimum and,
consequently, the temperature rise that will size the heat exchanger.

182
 cc
<
tu
is
>
cc IM OC

S i|
OC ü tu o
cc

m.

I
CO

a
o

X!

IN

<

j83
Aspiration of the oll will occur at the delivery points rather than having one
remote reservoir and aspirating device. The geometries of this system have been
defined and are shown in Figures 85 through 87. The distribution manifold sup-
plying mist to the various bearings is illustrated in these figures. The aspira-
ting devices would be located in close proximity to these manifolds.
(b) Design of Distribution System
Having established a design philosophy for the air/oil-mist system and supply of
the mist to the various bearings and gears, the minimum operating pressure and
heat exchanger size were determined. A worst-case plumbing roadmap (rear
compartment) was laid out showing line lengths and sizes, bends, and valvlng so
that the system pressure drop could be calculated. (See Figure 88.) The system
was sized for operation at a 5000-ft Idle-power condition and resulted In a static
pressure requirement of 23.2 psla at the compressor shroud. Figures 89 and 90
show temperature and static pressure at various locations along the compressor
shroude and Indicate that, at the pressure tap location required, an air temperature
of 660 F would result at a sea-level, hot-day, maximum-power condition. The
maximum air supply temperature sizes the required heat exchanger.

«■UTIOM MMHraiO

Figure 85. Control andAccossory Gearbox

Bearing Compartment

184
.
 I
2

185
 e

I
ao
'J

u
S
n

t-
oo

186
 i

cc
2
Ui

5
< o
be
0- CD c
< 9

<
u. s
3
» 9 CO

i '
3 s E J
5-
00
00

g
<

187
Figure 89. Estimated Shroud Static Pressure
Variation

Figure 90. Estimated Compressor Shroud Air

Temperature Variation

188
The required heat transfer surface area was found to be impracticably larp.e due
to the high heat load (890 Btu/min) and low cooling air velocities (150 to 175 ft/sec)
resulting in low heat transfer film coefficients, coupled with the requirement for
designing to hot-day inlet air conditions of ISS0!''. Therefore, the maximum air
supply temperature was increased to 250oF, and the impact on heat exchanger
size was determined. (See Figure 91.)

Figure 91. Finned Inlet Duct Heat Exchanger

Size Required to Cool Bearing Oil-
Mist Airflows

The 250oF air supply temperature is considered to be a reasonable extension of

testing accomplished to date and is consistent with our maximum normal oil
system supply temperature of 250oF.

The engine inlet annulus presents the largest available surface area for heat
transfer purposes. With the engine diameter fixed, the only variable is duct
length. Figure 92 defines the heat transfer problem; the available surface area
on the engine inlet would result In a 3160F air temperature. To maintain the
250°F limit, the engine length would have to be increased by 4.8 in. Total normal
engine length is approximately 36 In.

(3) System Design Based on Emergency Operation Design Criteria

Limited test data on a backup oil system designed for emergency use that utilized
significantly lower air flowrates was also considered. The test point considered
had demonstrated a 30-min operation of a 46-mm ball bearing operating at 38,000 rpm.
The airflow used was 0.84 scfm at 200°F Inlet temperature, and 4 cc/mln of oil
flow.

189
 •» f'~Mi;;"~~^l'.,T:"'-'» i m^^ r

IMS
MAT nOUNQU ONOtN ■ k!*4*

Figwre 92. Finned Inlet Duct Heat Exchanger

Size Required to Cool Bearing OIl-
MIsi Airflow From 660 to 250oF

These data were linearly extrapolated to define the air and oil flowrates for this
application. The estimated required air and oil flowrates are tabulated below:

Bearing Position Air, scfm Oil, cc/min

Tandem Ball Pair 1.64 7.44

Tower Shaft Drive Ball 0.65 2.95
Power Turbine Ball 0.65 2.95
Tower Shaft Ball 0.2 0.9
Tower Shaft Roller 0.2 0.9
Fuel Pump Ball 0.35 1.6
Fuel Pump Roller 0.35 1.6
4.04 18.34
Gears 1.2 6.0
Total 5.24 24.34

The required oil-mist distribution system is shown schematically in Figure 93.

The lower total system air flowrates would eliminate the requirement for a system
shutoff valve. The backup system would then operate continuously.
190 ^^_ ,lJ
 yv COMPRESSOR

AIR/AIR HEAT
EXCHANGER

PRESSURE
ACCESSORY
REGULATOR GEARBOX

FRONT BEARING

m
ASPIRATOR
COMPARTMENT

REAR BEARING
COMPARTMENT

Figure 93. Emergency Operation Air/Oil-Mist

Lubrication System

Due to the lower air flowrates, the use of a single aspirator and oil reservoir was
considered. The required oil reservoir (720 cc) is best contained within the main
oil tank. The oil-mist distribution system around the bearings and gears will be
similar to that previously described.

The air/oil-mist flowrate established for the rear compartment will allow the oil
mist to be transferred in a reasonable line size, thus negating the requirement
for a separate rear reservoir. A required rear reservoir size of 90 cc would
also be difficult to package in the rear compartment.

The required engine bleed air flowrate was established at 5. 24 scfm. A bearing
system supply air temperature goal of 200° F was also established. Based on the
flow schematic shown in Figure 88, a total system pressure loss of 6.4 psi was
estimated. This established an impeller shroud bleed location of 1.1 in. to pro-
vide the required pressure at the 5000-ft idle condition. A supply air tempera-
ture of 530°F would then be provided at sea-level, maximum-power condition.

An air/air-heat exchanger was then sized to provide the required cooling of the
bearing supply air to 200°F. The IPS bypass duct was selected as the best loca-
tion for the heat exchanger. A simple heat exchanger that would use the outer
surface of the bypass duct would provide 1.13 ft2 of heat exchange area without
the use of fins. For a 200oF supply temperature, an area of 1. 05 ft2 would be
required. The selected heat exchanger location was, therefore, determined to
be acceptable. The impact of air supply temperature on heat exchanger size,
for this heat exchanger location, is shown in Figure 94.

191
 Figure 94. Air Temperature Effect on Heat
Exchanger Size

A system based on the emergency operation design criteria was determined to

be feasible for the engine configuration studied. However, before a final system
design Is established, further experimental verification of the design criteria for
air and oil flowrates will be required on both the component and system levels.

(4) Conclusions and Recommendations

Industry testing of an air/oil-mist lubrication system indicates that a 6-min or

30-min system capability is possible for the engine size and speed range studied.
For a system using the design criteria established for full-time, steady-state
operation, the requirement for a relatively low oil-mist supply temperature, based
on testing accomplished to date, imposes impractical heat exchanger requirements
on the engine design.

A system using the design criteria established for emergency operation provides
acceptable heat exchanger and oil reservoir designs. However, test data verifying
this system approach are preliminary in nature and require further confirmation.

It is recommended that further test programs be directed toward higher air/oil-mist

supply temperatures. A helicopter, with even a limited operational envelope, must
be designed for standard hot-day sea level, Mn = 0 inlet temperatures of 135° F.
Assuming a nominal supply pressure requirement for the oil-mist system, an en-
gine bleed air temperature of several hundred degrees higher can be anticipated.

192 ■ niri' im'■ i i ;iiiilMjCfr-'"-»■"^--1--^^.■*j«' 11 fMiin■ Ta' 1 am-mailM»ri,»*tr*—"^"-■ ^•4**^^

The oil-mist distribution system valves, pressure regulator, lines, aspirator,
and oil-mist nozzle pressure drops define the system supply pressure that must
be provided by compressor bleed. The location of the pressure tap is determined
by the idle power conditions at the selected maximum operational altitude. The
maximum air supply temperature then occurs at the sea-level, maximum-power
conditions. The maximum supply temperature is, therefore, directly related to
the system pressure drop requirements. An optimized system design should in-
corporate a minimum pressure drup oil-mist feed system.

Minimizing the required air flowrate is also a desirable goal, as that decreases
the engine performance penalty and reduces the air/air heat exchanger size for
a given air supply temperature. However, for emergency system operation, it
is more desirable to reduce or eliminate the heat exchanger requirements at the
expense of airflow. For part-time emergency operation, the airflow penalty can
be tolerated, while a large heat exchanger imposes full-time cost, weight, and
volume penalties.

193
 SECTION VIII

REAR-DRIVE ENGINE WITH INTEGRATED C&A COMPONENTS

A. FINAL ENGINE DESIGN LAYOUT

Design layouts of each C&A component, component interconnects and interfaces,

and sensor interfaces have been integrated in the final engine layout drawing. The
final tower shaft drive/air turbine starter configuration is shown in Figure 95,
which also defines the C&A system elements to show the relationship with the other
engine systems. •

The air turbine starter is located in the engine nose cone and is mechanically
connected to the gas generator rotor through an overrunning clutch. The starter
air control valve and pilot solenoid are also located in the nose cone. The starter
components provide limited front vulnerability protection to the fuel and oil sys-
tems.

The oil sump is located in the front bearing compartment and also acts as the oil
system heat exchanger. A vane-type positive-displacement oil pump is In the
same location and is driven at a maximum speed of 15,000 rpm by a tower shaft.

The engine alternator, which has been made integral with the tower shaft, pro-
vides the required engine and electrical power for starting and control, and
operates at a maximum speed of 30,000 rpm. The alternator operates in an
oil-mist environment.

The main fuel pump is located in the fluid controller and is driven at a maximum
speed of 65, 000 rpm by the tower shaft. The pump is a two-stage centrifugal
unit, with a jet pump inducer, and is designed to operate at inlet conditions of
1 psi above TVP and a V/L of 1. 0. ,

The fluid controller is located, adjacent to the compressor case, on top of the
engine. The fluid controller meters the required engine fuel flow and actuates
the IGV's in response to stepper motor input commands from the electronic
control. The only external fuel lines are for the fuel supply and the discharge
line to the fuel manifold. A mechanical power turbine overspeed sensor inter-
faces directly with the fluid controller.

The full-authority electronic control is located adjacent to the fluid controller

for vulnerability protection and to reduce the interconnecting cable length. The
electronic computer controls all engine starting, transient, and steady-state
functions (except for the redundant power turbine overspeed function) in response
to electrical interface command signals from the airframe. Engine sensors are
provided for compressor inlet temperature, burner pressure, rotor speeds,
power turbine torque, and gas generator turbine blade temperature and interface
with the electronic control. Cooling for the electronic control is provided by
heat sink to the engine inlet air. Cooling for the torque sensor and cable is
provided by compressor bleed air, which has been cooled by an air/air heat
exchanger.

Airframe interfaces for starter air, fuel supply, and the electrical airframe
input commands are on the top of the engine.
194 A
 FUEL PUMP DRIVE

TOWER SHAFT/ALTERNATOR ASSEMBLY

FLUID CONTROLLER AND

•FUEL PUMP

COMPRESSOR
BLEED
VALVE

IGNITION EXCITER

AIR TURBINE STARTER ASSEMBLY

INLET PARTICLE SEPARATOR

ENGINE INLET AIR TEMPERATURE SENSOR

Figure 95. Final Tower Shaft Drive/Air Turbine Starter Engine Cross Section

/ 195
 RADIATION PYROMETER
POWER TURBINE
OVERSPEED SENSOR
TORQUE
SENSOR
MOUNTING
RING

TORQUE
SENSOR
AND
COOLING
SHELL

EJECTOR

AIR/AIR HEAT EXCHANGER

SCAVENGE DUCT
'■"t-t—t > t i t
0 12 3 4 5 6
SECTION A-A SCALE • in.

^
The engine layout (Figure 95) shows that the control and accessory integration
techniques have resulted in a minimal impact on the engine front and side pro-
jected areas, and that a system designed with consideration for vulnerability
resistance has added benefits in weight, volume, and simplicity.

B. C&A COMPONENT CRITICAL ITEMS

During the detailed design phase, several C&A component areas were identified
as requiring experimental effort to verify the selected design approach for a 1977
engine development time frame. These areas are outlined below and can also be
used for reference in definition of future C&A experimental development efforts.

• Fuel Pump Inducer - Based on pump development work accom-

plished to date, a three-stage pumping system has been pro-
posed to meet the specified system requirements The boost
elements comprise a jet pump plus a centrifugal loducer. The
main stage is a GS.OOO-rpm centrifugal pump.

The pump specification for this program requires operation at

inlet conditions of 1.0 V/L and l-psi NPSP. Present technology
in fuel systems is for operation at 0.45 V/L and 5-psi NPSP.
While tests have demonstrated the al ility of fuel boost pumps
to operate at conditions higher than 0.45 V/L, sustained opera-
tion at 1. 0 V/L, without a low-speed inducer, represents a sig-
nificant advance of the state of the art. The high speed
(65,000 rpm) and low flow (500 Ib/hr) required for this
pump indicate that an experimental program is required to
demonstrate inducers to meet these pump inlet requirements.

• Starter Overrunning Clutch - The proposed starter design

achieves minimum weight and maximum performance efficiency
by locating the overrunning clutch in line with the engine rotor.
This results in a clutch overrunning speed of 65,000 rpm. For
this application, engine-supplied oil is available for clutch
lubrication.

High-speed sprag clutch problems generally fall into two cate-

gories. In the driving mode, fatigue and sprag rollover are
the typical failure conditions. Overrunning at high speeds can
lead to excessive clutch and race wear, which ultimately result
in rollover or inability to drive.

Since the proposed clutch application involves low driving torques

and speeds considerably in excess of previously demonstrated
capability, an experimental program was recommended to verify
the 65,000-rpm overrunning capability.

i<)7
PRECEDlrG PAGB BLANK-NOT FILMED
 01

I
£
H

a.

g,

J.JJ... .. .■^■.„i ,. ...-.- .: .. ' ..■---;.^—„-^ .

1«Ä ■ •■■■ •■■-•'— vr ii«Mii*T«- ■— — MM
Oil Pump Performance and Endurance - The limit on operating
speed for a single-stage, positive-displacement oil pump is
determined by the inlet cavitation effects. For the engine oil
system, a vented oil tank is proposed that provides essentially
ambient inlet conditions to the oil pump pressure stage. Since
the engine is being designed to operate up to a 20,000-ft altitude,
an inlet pressure as low as 13.75 in. absolute of mercury is
anticipated. Experimental verification of the capability of an
oil pump to operate at the maximum speed (15,000 rpm) and
maximum altitude inlet conditions (20, 000 ft) without sustaining
cavitation damage was recommended.

Radiation Pyrometer - As engines get smaller and operate at

higher temperatures, accessories such as the radiation pyrom-
eter have greater demands imposed on them. The radiation
pyrometer, which is emerging as an important engine accessory,
is one item, in particular, that has to operate in a hostile
environment, yet provide accurate data vital to the perform-
ance of the engine.

The temperature profile defined for the pyrometer environment

on this program is as follows:

1. CDP (for cooling and purging) = 750oF maximum

2. Pyrometer location cavity = 800oF maximum
3. Lower seal metal temperature - 1400oF maximum
4. Turbine casing temperature = 1900oF maximum

However, the limiting temperatures of the individual elements

of the pyrometer, for reliable operation, are as follows:

1. Sapphire lens and mounting flange assembly = 1400oF

2. Solid fiber-optics light pipe = 1000oF
3. Flexible fiber-optics c^ble =• 800oF
The pyrometer proposed for this application includes a sighting
tube with a lens located just inside the 800oF cavity, thereby
negating the need for the solid light pipe and providing a thermal
environment well within the rating of the lens and flexible cable.

A further feature of the design is the absence of a CDP purging

port. Instead, the pyrometer includes an inherent purging con-
figuration, formed by small holes, located in the pyrometer
outer tube upstream of the lens. It is intended that air be
drawn in through these holes from the 800oF cavity to provide
purging airflow to the lens.

This concept will reduce the pyrometer coat, provide minimal

size, and ensure that the hot CDP purging air is inhibited from
interfering with the flexible fiber-optics cable.

Experimental verification of the above-mentioned purging con-

cept was recommended.

199
Electronic Control Cooling - The proposed mounting surface
for the electronic computer provides heat transfer to the
compressor inlet air.
A review of installation requirements for this configuration
indicates deparlures from previously developed cooling concepts
in that forced air-conveetive cooling is proposed instead of using
fuel as a heal sink. This concept, based on locating the control
with a heat exchanger interface with the particle separator,
should provide an adequate means of cooling. This approach is
advantageous over its fuel counterpart from a viewpoint of cost
and reliability. The anticipated primary source of heat for
the computer is via radiation from the compressor; however,
the magnitude is presently unknown.

An improved conductive heat transfer concept (from components

to heat sink) has also evolved for this control. This concept is
based on heal transfer from the individual "chips" to the com-
ponent leads and then via a special copper layer in the PCB to
form a thermal conductor to the flexible interconnecting cable,
which finally terminates at the heal sink. Initial analysis indi-
cates that this concept has the potential of providing a very low
thermal impedance path from all compoa its to the heat sink.
Experimental verification of this approach was also recom-
mended.

Electronic Control Vibration Isolation - Previous CECO efforts

in designing the printed circuit boards (PCB's) for the engine
vibration environment used internal damping and isolation
mechanisms for the PCB's that realized the goal of minimizing
the component and PCB vibration level to 30 g's with an input
excitation of 5 g's (20 to 460 Hz) and 20 g's (4G0 to 2000 Hz).
The electronic vibration environment for this application is
estimated at 50 g's.

The elements not protected by the damping and isolation

mechanisms (radiation pyrometer detector and first-stage
signal pi'ocessing, together with the pressure transducer) are
rated for 30 g's, but do not experience g levels significantly
above the input excitation, as they are hard-mounted to the com-
puter housing.

The recommended alternative approach for this application

would use external vibration isolation instead of the internal
isolation/damping concept. This approach appears advan-
tageous, In that the radiation pyrometer and its signal detector
would be protected from the 50-g input. The alternative to
this would be the design and qualification of components rated
for 50 g's, which would be less reliable.

It was recommended thai vibration isolation techniques be

experimentally verified.

Power Turbine Overspeed Sensor - Free turbine engines re-

quire a fast-response overspeed governor for protection to
prevent runaway speed. Also, in the cases where the engine
200
control system is electronic, it is advantageous to have a power
turbine overspeed governor that is redundant and operates inde-
pendent of the electronic control. Conventional mechanical
speed sensors operate in the 5,000-rpm speed range at 250oF,
whereas, in this system, rated-power turbine speed is 36,000
rpm and ambient temperatureis SOOT.

The speed sensor is formed by a set of closely spaced canti-

levers that are arranged in a circle around the power turbine
shaft. As the power turbine shaft speed increases, the centrif-
ugal force causes the cantilevers to move radially outward.
The cantilever motion is used to vary an orifice and thereby
generate a pneumatic pressure signal. An increase in power
turbine speed operates to reduce the metering head across the
fuel metering valve. Hence, a turbine overspeed causes a
proportional reduction in fuel flow.

Critical items in the operation of this governor are the speed

sensor and the dynamic response of the governor. A speed
sensor of this configuration has previously been tested to
20,000 rpm. For this system, the sensor has to be designed
to operate at 36,000 rpm in surrounding temperatures up to
approximately 800oF. The required response from overspeed
to reduction in fuel flow is on the order of 50 ms.

An experimental program was recommended to demonstrate

the performance and durability (in an engine-simulated
environment) of the speed sensor and to confirm the dynamic
response capability of the overspeed governor.

Starter Overrunning Clutch and Decoupler Endurance - The

selection of a sprag-type overrunning clutch necessitates in-
corporation of a decoupler mechanism to prevent overspeed
of the starter turbine wheel in the event of a clutch or bearing
failure. At the design speed of 65,000 rpm, it is important
that the decoupler provide complete disengagement of the
starter from the engine. Existing decouplers use some type
of supporting bushing or bearing, which continues to run after
decoupling occurs. It is felt that the new decoupler incorpo-
rated in the design concept approach must be experimentally
verified for the proposed high-speed application.

Starter Control Valve - A starter control valve, which is

integral with the starter and incorporates a controlled opening
poppet valve, has been proposed for the engine accessory pro-
gram. Current air turbine starters use separate starter con-
trol valves that open between 10 to 40 psi/sec to limit the
impact torque. The opening rate depends on the maximum
allowable impact torque, which varies with different applica-
tions. The majority of starter control valves incorporate
butterfly modulating elements, where the valve opening rate
is controlled by a simple orifice circuit in the actuator head
supply line. By contrast, a poppet valve ttscA* to snap open,

201
 and the rating device required to control its opening rate is
more complex, generally consisting of an incompressible
fluid damper. The proposed valve will be normally spring
loaded and pressure-assisted closed and will incorporate a
device to control the opening rate. An experimental program
was recommended to evaluate the proposed valve configuration
with an incompressible fluid damper.

C. ENGINE DESIGN CRITICAL ITEMS

During the course of the program, several engine design areas that interface
with the C&A systems were identified as requiring experimental effort to verify
or improve the recommended baseline engine configuration. These areas are
outlined below and can be used for reference in definition of future experimental
development efforts.

• Air Bearings - The primary engine oil-lubricated bearing

system is inherently vulnerable because of the large oil tank,
heat exchanger, and fuel system. Further development effort
and demonstration of radial and axial air bearings is in order
to verify a near-term applicability.

• Oil-Mist Backup Lubrication System - The study of an emergency

oil-mist lubrication system for this application indicates that a
system design, based on the limited test data available, is feasi-
ble for a 30-min system capability. Further experimental effort
to verify the required air and oil flowrates and directed toward
using higher air supply temperature is in order to confirm the
applicability of the backup system approach.

Additionally, testing of the total air/oil-mist system is recom-

mended to verify the system design.

• Advanced Electric Starters - For certain engine applications,

an electrical starter system may be considered. Develop-
ment effort for an advanced, high-speed starter that uses
engine-supplied oil for lubrication and cooling is in order.
The program goals would be directed toward minimizing
starter weight and volume.

• Air Assist During Engine Starting - Lower engine starting

speeds, with resulting lower starter horsepower requirements,
can be achieved by improving fuel atomization at start. A
small quantity of supplementary air for the air-assist nozzles
during the start cycle will provide the dv sired improvement.
A design study and experimental effort is recommended to
determine the best means of supplying the additional air and
to identify the overall impact on the engine and starter system.

202
 • Engine/Airframe Interfaces - Engine/airframe interfaces
for future applications are a fertile area for further study.
Advanced engines will be much smaller and lighter in weight
than current powerplants. This will allow a greater degree
of freedom in establishing the optimum installation since the
engine weight will have a small effect on the overall airframe
CG. Vertical engine installations, nonparallel gas generators
and power turbines, or two gas generators supplying one power
turbine are potential configurations which can be considered.

The power turbine interface, drive location for airframe

accessories and the technique for supplying air for ECS and IRS
are areas requiring engine/airframe trade-off studies to confirm
the best overall system approach.

D. LONG-RANGE ENGINE DESIGN CRITICAL ITEMS

From an ultimate C&A system reliability and vulnerability standpoint, a gearbox-

less engine is the ideal configuration and is shown in Figure 96. This arrange-
ment would require experimental development in several areas but could be
accomplished without excessive development risk. The areas requiring verifi-
casion are outlined below:

• Development ano demonstration of radial and axial air bearings

• Development and demonstration of engine integral start tech-

niques

• Coordination of a power turbine interface that is integral with

the airframe gearbox.

203
 SECTION IX

TESTING OF C&A CRITICAL ITEM COMPONENTS

The C&A critical item components identified in Section VIII were reviewed by the
Army technical program manager, and the following five programs were selected
for experimental evaluation: (1) fuel pump inducer (inlet suction tests), (2) high-
speed oil pump (cavitation tests), (3) electronic cooling techniques (performance
tests), (4) power turbine overspeed sensor (performance tests), and (5) starter
overrunning clutch (endurance tests).

The selected C&A critical item components were functionally and endurance
tested, in accordance with approved test plans, to fully evaluate the high-risk
aspects of the component or subsystem. Descriptions and results of the critical
item testing are presented and the results evaluated relative to meeting the per-
formance and endurance goals.

A. VANE ELEMENT OIL PUMP - SUNDSTRAND AVIATION

1. Background

The intent of the test program was to establish state-of-the-art parameters for
an ultimate lubrication and scavenge system capability. State-of-the-art pump
design has been advancing to the point where unboosted inlet pumps can operate
at altitudes up to 20,000 ft without cavitation at speeds to 10,000 rpm. The speed
limitation in an unboosted inlet pump is due to cavitation caused by the reduced
inlet pressure existing at 20,000 ft altitude. Tip speeds should not exceed 33 ft/
sec at this altitude, using MIL-L-23699 oil at 250°?.

The specification requirements ior the pump stipulated a minimum supply flow of
43 lb/min at 70 psid differential pressure, and a minimum scavenge of 44 lb/min
at 20 psid differential pressure. Both of these requirements must be met at
20, 000 ft altitude or 13. 75 in. HgA. To achieve the design goals without cavita-
tion, the vane pump had to be designed so that the vane tip velocity does not ex-
ceed the velocity head conversion of the available 13.75 in. HgA. This maximum
velocity is obtained from Bernoulli's equation:

h = vi
2g
where

h - oil head in ft
V = fluid velocity or vane tip velocity in ft/sec
g = acceleration due to gravity 32.2 ft/sec2

The conversion of 13. 75 in. HgA to feet of oil, h, using a weight of 57.5 lb/ft3 for
250oF MIL-L-23699 oil, is

13.75 in. HgA x 144 in? /ft2

= 16.9 ft of oil available
2. 04 in. HgA/psi x 57. 5 lb/ft3

204
Thus, the maximum allowable tip velocity, V, Is

V =Y/2gh = ^(64.4) (16.9)

V max =33 ft/sec

The test program obtained actual test performance enabling an engineering evalua-
tion of Sundstrand's PBD-1 through-vane pump configuration which would be
utilized as a main engine lubrication system pump to operate at 15,000 rpm and
20,000 ft altitude ambient pressure without cavitation.

A new computerized cam contour was developed by Sundstrand to minimize vane

acceleration loads and incorporated into a family of single-lobe, through-vane
type pumps. This new technology is Incorporated into pump Model 015553-200
lubrication and scavenge pump, which is of a cartridge design and similar to the
one included in the engine design study. Running this existing piece of hardware
at comparable tip speeds demonstrated the flow versus altitude capability on nine
different bore sizes in this family of pumps. An example of this comparison
was to take the Model 015553-200, which utilizes a Sundstrand PBD-4 bore, and
to compare It to the Sundstrand PBD-1 bore, which was Included In the engine de-
sign study. Using the chart shown In Figure 97, which cross references the vane
type radius in inches with available pressure in inches HgA, following the matrix,
the maximum speed to run the pump, so that cavitation will not occur, can be
determined. Example: PBD-4 equals 0.450 vane tip radius and the PBD-1 equals
a vane tip radius of 0.232; thus one can see quite readily that 7,733 rpm on the
PBD-4 is equal to 15,000 rpm on the PBD-1. Running the PBD-4 pump at a speed
range from 6,000 rpm to 10,000 rpm demonstrated flow versus altitude perform-
ance points at high tip speeds (30 to 40 ft/sec).

The PBD-1 bore has a tip speed of 30.3 ft/sec at 15,000 rpm and 40.4 ft/sec at
20,000 rpm; therefore, 15,000 rpm would be the maximum speed the^pump should
be run with an Inlet at 20,000 ft altitude, when pumping MIL-L-23699 oil at 250oF.
An overspeed test was run to demonstrate the pump durability (no cavitation erosion
while operating at 20,000 ft altitude inlet conditions).

2. Test Program

The test arrangements are/shown in Figures 98 and 99.

General calibrations of the PBD-4 pump were run at speeds between 4,000 rpm
and 10,000 rpm; fluid Inlet temperatures of 750F, 1750F, and 250oF; and Inlet pres-
sures between ambient and 24 In. Hg vacuum.

The discharge pressure of the lubricating pump element was maintained at 600 psl
gage, and the discharge pressure of the scavenge pump element was maintained at
30 psi gage.
These data are shown in Figures 100 through 103 and did not meet the theoretical
maximum performance based on tip velocity, indicated on Figure 101. The pump
will have a minimum volumetric efficiency of 85% at all temperatures between
750F and 250oF using MIL-L-23699 at inlet pressures equivalent to 20,000 ft
altitude.
205
 The calibration data show the pumping elements were not able to meet the 20,000-ft
altitude requirement. At the 250oF oil temperature condition, the maximum alti-
tude capability on the larger element, running at design speed, is approximately
9, 000 ft.

REFERENCE TIP RADIUS, in.

PBD -1 0.232
PBD -2 0.343
PBD -3 0.418
PBD -4 0.450 (TEST SIZE)
PBD -5 0.506
PBD -6 0.625
PBD -7 0.682
PBD -8 0.722
PBD -9 0.785
PBD -10 0.892
PBD -11 1.000
32,000

y/~ »BD- 4
Y L* v^
W
*
L^
* y*
^i x* •rl J(A
z
16,000 -^ ^ ^
y* ^
y
is y^ ^ r ^

s r^t
$ i
tä
i
s*
i
y
^f^L
^ ^
<<<
x
^P\^y S
8,000 y* '^> \Tr ^^^ ^
_r> g; < s rl ■^^^T ^
y
S
•<, K' & ^e ^ kk t %^^
**

^^ <&
r^ _,-> s & ^ S^ ^
1r
Ia.
1 "^^ y^y y^
2 4,000
Z ^
^

jjx^, Ys*
s^ 5 y

t
>v^l

2,000 ^v^
4» c- Uli
1
1
11
-J
i
\
1-5 \ ■ »I<
6 8 10 15 30 60 100

INLET PRESSURE, ptia

Figure 97. Maximum Pump Speed vs Minimum

Inlet Pressure for Various Cross
Vane Pump Designs

206
 LUBE

N f h
VACiVACi
i

"L
SCAV

14- PRESS
OR VAC

^^«i^-
0-
^L
£ HE

TEST CONDITIONS:

FLUID: MIL-L-23699
TEMPERATURE, FLUID: ROOM TEMP TO 250oF
SHAFT SPEED: 4000-10,000 rpm
DRIVE: STRAIGHT SPLINE DRIVE
PUMP FIXTURE: TO FIT 015553-200

Figure 98. Oil Pump - Test Schematic

Figure 99. Oil Pump Test Arrangement

207
>

<
>

e 0)
in 3

<

^ •" »" C2
o
uiui/ieS 'MOId
2
>
ii
ß c
eg
I ^
au
>
v a
S
>i i I
IA

1 is
d
o
i-H

h
a

208
 ©
00
I
<
> V
u
* a 3
ß
8.
s0)
H
s

to

H
c
o
—1

5
Vct
U
a
e

E
uiiu/ito 'Moid AuaAnia

209
 " >
<

? £
2
CT

&
E
H

D
H
a
o
a
•5
0)
U
I
ä
PU o a> co N ID m M «M
CM
O
unu/|rt 'Moid AH3An3a

210
e> eo r> (O in

£
s
fc

211
To meet the high-altitude ambient inlet condition, the inlet port of the lubrication
element was redesigned.

Another test series was run to establish the effect of increasing the inlet opening
of the lubrication element, as shown in Figure 104, from its original 0.562 in.
opening to approximately 0.900 in. The performance improvements resulting
from the increased opening are shown in Figures 105 through 108, rnd demon-
strate the ability of the pump to run cavitation free at various speeds and altitude
inlet conditions.

The pump was run for a minimum of 50 hr at a speed of 8,000 rpm, a fluid Inlet
temperature of 250oF, Inlet pressures of 16. 25 through 16.75 In. Hg vacuum, and
discharge pressures of 60 pslg and 30 pslg for the lubricating and scavenge ele-
ments, respectively. The pump performance was unchanged throughout the 50-hr
endurance test, and there was no measurable or visual damage or distress to the
pump as a result of the test. Photographs of the test hardware are shown In Fig-
ures 109 and 110.

With the Inlet pressure at room ambient and the discharge pressures at normal
conditions, ambient air was bled Into the inlet lines of both elements. The pump
was allowed to run 1 hour at the design speed of 7733 rpm. The housing tem-
perature Increased from room ambient to 120oF within 35 mln and continued to
1240F at the end of the hour run.

No damage or distress resulted from the air bleed test. The pump was recalibrated
after the dry run and showed no degradation In performance.
0.900 REVISED
OPENING

— 0.562
ORIGINAL
OPENING

Figure 104. Oil Pump Element Liner Showing

Material Removed From Inlet Opening
212
o
<
>

e I,
o

a
c
o
E
S

§
s
S

T
} 1 ill m
e

1
CD
•o
B

5
^ ^ ** r» ^ «•
<

e
o
uiui/|e« 'AÄ01J uitu/irf 'MOld
•S
«
8 ? ü
c
a
I
ft M
e *
E
I in

213
f» <M oi eo r> ID in
ujiu/irt '«Old AU3AI1ia

214
 's
•E 2
2

H
I

0)
H
§

>
ca
U
I
M M o» oo iv. «c m M W
o
UIUJ/IIS 'Moid AaaAnaa

2
&
•p^
b

215
 <
>
Jg s O
in
N
3)
c
3
2

es

3
E

a»
H
eo
2
"5
rt
Ü
a
6
M N o> eo r« <e 10 »»■ « CM
s
»—I

o
uiui/irt 'MOld AU3AI13a
S
H

2I6
 - c^ojhyio
i

Figure 109. Oil Pump Components Before 50-hr

of Endurance Testing at 250oF

c)0§

Figure 110. Oil Pump Components After 50-hr

of Endurance Testing at 250oF

217
3. Test Results
The Sundstrand vane element oil pump ran satisfactorily throughout the test series.
The altitude performance improved with increasing oil temperature, but at design
speed and with 250oF oil, the lubrication element output flow began to deteriorate
at an inlet condition of approximately 18,000 ft altitude (Figure 108). The lubri-
cation and smaller, scavenge element performance was for the most part, com-
parable throughout the test. However, as indicated in Figure 108, the scavenge
element does appear to meet the 20,000 ft altitude inlet pressure condition with
negligible reduction in output flow. Performance was unchanged following a 50-hr
endurance test at elevated oil temperature. Running the pump dry, simulating
loss of system oil pressure, did not result in excessive housing temperature or
damage to the pump.

4. Recommendations for Additional Testing

A scavenge element normally does not operate with a 100% fluid inlet, as was
simulated in this test. The scavenge element is usually sized for two to four
times the scavenged oil flow to pump an air/oil mixture resulting from air leak-
age into the bearing compartment. Additional tests are recommended to investi-
gate and define capability of a high-speed vane element to pump a vapor/liquid
mixture against various backpressures.

B. STARTER OVERRUNNING CLUTCH - AIRESEARCH

1. Background

The air turbine starter design configured for the rear-drive engine achieved
minimum weight and maximum performance efficiency by locating the over-
running clutch in line with the engine rotor. This resulted in a clutch over-
running speed of 65,000 rpm that is considerably in excess of previously
demonstrated capability.

High-speed sprag clutches, in the driving mode, generally fail from fatigue or
sprag rollover. Overrunning at high speeds can lead to excessive clutch and
race wear, which ultimately results in rollover or inability to drive. The sprag
type clutch also necessitates incorporation of a decoupler mechanism to prevent
overspeed of the starter turbine wheel in the event of a clutch or bearing failure.
At the design speed of 65,000 rpm, it is important that the decoupler provide
complete disengagement of the starter from the engine.

Two basic sprag clutch designs are currently in use, and these two designs can
be arranged in two ways, that is, the inner race of the clutch can overrun or
the outer race can overrun. Borg-Warner manufactures one of the basic de-
signs. This design is a doubl&-cage, full-phasing sprag clutch consisting of two
concentric cages with a ribbon spring between. The spring provides a light
energizing torque on the sprag to ensure positive engagement. Formsprag
offers the second basic clutch design. This configuration is a single-cage full-
complement design. The sprags contact each other to provide phasing and roll-
over protection. A garter spring provides engaging force.

218
The clutch test program was comprised of two phases. Phase I coasisted of an
analytical and experimental investigation of the design concepts to identify the
most promising configuration. Testing enabled conclusions to be drawn with
regard to feasibility, problem area definition and design changes. The Phase II
effort demonstrated a measure of the operating life of the selected clutch con-
figuration and decoupler. A 1000-hr endurance test was performed to establish
a measure of the life expectancy, and decoupler tests were run to define dis-
engagement torque levels.

2. Test Program - Phase I

Both Formsprag and Borg-Warner evaluated the clutch requirements discussed

in Section VIII. Borg-Warner indicated that they favored the inner race over-
running approach due to the very high disengaging speed and close tolerance
required to ensure clutch release between starter cutout speed and engine idle.
In addition, Borg-Warner had an existing inner race overrunning design which
they felt could be modified to meet the requirements.

Formsprag indicated that they did not have a clutch capable of meeting this
design requirement, and presented a proposal for the development of an outer
race overrunning configuration. After reviewing both clutch vendors' recom-
mendations, it was the opinion of AiResearch's design and engineering depart-
ments that the outer race overrunning approach proposed by Formsprag involved
greater technical risk than the inner race overrunning approach proposed by
Borg-Warner. The ability to accurately predict the lift-off speed of the outer
race overrunning configuration was the primary area of concern. In addition,
the development program quoted for the outer race overrunning configuration
was not within the scope of available time and cost.

Rather than running a 40-hr screening test on both clutch designs of each man-
ufacturer, a 160-hr test would be run on the Borg-Warner inner race overrunning
configuration. A test unit was designed around this clutch providing bearings
support, lubrication system, oil dams, output shaft, drag torque measuring in-
strumentation and support housing (Figure 111).

The anticipated problem areas in the test unit were identified as the following:

1. Wear on the sprag clutch or the clutch inner race

2. Overrunning bearing life
3. Excessive overrunning drag torque
4. Lubrication flow and location.

The following series of tests was outlined to demonstrate the capability of the
clutch test rig in achieving satisfactory performance with respect to the sus-
pected problem areas.

219
 K
■i->
CO
0)
H
j=
ü

a
•—i
u
d
c
c
2
I
o

CO

220
a. Segment I

A 40-hr test would be conducted at 65,000 rpm with predetermined values for
oil jet size and pressure and for oil dam clearances. Overrunning drag torque
would be monitored during the test; in the event of excessive drag torque, the
oil flow to the clutch could be reduced. Upon completion of the 40-hr test, the
unit would be disassembled and inspected. Critical dimensions would be mea-
sured to determine the extent of clutch or inner race wear. The overrunning
bearings would be inspected for signs of deterioration. Should all parameters
prove satisfactory, the test unit would be rebuilt and Segment II conducted.
Should there be indications of distress in any of the components of the test unit,
changes would be made, aimed at correcting the problem area, such as in-
creasing oil flow or modification of oil dam clearances to increase oil supplies
to the bearing. Following any necessary changes, the test unit would be rebuilt
and subjected to Segment II.

b. Segment II

Conduct a 40-hr test similar to Segment I. Following completion of Segment II,

the teardown and inspection procedure of Segment I would be performed. The
test unit would again be evaluated and any necessary changes made. Upon com-
pletion of Segment II, the test unit would be subjected to Segments III and FV.

c. Segments III and IV

Segments III and IV would consist of separate 40-hr test similar to Segments I
and II followed by teardown inspections. At the completion of Segment IV, a
total of 160 hr of testing would be accomplished on the test unit.
Due to the high-speed operation required for this test, a special gearbox was
designed to allow the use of conventional drive motors as the power source. A
speed increasing gearbox with a 9.13:1 gear ratio was designed and tested at the
required speed of 65,000 rpm. The gearbox presented no problems. However,
upon attaching the clutch test rig and associated coupling shafts, critical speed
problems were encountered above 45,000 rpm. An isolator pillow block was
designed to mount between the 9.13:1 gearbox and the clutch test rig. The im-
proved support offered by this isolator eliminated the critical speed problems.
Figure 112 shows a schematic of the test stand including the clutch test rig.
Figure 113 shows a photograph of the actual test stand and clutch test rig.

Figure 111 shows an assembly drawing of the clutch tested in Phase I. The
outer race (Item 3) was designed to mount in antifriction roller bearings to allow
accurate determination of the overrunning drag torque during testing. A strain
gage was mounted to the bearing preload nut (Item 10) to provide continuous drag
torque data. As previously noted, all Phase I testing was accomplished using
a Borg-Warner inner race overrunning sprag clutch (Item 19). Oil dams (Item 6)
were provided on either side of the clutch to control oil flow from the clutch and
also to restrict the oil flow to the overrunning bearings (Item 16). The oil supply
was introduced into the clutch by means of an oil jet (Item 11) which direcldd the
oil into the center of the clutch inner race (Item 13). The oil was then discharged,
using the centrifugal force created by the inner race, through three equally
spaced holes located directly under the sprag clutch. Holes placed in the outer
race then allowed the oil to drain to the main support housing (Item 1). Four oil

221
return tubes were installed in this housing to monitor oil flowrates from isolated
areas in the clutch. Tube No. 4 measured oil flow through the aft overrunning
bearing. Tubes No. 3 and No. 2 measured oil flow through the clutch and the two
oil dams. Tube No. 1 measured oil flow through the forward overrunning bear-
ing. The oil flow data were used to evaluate the performance of the oil dams.
Thermocouples were installed in Tubes Nos. 3 and 2 to monitor outlet oil temper-
atures and in the oil supply jet to monitor inlet oil temperature. The test rig
was mounted to the drive assembly and power transmitter to the clutch through a
quill shaft (Item 2). All testing was conducted using the test fixture described.

3. Testing

Initial testing of the test rig was designed to determine desirable oil flows and tem-
perature levels and to calibrate the drag torque strain gage, A 0.040-ln,-
diameter orifice was installed in the oil supply tube, and oil flows and temper-
atures were monitored at pressures from 20 psig to 40 psig. Inlet oil pressure
of 30 psig provided the most desirable results, and all further testing was accom-
plished using the 0.040-in.-diameter orifice and 30 ±1 psie inlet pressure.
Overrunning drag torque was found to be very low, and the strain gage was cali-
brated to record 0-10 Ib-in. This level covered all torque levels which could be
expected from normal operation of the clutch. The inlet oil temperature was not
controlled for any of the Phase I testing but was allowed to seek its own level.

Following the completion of these pretest evaluations, the test outlined for
Segment I was initiated. The test was terminated after 18 hr of continuous
operation at 65,000 rpm due to a bearing failure in the isolator pillow block. A
mismatch in the pilot diameters of the isolator and clutch housing was found that
led to shaft misalignment and undue load on the clutch bearings. The clutch
test rig damage was repaired and the isolator pilot diameter corrected and the
bearing replaced. The test was restarted at the 18-hr point of termination. At
20 hr of test time, another failure of an isolator pillow block bearing occurred,
resulting in secondary damage to the clutch. It was determined that an isolator
bearing of higher quality was needed to allow continuous operation at 65,000 rpm.
The isolator bearings were replaced with one of a better design. The clutch test
unit damage was repaired and returned for further testing. It should be noted
that the strain gage indicator was disconnected after approximately 19 hr of test.
The recorded torque had been consistently low, and there was no need for further
data. A typical drag torque at 9 hr was 0.31 Ib-in. At this time Segment II was
initiated. The test rig successfully completed 40 hr of continuous overrunning
at 65,000 rpm. The test was terminated and a teardown inspection performed.
Figures 114, 115, and 116 show the test components. As can be seen, no sign
of wear is present on the clutch or inner race. The overrunning bearings were
smooth, with no sign of distress. Some wear was evident on the quill shaft spline
teeth and mating clutch shaft spline. This was expected, however, since no oil
was supplied to these splines in Phase I testing. In view of the excellent condi-
tion of the test hardware, the unit was reassembled with no hardware or design
changes. At this time a total of 60 hr of testing had been conducted. This left
100 hr of total test time to be performed.

222
 OIL JETS

DRIVE MOTOR
(VARIABLE SPEED)

Figure 112. Overrunning Clutch Test Stand

Schematic

Figure 113. Clutch Endurance Test Rig

223
Figure 114. Clutch Test Rig After 40-hr of
Overrunning at 65,000 rpm

Figure 115. Quill Shaft and Spline After 40-hr

Overrunning at 65,000 rpm

224
 Figure llfi. Clutch Components After 40-hr
Overrunning at G5,000 rpm

It was decided, based on the results of the successful 40-hr endurance run, to
attempt the remaining 100 hr without a teardown inspection. The drag torque
strain gage was reconnected for this test. After 17 hr 35 min of continuous
test time, a slight increase in vibration was noticed. The test was interrupted
to check the test rig. It was found that the nut (Item 8 of Figure 111) was loose,
allowing the clutch outer race to vibrate. The nut was tightened and secured
and testing resumed. The strain gage was disconnected at this time since the
torque was again very low and consistent. At 49 hr the nut (Item 10) fatigued
and broke off at a spot just above the oil jet (Item 28). The broken section fell
forward and struck the separator of the overrunning bearing (Item 16).

Testing was resumed and continued to 50 hr, when unusually high bearing tem-
peratures were noted. The test was again interrupted, and a small crack was
found in the bearing separator which had been struck by the broken nut. No
other damage or distress was noted. The bearing was replaced and testing con-
tinued. At 97 hr 20 min, the test was terminated due to the failure of the over-
running bearing next to the quill shaft. Figures 117 and 118 are photographs
of the test hardware following the 97-hr run. It should also be noted that this
same set of hardware had completed 40 hr of overrunning prior to the 97 hr,
for a total of 137 hr. The significant item in the photographs is the condition
of the clutch inner race 180 deg away from the damage which resulted when
the bearing failed. No measurable wear could be found in this area, indi-
cating that the clutch was performing as expected prior to the bearing failure.

225
 o O
COO

Figure 117. Clutch Components After 137.5-hr

of Overrunning at 65,000 rpm
(Overrunning Bearing Failure)

^(O»
"■'■A
J
oo
)0 — Cr
-*
Figure 118. Clutch Test Rig After 137. 5--hr
Overrunning at 65,000 rpm
(Overrunning Bearing Failure)

226
Further evaluation of the bearing failure revealed a potential design flaw which
could have contributed to the bearing failure. Referring to Figure 111, a spring
washer (Item 14) is employed to preload the overrunning bearings. It was noted
that any reverse thrust from the quill shaft (Item 2) would tend to unload the
bearing nearest this shaft. Further, a built-in thrust was inadvertently provided
by the spring (Item 17) which would be added to any thrust generated by spline
misalignment. The Phase 11 design for this test rig relocated the spring washer
(Item 14) between Item 5 and Item 20. The same bearing preload will result,
but reverse thrust cannot unload either bearing.
No further testing was conducted on the Phase I design. L was felt that no new
information would have been gained by running of the additional 2-1/2 hr needed
to complete the 100-hr test. Total accumulated test time was 157.5 hr.

4. Test. Results

The results of the Phase I testing indicated a high degree of confidence in de-
signing a reliable clutch assembly for operation at 65,000 rpm. The oil flow
requirements were well within available limits. The overrunning drag torque
would present no special problems in an actual engine application, and the oil
temperature rise levels are not prohibitive. For a clutch oil flow of 1.25 lb/min,
a nominal oil temperature rise of 10CF was observed.

The two basic problems which were identified during Phase I testing appeared to
be proper overrunning bearing preload and quill shaft unbalance and misalignment.
The bearing preload problem was solved by the relocation of the preload spring
washer. The quill shaft was eliminated in the Phase 11 design and replaced with
a one-piece clutch inner race-output shaft combination.

5. Test Program - Phase II

This section summarizes the results of Phase II testing of the overrunning clutch.
The purpose of this test was to incorporate the knowledge gained during Phase I
testing into a design capable of 1000 hr of overrunning clutch operation at 65,000
rpm. The test stand used for this test was identical to that utilized for Phase I
testing. No changes in test stand design were necessary for Phase II. However,
the two design improvements discussed above were incorporated into the endur-
ance clutch rig as described following.

a. Clutch Endurance Test Rig Design

The Phase II test program had two main goals: the first to achieve 1000 hr of
65,000 rpm clutch operation and the second to incorporate a one-piece output
shaft with a spline drive and decoupler into the design. The Phase 11 test rig
was modified to incorporate an output shaft and decoupler mechanism which
simulates an actual engine application. Figure 119 is the assembly drawing for
the Phare II clutch test rig. It can be seen that the test rig is identical to that
tested during Phase 1, except for the addition of the decoupler mechanism and
the longer, one-piece, splined, output shaft. It should be noted that the spring
washer (Item 7) was located between the retaining ring (Item 6) and the over-
running bearing (Item 4) when the test unit was assembled This will prevent
unloading of the overrunning bearings by the thrust resulting from the decoupler
control spring (Item 30). A description of the oil feed and scavenge plumbing
for the test rig was contained in the Phase I test description.
227
 T
9

I
f
I

?3
u 69 ill1
ll!
:::JTI€
Si*
fa ^i^: ash
"-a in«

ifi
im:TZeif or j
JV
ill
C4 r|o (jusiln- sj ^14
Sal;

£
CO
0)
H

C
re
4)
Ü
e
re
u
c
W
x:
Ü

I
B
0)

ft

228
b. Clutch Endurance Testing

The data recorded during the test were identical to those for Phase I testing,
except drag torque. Due to the low drag torquj recorded during Phase I testing,
no drag torque data were recorded during Phase 11 testing. The inlet oil tempera-
ture to the clutch was maintained at approximately 110oF and was allowed to
seek its own level for the first 20U hr. At this time a teardown inspection was
performed. Figure 120 is a photograph of the clutch components following tear-
down. As can be seen, there is no visible sign of distress on any of the com-
ponents. The test unit was reassembled and returned for continued testing. Inlet
oil temperatures to the clutch for the remainder of the test were maintained at
approximately 200oF. It should be noted that some early difficulties were en-
countered in maintaining this temperature. However, continued efforts at proper
insulation resulted in the desired inlet oil temperatures.

After 686 hr, the test was stopped for an Inspection due to very erratic oil flow
from oil tubes No. 3 and No. 4. A teardown inspection was initiated, revealing
that the oil plug at the end of the clutch shaft (Item 1) under the decoupler had
become dislodged. This plug had been tack welded in place and it is unknown
what caused it to come out. When this plug is not in the system, the oil flowing
down the center of the clutch shaft bypasses the clutch and overrunning bearings
and is expelled through the decoupler end of the test rig.

This situation most drastically affects the bearing nearest to the oil jet and
effectively causes it to run without sufficient lubrication. The decoupler mech-
anism had actuated, and bearing analysis indicated that the forward bearing,
closest to the oil jet, was distressed due to the lack of oil. The increased bearing
drag which resulted, actuated the decoupler.

Analysis of the test data showed that oil flow in oil tube No. 3, which is providing
flow through the clutch, dropped almost 50% following the decoupl . actuation,
and oil flow in oil tube No. 4 increased by a factor of 100. This would be expected
since the inner race does not rotate following decoupler actuation. Without inner
race rotation, the radial holes normally expelling oil to the clutch no longer
function, causing flow from the jet to back up and run into oil tube No. 4.

A photograph of the clutch components is shown in Figure 121. The output shaft
was broken during disassembly when the decoupler was being removed. It is
suspected that the shaft was weakened during the extended period of operation in
the decoupled condition. The dislodged plug can be seen In the foreground between
the shaft and the spline. The bearing nearest the spline, in the photograph located
nearest the oil jet, exhibited high drag. It can be seen that the inner race does
not appear damaged and detailed inspection revealed no measurable wear. The
clutch and aft bearing were also undamaged.

This incident demonstrated that a failure of any overrun component In the starter
that causes an additional drag torque on the starter turbine will cause the de-
coupler to actuate. Further, after actuating, continued operation of the engine
will not cause any distress upon those critical components within the engine that
connect to the starter.

229
 *r
.*"

Figure 120. Chilch Components After 200-hr of

Overrunning at 65, 000 rpm

Figure 121. Inner Race and Bearings After 686-hr

of Overrunning at 65,000 rpm
(With Oil Plug Failure)

230
The test unit was rebuilt using the original clutch and aft bearing. The forward
bearing, inner race shaft, and decoupler were replaced. Testing was resumed
at 686 hr. The remaining 314 hr were completed without incident. A teardown
inspection was conducted following completion of the test, with no significant
signs of wear or damage to the clutch, inner race, or output shaft spline. Fig-
ures 122 through 124 show the clutch components after 1000 hr of testing.

c. Decoupler Test

Following the completion of the endurance test, the unit was rebuilt with the clutch
installed so that it would drive in the normal overrunning mode, thus allowing a
determination of the torque required to actuate the decoupler. The test-unit was
mounted to the test stand gearbox. With the output shaft of the gearbox restrained,
a torque was applied to the nut (Item 8 of Figure 119) until the decoupler actuated.
This procedure was repeated a total of ten times. The recorded decoupler torque
was between 28 and 30 in.-lb on each occasion. The decoupler had been designed
to actuate at a level of 30 in.-lb maximum. It should be noted that the level of
decoupler actuation torque can be varied by changing the spring (Item 30) to re-
sult in higher or lower axial load required on the jaws of the decoupler.

Figure 1 ;2. Clutch Test Rig After 1,000-hr cf

Overrunning at 65, 000 rpm

231
 INNER RACE USED
TO COMPLETE 1000 hr
ENDURANCE

686 hr OF
OVERRUNNING
OPERATION
AT 65,000 rpm

Figure 123. Inner Race at Conclusion of Endurance Test

Figure 124. Clutch at Conclusion of 1,000-hr Over-

running at 65,000 rpm Showing Clutch
Sprag

232
d. Test Results

The results of Phase II testing have indicated that the design and operation of a
GS.OOO-rpm overrunning sprag clutch is feasible. The drag torque and resulting
oil system temperature rise were well within allowable limits, and the endurance
wear characteristics of the critical clutch elements were acceptable. Further,
the incorporation of a decoupler mechanism has also proved to be a workable
solution to the emergency disengage requirement. The spline utilized on the
output shaft showed no signs of excessive wear and should present no special
problem in an engine applicati n. Every effort was made in this program to
simulate an actual engine installation. It is felt that the results of the test pro-
gram are directly applicable to an integrated engine-starter design and will
result in a high degree of confidence in a production design.

C. ELECTRONIC CONTROL COMPONENTS TO HOUSING HEAT

TRANSFER - CHANDLER EVANS INC.

I. Background

Through two phases of Army-sponsored programs, development work has been

conducted to formulate concepts to provide adequate cooling of the electronic
computer.

A brief summary of previous milestones is as follows:

1. USAAMRDL Phase I - Electronic control hard-mounted to the

fluid controller utilizing fuel at a maximum temperature = 1350F
for cooling and a fluorinert fluid as a heat transfer mechanism
from the components to the cold plate. The maximum ambient
temperature was 250oF.

This arrangement proved to be successful, limiting the maximum

component surface temperature to 190oF, well below the
maximum allowed of 250oF. However, the inclusion of the
fluorinert fluid added cost and weight and proved to be very
difficult to use.
2. USAAMRDL Phase II - Electronic control utilizing fuel at a
maximum temperature = 1350F for cooling and conductive heat
transfer via the printed circuit boards (PCB's) and intercon-
necting cables from the components to the heat sink.

With an ambient temperature of 250oF, the maximum component

surface temperature was 1950F; with an ambient temperature
of 180oF and no fuel cooling, the maximum component case
temperature was 190oF.
The basic requirements for this critical Item test program were to provide an
adequate thermal path from every electronic component to the computer housing
heat sink, to ensure adequate heat transfer, to minimize hot spots, and thereby
to enhance reliability. The heat sink in this cade is the computer housing mounting
surface that interfaces with the engine inlet particle separator and is thereby
cooled by the compressor inlet air.

233
The proposed components to housing heat transfer concept is based on utilizing
multilayer printed circuit boards for interconnecting the electronic components.
The outer layer of the boards is an additional copper layer termed a thermal
plane (Figure 125). This layer of copper covers the entire PCB except for a
minimal spacing of 0.015 in. between it and the component pads. This exposed
plane is connected at two adjacent corners to the computer housing by a flexible
copper cable, an arrangement which provides a virtual thermal short circuit
between the computer housing and the thermal plane on each board. There are
two thermal paths between the heat-producing electronic components and the
thermal plane. One of the paths is direct from the component housings which
are in contact with the thermal plane, and the second path is from the component
leads and interconnecting conductors through the PCR epoxy-glass material to
the thermal plane. For the power driver transistors, a mounting clip is pro-
vided that electrically isolates the transistor, yet provides a reasonable thermal
path (approximately 30oF/watt). The clip is fastened to the PCB in direct con-
tact with the thermal plane.

The thermal plane configuration dictated a departure from previously developed

concepts; thus a test program was conducted to determine the heat transfer
efficiency of the new design.

The multilayer PCB, which contains a thin sheet of tinned copper on the com-
ponent mounting surface, and associated mounting hardware are shown in Fig-
ure 126. The board was supported in the electronic control housing on silicone
vibration isolators whose thermal conductivity is negligible compared with the
flexible copper braids which coupled the thermal plane to the housing. Power
transistors mounted in beryllium oxide, electrically but not thermally insulated,
TO-5 clips were used as the primary heat source. Four additional resistors
held in steel clips simulated the power dissipated by components normally
mounted on the remaining area of the board. This method of mechanically
fastening the electrical components to the board resulted in a good thermal path
from component to the thermal plane which could be easily repeated from one
configuration to the other.
The sealed computer module containing the PCB, associated electronic com-
ponents, and temperature measuring thermistors was enclosed in an environ-
mental chamber and dc power was applied. A schematic of the test setup is
given in Figure 127. The power dissipated in each component was individually
measured by reading the voltage supplied and current drawn by the device. A
variable load resistor in each transistor control circuit was tuned such that
each transistor dissipated the same power within a 1% variance. Thus, the
heat input was accurately maintained, and any measured temperature differences
could therefore be attributed to the physical configuration under test.

2. Test Program

A printed circuit board of the type to be tested on an advanced technology turbo-

shaft engine control, containing the highest power dissipating components, the
stepper motor drive transistors, was tested in two phases. In the first series
of tests, various thermal plane thicknesses were evaluated by mounting the PCB
into the sealed electronic computer housing, dissipating power through a few

234
I.

components mounted on the board, and measuring the resulting temperature

gradients along the board and component operating temperatures. The second
phase of testing consisted of measuring the operating temperature of selected
high-power dissipating components in a completely assembled prototype elec-
tronic control which was operating at a typical engine transient condition. Both
test phases were conducted at various ambient temperatures, not to exceed a
maximum ambient temperature of 80°C as specified In the Appendbc.

TYPICAL POW"=R
TRANSISTOR
MOUNTING CLIP

TYPICAL OUTPUT
RESISTOR PIN STRIP

CABLE - TO
CHASSIS

0.004 MIN TYP

POWER

SIGNAL GROUND

Figure 125, Thermal Plane Heat Transfer Concept

235
 POWER TRANSISTOR RESISTORS
BaO CLIP (Q1S THRU 022) (R1 THRU R4)-
TINNED COPPER LAYER

o
COPPER BRAID
\ (2 PLACES
1
\ SHOWN)

X1

SILICONE
VIBRATION
ISOLATOR
(4 CORNERS
OF BOARD)

■REVERSE, GLASS EPOXY SIDE OF PCS

Figure 126. Thermal Plane Printed Circuit Board

POWER SUPPLY

//////////////A/ o o

mmoo
I POTENTIOMETER
LOAD
® ® ®®
®® ®®
T
^ AMBIENT
^7777777777 000000000
TRANSISTOR
CIRCUITRY
ELECTRONIC THERMISTOR
BANK
COMPUTER

ENVIRONMENTAL
CHAMBER

Figure 127. System Test Schematic

236
a. Phase I - Thermal Plane Evaluation

The first series of tests was conducted with only two power transistors bolted
to the PCB through heat sinks at the Q15 and Q16 locations (Figure 128). The
two transistors were operated to dissipate a considerable amount of power, 4. 3
watts combined, to facilitate the measurement of temperature gradients through-
out the board. Thermistors were located in positions 1, 10, 11, 12, and 13 as
shown in Figure 128. The most significant effect was a 20oF decrease in the
operating temperature of the components and of the board temperature in the
vicinity of the heat source with a 0.009-in. thermal plane.
Ambient Thermal Plane Copper Component Temperature - 0F
Temperature, Thickness, Braids,
0 PCB PCB11
F in. No. Trans, at Q15 10

75 0.002 None 231 165 123

75 0.009 None 211 145 119

Thus, the thermal plane dissipates hot spots along the board by conducting
localized heat inputs into a larger board area.

Another set of tests was conducted wherein all eight of the stepper motor drive
transistors (i.e., U2T101 Darlington Power Transistors) were Installed in their
assigned locations, Q15 through Q22, and with 0.03-in#2 copper braids connecting
the computer housing to the thermal plane. Each transistor was set to dissipate
3/4 watt, a power level approximately 50% greater than the individually highest
dissipating operating condition which occurs when one continuously energized
phase is required to hold the stepper motor in a fixed position. This power
level yielded sufficient thermal gradients (10o-20oF) to negate much of the
±0.50F accuracy of the thermistor and allow a good comparison of the various
thermal plane configurations under test. Four resistors dissipating an additional
total of 1 watt were spaced throughout the remainder of the board, and thermistors
1 through 8 were positioned as shown in Figure 128.

The test results are plotted In Figure 129. These data show that a nearly 30%
decrease In temperature rise of the power transistors (TQ-Tamb) can be achieved
by utilizing a 0.009-in. thermal plane In conjunction with a 0.03-in.2 copper
braid fastened to the housing. The components located close to the braid (Q21
is 0.5 in. away) derive approximately 7% more benefit in the reduction of temper-
ature rise than those farthest removed from the braid (Q15 Is 2 In. away). Also,
the 0.009-ln. thermal plane yields a 10% average decrease in temperature rise
over the 0.002-ln. plane.

b. Phase 11 - Prototype Control Component Temperatures

The completely assembled prototype electronic computer (Figure 130) employs

0.002-ln. -thick thermal planes and 0.03-ln,2 copper braid connections between
the plane and the housing. The sealed unit with thermistors epoxied to the
Q15 and Q21 stepper motor drive transistors was placed in an environmental
chamber and operated at a typical engine transient condition with both stepper
motors slowing. The operating temperature of the two transistors was monitored

237
to determine the effect of the thermal plane with and without copper braids
attached to the housing. The results are tabulated following.
Percent Decrease in
Component Temperature, Temperature Rise,
r X
Control Ambient Q amb
OF Braids Q15 Q21 Q15 Q21
^mb»

70 No 116 106
70 Yes 111 100 11 17
155 No 191 185
155 Yes 187 181 11 13
180 Yes 211 205

COPPER BRAIDS

Figure 128. Printed Circuit Board Layout

238
 250 11 1 [
>\MBIEN1 TEMPEI MATURE - 175°!

240 Vv
\ 2 COPPER BRAIDS
0.03 in.2 EACH
^
230
— Q15

^ Q21

tu
cc
220
5

160

ISO

140

130
2 3 4 5 6 7 8

THERMAL PLANE THICKNESS - in. x ID3

Figure 129. Component Temperatures vs Thermal

Thickness

239
 STEPPER MOTOR
Iflbff DRIVE PCB-

POWER
SUPPLY
PCB

0.03 in.2 COPPER

BRAIDS (2 PLACES)

Figure 130. Electronic Computer With Thermal

Plane PCB Implementation

The improvement in temperature rise of both power transistors with a braid con-
necting the thermal plane to the housing was better than that predicted by Phase I
testing. The transistor located close to the braid, Q21, followed the previous
results more closely (i.e., 13-17% decrease in temperature rise vs 12%); how-
ever, Q15, which is located 2 in. from the nearest braid, yielded a IV'i decrease
in temperature rise compared to the 4,/( measured in Phase I. It was found that
with the board fully assembled and operating, components farther away from the
braid were affected to a greater degree by components in their vicinity, whereas
those close to a braid approach the condition of being thermally shorted to the
housing. Measurement of the thermal plane near Q15 revealed a plane temper-
ature 50F hotter than the transistor itself, thus indicating heat flow from the rest
of the board toward Q15. With the thermal braids connected, the plane temper-
ature dropped by 50F, thereby accounting for the total improvemenl in the oper-
ating temperature of the transistor mounted at Q15.
These results indicate that when heat is introduced throughout the board rather
than at a few isolated locations, as was the case in Phase I testing, more of the
thermal plane is utilized in the conduction path to the braids. Consequently, the
thermal gradients across the board are less, the conductive heat transfer
through the braids into the housing is more efficient, and component operating
temperatures are lower.

The 180oF ambient test demonstrated that the drive transistors remain well
within their operating temperature ceiling of 250oF at the worst-case ambient
condition expected to be encountered in an advanced technology helicopter
application. These component temperatures are approximately 20 F higher
than those encountered on the Phase II unit due to an increase in control com-
plexity and associated power dissipation.

240
3. Test Results

The results of the Phase I testing demonstrated that the thermal plane helps to
spread the heat and thereby reduces the temperature rise of hot spots. Test
data show a 13% Improvement even without a braid connection between the
thermal plane and housing. An additional 15% reduction in temperature rise was
achieved with copper braids.

The Phase II testing of a fully assembled and operating system has generally
confirmed the results of the PCB thermal plane evaluation testing. Figure 131,
which shows the percentage improvement in component temperature rise, is
also valid with a minor modification for a fully assembled control. Components
which are farther removed from a braid (i. e., Q15 transistor) may approach
the results of a close proximity device such as Q21 if a considerable portion of
their temperature rise is attributable to other components rather than self-
heating. A nearly 30% decrease in temperature rise was achieved with a 0. 009 in.
thermal plane in conjunction with 0.03 in,2 copper braids fastened to the housing.
To realize any significant additional reduction in component operating temper-
atures, a heat sink will be required.

4. Recommendations for Additional Testing

The tests conducted assumed a housing temperature of 175CF, Further reductions

In component operating temperatures are possible by reducing the housing tem-
perature by heat sink to the engine Inlet air which Is 1350F maximum. Additional
tests are In order to determine the effectiveness of this heat sink approach and
the overall Impact on component operating temperatures.

D. FUEL PUMP INDUCER - CHANDLER EVANS INC.

1. Backgi-»and

The pump specification for advanced technology helicopter engines requires

operation at Inlet conditions of 1. 0 V/L and one psl NPSP. Current technology
In fuel systems Is for operation at 0.45 V/L and five psi NPSP. While tests
have demonstrated the ability of fuel boost pumps to operate at conditions higher
than 0.45 V/L, sustained operation at 1. 0 V/L represents a significant advance
of the state of the art. The additional requirements of high speed (65,000 rpm)
and low flow (500 Ib/hr) required for this pump Indicated that a special program
was required to develop Inducers to meet these pump Inlet requirements. The
objectives of the test program were, therefore, to:
1. Develop an Inducer capable of charging the high-speed main
stage pump through the range of Inlet temperatures and pres-
sures required.

2. Demonstrate V/L = 1 and NPSP = 1 psl performance on a bar

stock Inducer test fixture.

241
 4 6 8
THERMAL PLANE THICKNESS - in. x 103

Figure 131. Percentage Reduction in Temperature

Rise vs Thermal Plane Thickness

242
A bar stock test fixture (Figures 132 and 133) designed to adapt to several im-
peller configurations was fabricated to evaluate the following:

1. A conical inducer
2. A jet inducer
3. An axial inducer
4. Combinations of the above.

Performance requirements and acceptance limits for performance capabilities

were based on the main stage inlet requirements specified in Section VIII, that
is, 500 Ib/hr fuel flow at a pressure rise of 80 psi. The test evaluation of the
hardware used an Ikor V/L meter which provides an instantaneous reading of
V/L in the inlet line and does not rely on the lengthy calculations normally
involved in the test procedure. The test setup conformed as nearly as possible
with the ARP 492 V/L test procedure recommendations. In addition, each
impeller, inducer, or combination was:

1. Calibrated to determine head-flow, ^T, BEP, and NPSP

capabilities

2. Tested to determine V/L limitation on 1350F JP-4 at sea

level, 10,000 and 20,000 ft tank altitudes at three flow
values.

Minor reworks to the basic designs were attempted to improve performance.

2. Test Program

The centrifugal boost impeller designed for use in an USAAMRDL pump to be

tested onan advanced technology gas generator was used as a base for the tests
performed. Design parameters for this impeller are summarized following.
Speed 65,000 rpm
Q 500 Ib/hr
^P 80 psid
Ng 1,100
OD 0.583 in.
Inlet Diameter 0. 234 In.
Blade Inlet Tip Angle 14. 8 deg
Type Conical
V/L Capability V/L = 0.45
Q = 225 Ib/hr
IP = 50 psid

The impeller was calibrated at a sea-level tank condition, using the test arrange-
ment shown in Figure 134. The head/flow calibration data are illustrated in
Figure 135. The V/L evaluation at various tank conditions is shown in Fig-
ures 136 through 138 and resulted in minimal V/L capability. The USAAMRDL
'mpeller was modified as illustrated in Figure 139, which resulted in a cutback
Inlet. This configuration has a straight vane shape which is the standard method
of treatment for conical impellers. The maximum speed head/flow calibration,
with a sea-level tank, for the revised impeller is shown in Figure 140. As indi-
cated in Figures 136 through 138, a V/L improvement does result with the cut-
back inlet, with a corresponding loss in available head rise.
243
 BEARING A RETURN
SUPPLY r** PORT

DRIVE NOZZLE
SPLINE INLET

BEARING AND ■AXIAL INDUCER

SEAL RETURN JET
•BOOST IMPELLER
INDUCER"

DISCH PORT

A-A"

Figure 132. Inducer Test Rig Configuration

IT INDUCER
NOZZLE
JET INDUCER
HOUSING

AXIAL INDUCER
CONICAL INDUCER

' INDUCER JOURNAL

DRIVE SHAFT

•SHAFT SEAL

Figure 133. Inducer Pump Test Fixture

244
 ®
HEAT
-D-
FLOW
REGULATOR
(2)_ VALVE
EXCHANGER METER
VACUUM

®® =a®
011

:
UEL TANK

LEGEND
1. INLET PRESSURE 7. AMBIENT TEMPERATURE
2. DISCHARGE PRESSURE 8. AMBIENT PRESSURE
3. rpm 9. FUEL TANK PRESSURE
4. FUEL FLOW 10. V/L VALVE
5. FUEL INLET TEMPERATURE 11. V/L METER
6. FUEL DISCHARGE TEMPERATURE

Figure 134. Inducer Test Schematic

140

120

100

I
a.

0 100 200 300 400 500 600

FUEL FLOW - Ib/hr

Figure 135. Inducer Head/Flow Calibration

245
 in
6

H-l
d
>
'S
>
fM
d
cö
5
u
o

£ ft
t>
CO

g
!«1 dV fa

i
10.000 ft

/
LAB 43

/
780F

f
I .

i
£ /
£ /
FUEL TEMP:

CUTBACK INLET

a /
CM /
1
TANK:
FUEL:

330 Ib/hr

/ >
L X /
i
V

/ o
c o
f f r c o «
J

S 1 0 • 2
* S
1 1 h s
/ - 1
/
^8
/
/
§
l\
v n
/

S^Z ^
s s
^*

§
nniinJ

CD

!«* - dV
MM

246
 120

a.

V/L
Q 450 lb/hr FUEL: LAB 43
20 FUEL TEMP: 780F
TANK: 20,000 ft
rpm: 65,000

0 0.1 0.2 0.3 0.4 0.5

V/L

Figure 138. Inducer 20,000 ft V/L Performance

INLET CUTBACK
IN THIS AREA

Figure 139. Cutback Inlet of the Inducer

247
.

120

100

a.
<I

200 300 400 500 600

FUEL FLOW - Ib/hr

Figure 140. Indacer Head Flow Calibration With

Cutback Inlet

The axial inducer teste i during this program is described below:

Speed 65,000 rpm

Q 500Ib/hr
No. Blades 3
6. 2 psid
Ns 8,375
OD 0.375 in.
Inlet Angle 7.16 deg
Type Flat Plate (Constant Lead)

Calibration of the axial inducer-conical inducer combination (Figure 141) showed

a lower head rise than the uninduced conical impeller, an effect believed to be a
result of increased leakage at the front shroud. Initial V/L testing of the com-
bination, shown in Figures 142 through 144, demonstrated a loss in the perform-
ance (head rise) compared with the cutback inducer performance. Modification
to the axial inducer inlet to provide sweepback to the inlet vane edge, and removal
of the vanes on the trailing edge of the axial inducer, failed to provide any per-
formance improvement of the axial/conical inducer combination. The use of an
axial inducer was therefore abandoned in this program phase.

The final phase of testing incorporated a jet pump configured to run in combination
with the cutback conical impeller. Design characteristics of the jet pump are de-
fined below:

Pressure Ratio 0.46

Capacity Ratio 0.5
Nozzle Diameter 0.0559 in.
Mixing Tube Diameter 0.112 in.

248
 100

I
a.

100 200 300 500

FUEL FLOW - Ib/hr

Figure 141. Axial/Conical Inducer Head vs

Flow Calibration

100

a.

Figure 142. Axial/Conical Inducer Sea Level V/L

Performance

249
 100
SPEED: 65,000 rpm
TANK: 10.000 ft
80 FUEL: LAB 43 —\
FUEL TEMP SOOP i

60
1 ^ ^^

1
a.
40 V
[ 450 lb/hr
-^

225 lb/hrl

—i ,»^ 330 lb/hr

20

0.1 0.2 0.3 0.4

V/L

Figure 143. Axial/Conical Inducer 10,000 ft V/L

Performance

Figure 144. Axial/Conical Inducer 20,000 ft V/L

Performance

250
A schematic of the jet-Induced conical impeller test conditions at V/L = 1. 0 Is
shown In Figure 145. Impeller flow rates and AP values are shown on the sche-
matic. The V/L test results for the iet-lnduced conical Impeller are defined In
Figures 146 and 147. This combination has shown that, with the existing design
configuration, the inducer arrangement is flow limited to 60 Ib/hr at maximum
speed and a V/L ratio of one. A V/L = 1.0 was also demonstrated at 70% speed
where output flow was limited to 30 Ib/hr. This limited flow capability results
because the USAAMRDL conical Inducer used throughout the test series was not
sized for use with a jet pump and cannot provide required output flow and head while
providing motive flow to the jet pump. To provide required boost stage output
flow at a head rise sufficient to charge the main pumping element, a higher
capacity conical impeller and a larger jet Inducer would be necessary. These
design considerations are discussed in subsequent paragraphs.

V/L tests are conducted by throttling the fluid below Its varor pressure to induce
the formation of vapor and therefore a fluid/vapor mixture. Upon entering the
pump element, the vapor is dissolved back Into the liquid as the pressure is in-
creased. These tests simulate a local restriction in the fuel line upstream of the
pump inlet. Another form of V/L operation is sometimes encountered when there
is an air leak into the pump inlet line. Under suction inlet fuel conditions, this
results In an air/fuel mixture which behaves dlfierently. The air remains pres-
ent throughout the pumping process, although the relative V/L decreases as the
fuel pump pressure increases. Since these V/L tests represent two different
phenomena, air/fuel tests were also conducted on the inducer system.

Testing of the jet-Induced conical Impeller using an air bleed into the pump inlet
resulted In sudden loss of prime In the pump Inlet. Pump performance was ob-
served to be satisfactory as V/L was Increased until the point of failure, which
occurred from 0.5 V/L to 0.75 V/L operating In the range of flows at which the
unit had demonstrated 1.0 V/L performance using the Induced line loss method
of test.

A performance curve showing pump AP vs V/L Is presented on Figure 148. These

data show that the pumping system was maintaining head rise up to the point of
performance breakdown.
The reasons for this sudden failure are not fully understood but are believed to
be:

1. Poor control of the amount of air being introduced into the line
may have resulted in instantaneous excessive V/L values.

2. The fuel may have the capacity to absorb small amounts of air
but larger amounts may more severely impact pump perform-
ance.

Introduction of a jet inducer into the pumping scheme has a significant effect on
the ability of the system to operate with low inlet fuel flowrates at V/L conditions.
With two-phase flow at Inlet velocities below 3 ft/sec, stable flow Is not main-
tained In the Inlet line. This low flow condition consists of a stratified plug, or
slug, flow on which an Impeller alone cannot maintain stable performance.

251
 SCHEMATICS SHOWING FLOW CONDITIONS
AT
V/L * 1.0 CONDITIONS

PRESSURE CAPACITY % NOZZLE

CONDITION RATIO RATIO FLOW
PdPIn Qu On
PnPd Qn Qt

220 Ib/hr 18 ptig 60 Ib/hr

f Qn
S L TANK
V/L - 1.0 [ Pn , Pd / 0.239 0.272 78.S
66,000 rpm
j^ ' 5 ptig \
SEA LEVEL
60 Ib/hr • Qu
-10.7 pilg

210 Ib/hr IS ptig 60 Ib/hr

10,000 ft TANK
V/L - 1.0
65,000 rpm
10,000 ft
L 5 ptig
A 0275 0.285 77.7

60 Ib/hr
■10.5 ptig

c
185 Ib/hr 10 ptig 60 Ib/hr

20,000 ft TANK

A
V/L - 0.96 0.38 0.324 75.5
66,000 rpm
-S ptig
20,000 ft
60 Ib/hr
10.7 ptig

c
1785 Ib/hr 36 ptig 600 Ib/hr

DESIGN
0.28 0.28 78.0
CONDITIONS
•6 ptig <
500 Ib/hr
10.7 ptig

Figure 145. Conical With Jet Inducer Test

Conditions

252
 100

80

&
a. 60 k. ^^.
<
O
UJ VoOlb/hrN. N^ Ib/hr
i * \

20
\ 260 Ib/hr

\
SPEED:
TANK:
N
66,000 rpm
SEA LEVEL
\

h-N
FUEL: LA343
FUEL TEMP: TSOF
,
0.2 0.4 0.6 0.8 1.0
V/L

Figure 146. Conical With Jel Inducer Sea Level

V/L Performance

a
o

Figure 147. Conical With Jet Inducer 10,000 ft

and 20,000 ft V/L Performance

253
 100

V/L

Figure 148. Conical With Jet Inducer V/L Per-

formance With Direct Air Induction

With the jet inducer in the system, a relatively high rate of flow through the im-
peller and nozzle loop is maintained, and the mixing action in the jet makes the
system insensitive to the type of flow in the inlet pipe.
3. Proposed Redesign

The demonstrated ability of the jet-induced conical impeller to operate with 1.0
V/L and with TVP +1 psi at low flowrates provides a baseline from which to
design a larger system which will have the capability of meeting the flows re-
quired for the advanced technology helicopter engine. This redesign proposes
a higher capacity mixed flow impeller and a resized jet inducer.

Test conditions at which the test hardware demonstrated V/L = 1 capability are
as follows (Reference Figure 145):

Pressure Ratio 0.239 to 0.38

Capacity Ratio 0.272 to 0.324
Percent Nozzle Flow 75.5 to 78.5
Area Ratio 0.25

From curves of conical inducer at 0.45 V/L, we can conservatively say that at
0.45 V/L, AH = 50% noncavltating head at flows less than 50% of design flow.

To design for 1. 0 V/L at 40 psi maximum nozzle AP and 500 Ib/hr useful flow,
the jet inducer modeled on the above data gives the following:

Pressure Ratio 0.28

Capacity Ratio 0.28
Percent Nozzle Flow 77
Area Ratio 0.2

254
Then,
Finn
= 1785

Qn = 1285 Ib/hr

The impeller requirements then become:

AP 85 psid
Q 2,280 pph
rpm fi5,000
N 2,193
s

Using the conical type USAAMRDL Impeller as a model, this results In an Impeller
OD of 0.78 In. and a shutoff pressure of 220 psl. The high specific speed In-
volved has shown the conical Impeller to be a poor choice for the higher flow
required due to the high shutoff head. This high shutoff head results in exces-
sive AP across the inducer nozzle and, therefore, high nozzle flows at low flow
conditions. To obtain a flatter head-flow curve, it is necessary to use a mixed
flow type impeller which has been developed for use on the MFP 330 and other high
flow CECO pumps. This impeller develops a flatter head flow curve and, there-
fore, will produce less pressure rise at low flows. It also has the potential of
higher V/L capability due to the favorable inlet configuration.

A mixed flow impeller to meet the requirements specified above would have an OD
of 0.56 in. and a shutoff pressure of 110 psi based on MFP 330 impeller perform-
ance. Performance Is estimated on Figure 149 and a comparative cross section
Is shown In Figure 150.

Power requirements for this higher specific speed inducer (Ns = 2200) using 23%
overall efficiency based on the Worthington curves calculate as follows:

85 x 2280
HP
HP = = 1 3io
1,31U
1714 (0.23)375
Figure 151 represents the predicted fuel pump total temperature rise with an
inducer sized for V/L = 1.0. Assuming a 350oF maximum fuel temperature
limit and a 1350F maximum fuel inlet temperature, a temperature rise of 2150F
is allowed. An evaluation of the engine operating map showed that this temper-
ature rise limit is approached only at the worst-case operating point of flight idle
at 20,000 ft altitude. The fuel temperature rise at Intermediate power level is
predicted to be below 100°F over the entire operating envelope.

255
 120

100

^
-^
80

E
A0
rpm 65,000

Ui INLET 0 psi
cc
0. 40 I I

20

0
400 800 1200 1600 2000 2400
FUEL FLOW Ib/hr

Figure li9. Mixed Flow Impeller - Estimated

Head vs Flow Performance

USAAMRDL CONICAL INDUCER

PROPOSED MIXED FLOW INDUCER

Figure 150. Comparison of Boost Inducer Cross

Sections
256
 440

400

I-

100% N - 65,000

Figure 151. Predicted Fuel Pump Total Temper-

ature Rise

257
4. Test Results

The test results obtained from the fuel pump inducer test program are summarized
below:
1. The jet inducer and conical impeller combination was capable
of low flow operation (60 Ib/hr at 65,000 rpm and 30 Ib/hr at
45,500 rpm) with V/L = 1.0 and/or 1.0 psi above TVP at the
pump inlet. A redesign of the pumping element is required to
meet the V/L specification over the entire flow range. The
flowrates at which the tests were successfully completed show
that an increase in capacity of 8 to 1 is required.

2. Performance of the conical inducer could be improved by re-

design of the inlet area. The "cutback" rework improved
performance, showing that a blockage or mismatch condition
exists In this area.

3. Performance of the axial Inducer was significantly less than

the predicted performance based on test data for larger units.
Further development In this area would be required to deter-
mine the specific design problem associated with this Inc'ucer.

4. Operation of the jet-Induced conical Impeller at 1.0 V/L ob-

tained by air Induction Into the Inlet line was not possible due
to sudden loss of prime experienced during all tests cf this
type which were conducted. It was Impossible to determine
maximum V/L obtainable due to Inconsistent results.
5. The Increased Impeller capacity required to meet the flow and
pressure requirements of the jet Inducer extend the capabilities
of the conical type Impeller beyond the normal range of useful-
ness. A mixed flow type of Impeller would be a more suitable
choice. This design has the potential advantages of Increased
overall efficiency and Improved vapor handling capability, both
of which are required for this application. Since this mixed
flow Impeller cannot be machined by conventional methods but
must be cast from specially prepared patterns, this may re-
quire a development effort due to the small size. The cost of
producing this type of Impeller Is many times the cost of a conical
Impeller In the small quantities required for a development effort.

6. The predicted fuel temperature rise characteristics with the

jet pump/centrifugal element pumping configuration are within
acceptable limits for the application.

7. The Increased capacity Impeller and jet Inducer will not sub-
stantially increase the basic size of the pump package.

258
5. Recommendations tor Additional Testing

1. Additional testing of the recommended jet pump/conical inducer

configuration is in order to verify the design approach and to
document the predicted fuel temperature rise.

2. The high-speed centrifugal pump elements tested have demon-

strated applicability for advanced engines from a predicted
performance standpoint. Endurance tests of the components
on contaminated fuel are in order to further verify applicability
from a mechanical design endurance standpoint.

E. POWER TURBINE OVERSPEED GOVERNOR - CHANDLER EVANS INC.

1. Background

Free turbine engines require a fast-response overspeed governor for protection

to prevent runaway speed. Also, in the cases where the engine control system
is electronic, it is advantageous to have a power turbine overspeed governor that
is redundant and operates Independent of the electronic control. Furthermore,
it is desirable to sense power turbine speed directly L eliminate stepdown gear-
ing. Conventional mechanical speed sensors operate in the 5,000-rpm speed
range at 250oF, whereas In this sytem rated power turbine speed is 36,000 rpm
and ambient temperature is 800oF.

The power turbine overspeed governor described In Section VII and shown sche-
matically In Figure 152 comprises a sensor which is an Integral part of the
engine power turbine shaft, and a governor mounted on the fuel control. The
speed sensor is formed by a set of closely spaced cantilevers arranged in a
circle around the power turbine shaft. As the power turbine shaft speed increases,
the centrifugal force causes the cantilevers to move radially outward. The canti-
lever inotion is used to vary an orifice and thereby generate a pneumatic pressure
signal. An increase in power turbine speed causes an increase in air pressure in
the orifice line. The increase in air pressure is sensed in the fuel control by a
bellows. Movement of the bellows is amplified by a follow-up flapper servo sys-
tem which operates to reduce the metering head across the fuel metering valve.
Hence, a turbine overspeed causes a propoitional reduction in fuel flow. The
point at which the overspeed governor cuts in is dependent on the spring preload
acting on the sensing bellows In the fuel control.

Critical items In the operation of this governor are the speed sensor and the
dynamic response of the governor. A speed sensor of this configuration has
previously been tested to 20,000 rpm. For this system the sensor has to be de-
signed to operate at 45,000 rpm (125% power turbine speed) in surrounding tem-
peratures up to approximately 800oF. The required response from overspeed to
reduction in fuel flow is on the order of 50 ms,
2. Test Program

Only the speed sensor and pressure-sensing bellows parts of the overspeed
governor were developed because the remaining parts of the system are of con-
ventional hardware and their performance can be determined from previous ex-
perience.

259
 u
>
o
0
•v
|
CO
C
0)

c
ft

rH

260
The objectives of the program are to demonstrate the performance and durability
(in an engine simulated environment) of the speed sensor and to confirm the
dynamic response capability of the overspeed governor.

The test Installation is shown schematically in Figure 153. The overspeed sensor
is shown enclosed with a heater element in an asbestos box wherein an 800oF
ambient temperature is maintained to simulate the engine environment. The
test series outlined for the overspeed governor is described below.

a. Room Temperature Calibration

Sensor cantilever displacement is calibrated against drive

speed and bellows stroke. Cantilever displacement is meas-
ured with a proximity Indicator. Bellows displacement is
measured with an LVDT.

b. Ambient Temperature Effects

Variation in ambient temperature from room temperature to

800oF is assessed during the calibration procedure described
in (a) previously.

c. Transient Response

A critical requirement of the overspeed governor is that the

governor be fast enough in reducing fuel flow to avoid destruc-
tion of the power turbine disk. This determines that the re-
sponse from speed increase to bellows stroke be faster than
50 mj. The transient response ot power turbine speed to
bellows stroke is determined by providing a step change in
cantilever displacement and measuring the bellows stroke with
an LVDT. The length and diameter of the air line from the
receiver nozzle to the sensor bellows are sized to simulate
the acfual engine installation.

d. Endurance Testing

The governor endurance is demonstrated by cycling the

governor for 50 hr about a nominal speed of 3R,000 rpm at
SOO'F ambient temperature. Calibration at 10-hr internals
determines degradation in performance.

261
 t
Si
CO
0)
H
u
o
K
d
u

s
ca
u
v
>
O
<u
c
3u
s
H
u

I
in

262
3. Testing

Considerable testing was done to develop a suitable drive and sensor shaft sys-
tem before a satisfactory measurement of cantilever deflection versus sensor
speed could be made. A high-speed air turbine was chosen to drive the sensor,
which was supported in carbon bearings lubricated with a drip oil system. The
air turbine did not have sufficient power to drive this sensor to 45,000 rpm due
to drive alignment problems. A second sensor was manufactured, integrating
the air turbine and sensor shaft (Figure 154). This drive obtained 30,000 rpm,
at which speed the shaft went into a critical vibration mode. A critical speed
analysis was made on the shaft design, and it was determined that the shaft had
to be shortened by 1.25 in. to avoid vibrations. A third sensor shaft was built
with shortened length. The reduced length necessitated breaking into the side
of the air turbine to measure the sensor speed. The third sensor failed because
one of the cantilever slots was curved at the root of the cantilever. This caused
this particular cantilever to deflect more than the others and to hit the sensor
nozzle. The cantilever slots are difficult to machine because of their width
(0. 008 in. - 0.010 In.) and the poor machineability of Inconel 718 (AMS 5663).

A fourth sensor shaft was made and successfully run to 45,000 rpm. However,
deflections were measured only to 43, 000 rpm, at which speed another failure
occurred. Deflection was measured by applying a voltage across the sensor
shaft to a micrometer with a light In series. At low speeds (<30,000 rpm), as the
cantilevers became very close to the micrometer (a few ten thousandths of an Inch),
a small spark was produced between the sensor and the micrometer, before me-
chanical contact was made, to light the bulb. At sensor speeds greater than
30,000 rpm, no spark was visible; the deflection was measured by observing the
light. However, at 43,000 rpm, the cantilever made contact with the micro-
meter and caused the sensor to fall.

80,000 rpm RATED

AIR TURBINE

ACCESS FOR
MAGNETIC SPEED PICKUP

OVERSPEED
SENSOR

Figure 154. Power Turbine Overspeed Sensor

and Test Drive
263
A fifth sensor shaft was fabricated, and a successful way of determining cantilever
deflection was developed as follows. A shallow angled cone was forced axially
Into the sensor shaft, displacing the cantilevers outward. The radial deflection
was measured with a dial Indicator. The dial Indicator was then replaced with a
proximity pickup and the tests were repeated. From the relationship of cone
axial position and radial displacement and cone position and proximity pickup
output voltage, radial displacement versus pickup voltage was determined. Hence,
the radial deflection was determined by measuring proximity pickup output voltage.

As the sensor shaft rotates, the air gap between the cantilevers passes under
the proximity pickup, causing the proximity voltage to drop; therefore, a circuit
was designed to pass only the highest voltage as a readout of cantilever deflection.

The test results of cantilever deflection versus sensor speed compared to

theoretical predictions are shown In Figure 155, It can be seen that the test
results follow theoretical predictions very closely.

0.04

!
a.
UJ 0.03

|
z
Z 0.02

o
g
a. 0.01

-i
<
5
<
cc 10.000 20,000 30,000 40,000 50,000

SENSOR SPEED - rpm

Figure 155. Overspeed Sensor Room-Temperature

Calibration
A test of the pressure gain of the sensor nozzle system Is shown on Figure 156.
Over the operating range of the governor, there Is a good correlation with the
computed design performance. At A2/Ai > 1. 5, actual test results give higher
values of Pos/Preg because of pressure losses In the line from the sensor
bellows to the air nozzle. The line bore was minimized (0.1 dla) to provide a
East response for the governor. The performance of the governor Is not affected
by the higher values of Pos/^reg because It Is not operating In this range.

■ 264
 1.4
26 in. LINE
(0.1 in. DIA'
31 3t
p
REG-»OP0S BORE)
1.2

A-i (0.041 DIA) A2 (0.064 DIA)-4

Az/A,

Figure 156. Pneumatic Characteristic of Power

Turbine Overspeed Governor

Test data, of overspeed pressure (PQS) versus sensor speed is shown in Figure 157,
The tests were first performed statically by mechanically forcing the cantilevers
to deflect the amount required at different sensor speeds. The tests were then
performed dynamically and show a close agreement with the statically determined
curve, because the frequency at which the air gaps (between the cantilevers) pass
the nozzle is much faster than time constant of the pneumatic system.

The effect of sensor ambient temperature on overspeed pressure (Pos) Is also

shown in Figure 157 for an ambient temperature of 650oF. It can be seen that
the effect of Increasing the sensor ambient temperature is to Increase the over-
speed pressure Pos. The reason for this shift is the reduction In Young's
modulus of the sensor cantilever material; a reduction in modulus causes an
increase in cantilever deflection for a given sensor speed. The change In Pos
would change governor cut-in speed 1% for a lOO'F change in ambient tem-
perature.

265
 34
nSTATIC TEST POINTS DEFLECTING CANTILEVERS|
MECHANICALLY (SPEED vi DEFLECTION •
32 KNOWN) ROOM TEMPERATURE
QDYNAMIC TEST POINTS
ROOM TEMPERATURE
I 30

o
o. 28

26

24
10,000 20,000 30,000 40,000 50,000
SENSOR SPEED - rpm

Figure 157. Variation In Power Turbine Governor

Pressure Gain With Ambient Tempera-
ture
The choice of the governor cut-in point is critical because It has to be within a
small margin between the maximum operating point of the engine (119%) and the
rotor burst speed (137%). Therefore, It Is necessary to provide sensor ambient
temperature compensation. This may be possible by selective use of material
thermal expansion properties In the area of the sensor nozzle.

The transient response of the overspeed governor was performed by simulating

a step change in cantilever deflection and measuring the bellows stroke with an
LVDT. The step change In cantilever stroke was simulated by a quick change
In a micrometer axial position against the sensor nozzle. The LVDT voltage with
respect to time was recorded on a strip chart recorder. The results of the
transient test are shown on Figure 158. The tests were done at different levels
of overspeed pressure PQS to show that the time response Is Independent of
sensor speed. The equivalent sensor speed Is Indicated on the traces. The meas-
ured time constant from change In cantilever position to sensor bellows stroke
Is on the order of 0,04 sec. The specified transient response Is 0.05 sec from a
turbine overspeed to reduction In fuel flow; therefore, the response of the sensor
Is satisfactory provided that It Is Incorporated In a system with fast response
from bellows stroke to reduction In fuel flow.

High ambient temperature endurance testing was performed on the sensor. The
test setup Is shown on Figure 153. Initial tests were performed with an ambient
temperature of OSO'F. The sensor was run at 36,000 rpm, and after 10 hours of
operation the cantilevers failed. The failed section Is shown on Figure 159,
The cantilever section was designed for unlimited life provided that the cycle
stress was low; however, Figure 159 shows that stress concentrations existed at
the corner point of the section. This problem could be eliminated by making the
cantilever slots wider and providing a radius at the root of cantilevers; also, shot
peenlng the sides of the cantilevers to remove the sharp edges of the cantilevers
would Increase the fatigue strength considerably.

266
 p
os

mm
m -,25 psia r(0 rpm) | 7

m it
BE
-Miliiliruiiiinit liilili
ü^iiii|i!'!l'i!!{'!lilt}iililli{-i^iii^'iiliS m
1 TIT m
FTTTF

Hit Mm

■ m■
m m
i mI
m TIME CONSTANT - 0.035 sec
trrrr mf
m S
m 1
!i
tiHm't{ttiHii!ttHHt1lfilii lliii m
m
im
i liii
^r
■
tTTT
^tt WWf
tttr ■fr i< 25,000 rpm) 4+r m
it 0.1 mi rrr ni}{|niffntn]t
m m TIT •!; Hr UijUliii. 411+11
m
U lllllimiln]
ME Si iiilliiilLiiiliiliiiiii ^26piia^- HL

8.
o
>

3
o
Q
>
>
CO

I3
c
<u
s
CO

{
o
"55
00

Figure 158. Transient Response Tests

267
 SCALE 14 X FULL SIZE
EVIDENCE OF
FATIGUE FAILURE
EMANATING FROM
CORNERS

Figure 159. Failed Cantilever Section

4. Test Results

The test results obtained during this program are summarized following:

1. Tests performed on the pneumatic components show that

the performance can be predicted accurately,

2. Tests of cantilever deflection versus sensor speed show

that the performance can be accurately predicted.

3. The sensor tested shows a shift in calibration equivalent

to a change in cut-in speed of 1% for 100oF change in sensor
ambient temperature.

4. The transient response of the overspeed sensor is satisfactory.

5. The design of the cantilevers needs improvements to increase

the fatigue life of the sensor.

5. Conclusions
Sufficient testing has been accomplished to demonstrate that the concept of
sensing the power turbine speed with the cantilever air nozzle arrangement is
feasible.

The sensitivity of the carMlever deflection to dimensional tolerances is such that

it is preferable to make the sensor separable from the engine power turbine
shaft so that the sensor can be calibrated at manufacture.
268
Experience gained during this program shows that design Improvements can be
made to the sensor. Figure 160 shows these Improvements. The sensor is
shown to be detachable from the engine power turbine shaft to allow the sensor
to be calibrated at manufacture. The cantilevers would be produced so as to
allow normal tolerances for the dimensions; the deflection at governed speed would
be ensured by grinding the diameter against the sensor nozzle with the sensor run-
ning at governed speed. The shield around the sensor would allow setting up of
the sensor nozzle by screwing the nozzle in until It touches the shield a.nd then
backing off at a predetermined angle. This would control the gap between the
nozzle and the cantilevers. The shield would also stop the cantilevers from
hitting the sensor nozzle in the event of an excessive overspeed.

,-JL.
SENSOR
ENGINE POWER DEFLECTION LIMIT
TURBINE SHAFT CANTILEVERS
AND SETUP SHIELD-

Figure 160. Improved Sensor Design

fi. Recommendations for Additional Testing

1. Further evaluations of the overall system dynamic response

are In order to verify applicability of the overspeed governor
to the advanced technology helicopter engine.

2. Methods of compensating for sensor ambient temperature

should be Investigated and a solution to the problem demon-
strated .
3. Experimental demonstration of the sensor installation and
setup procedures is In order to assure that the system can be
installed without ln-place calibration.

269
 SECTION X

CONCLUSIONS

The study has confirmed that controls and accessories for future Army turbo-
shaft engines must utilize advanced technology to obtain a size, weight and vul-
nerable area consistent with the basic engine. Full-authority electronic controls,
high-speed fuel and oil pumps, high-speed alternators and Integral engine starters
are recommended for controls and accessory Implementation.

The engine configuration used In the study was optimized from a controls/
accessory packaging and vulnerability standpoint. The baseline engine incor-
porated a rear-drive power turbine and an integral air turbine starter with no
airframe accessory power takeoff. While the design of a C&A system for a
front-drive engine configuration was not pursued to the detailed design stage. It
is not anticipated that the technology requirement and implementation of the C&A
components would vary significantly from those utilized on the rear-drive engine.
In a similar light. If an engine/airframe installation favored an electrical starter
or electrical starter/generator, the resulting Impact on the C&A system would
primarily be the requirement for additional envelope. The requirement for a
relatively low speed airframe accessory drive pad would significantly impact the
C&A system envelope and vulnerability resistance.

The study has also shown that an engine system designed with consideration
for vulnerability has also added benefits in weight, volume and simplicity.

The detailed recommended C&A component implementation for future systems

is outlined in the Recommendations section.

270
 SECTION XI

RECOMMENDATIONS

Recommendations for additional C&A component critical item testing are included
at the conclusion of the Phase III Critical Item Test Summary. The efforts on
the Integrated Accessory Systems for Small Gas Turbine Engines program have
indicated certain design trends for future Army helicopter engine and control
systems. The projected state of the art for a 1977 development time frame has
indicated that the following small, high-speed control system components and
integration techniques are feasible and should be considered for small Army
helicopter engines:

Flow Control - A full-authority electronic control should be used

in conjunction with a hydromechanical flow metering system. Sensors
would include a radiation pyrometer for turbine blade temperature
limiting. Inlet guide vane actuation would be hydraulic and integral
with the hydromechanical flow metering system. The power turbine
would incorporate a separate mechanical/pneumatic overspeed sensor.
Engine/airframe interfaces would be "fly by wire. " The control
system redundancy approach would be established as a function of
the engine application and mission.

Fuel Pump - The fuel supply system would be designed for suction
inlet conditions (1 psi above TVP, V/L = 1) and would consist of a
high-speed (65,000 rpm) centrifugal pump with inducer and jet pump
stages. The fuel pump would be packaged integrally with the hydro-
mechanical flow metering system, if practical.

Oil System - The oil system remains as one of the most vulnerable
of the engine systems. Vulnerability resistance can be improved by
the use of air bearings to reduce the oil system heat load. Radial
load capacity for air bearings is considered to be within the 1977
time frame, but axial load capacity exceeds the state of the art.
Pressurized air/oil or fuel/oil heat exchangers are to be avoided
for vulnerability considerations. For the system studied, an air/
oil heat exchanger integral with the oil sump and inlet particle
separator had sufficient heat exchanger capacity. A single-stage
"vane type" oil and scavenge pump would be used. A limiting oil
pump speed of 15,000 rpm was set by consideration of inlet cavita-
tion. An air/oil mist backup lubrication system with a 30-mln
operation capability would also be utilized.
Electrical - Electrical power requirements for the basic engine
system (computer and ignition) would be supplied by an alternator
integral with the main gearbox. The projected thermal environ-
ment is well within the operational limits of alternator components,
and the oil mist environment is not seen as a detriment to the
alternator design. The reliability of the alternator is projected
to be similar to that of other gearbox components.

571
Controls and Accessory Drive - A mechanical drive is recommended
for the coArols and accessories. A tower shaft drive off the gas
generator rotor is the recommended implementation. A vertical
drive to a controls and accessory location near the top of the engine
is recommended for vulnerability benefits.

Accessory Drive Pads - The use of the engine gearbox for airframe
accessory drive is discouraged because of the adverse impact on the
gearbox design. Advanced engines will utilize small, high-speed
rotating components which will be integrated within the engine struc-
ture without the conventional "hang on" gearbox. Provision of the
relatively large low-speed pads for airframe accessories is possible
at the expense of engine size, weight and vuluerability, but is not
recommended. The use of the ain'rame gearbox for accessory drive
will also provide drive redundancy in twin-engine installations.

Engine Starting - An air turbine starter system is recommended,

if an APU Is available or required for other airframe services.
For certain installations where the development expense of an
APU for starting may present an overwhelming cost disadvantage
to the approach, electrical starters can be utilized at the expense of
weight and vulnerability. ElÄirical starter systems can be more
attractive if the -65°F start requirement is relaxed. Development
of the starter as part of the engine system is recommended.

Power Turbine Interface - The packaging and resulting vulnerability

resistance of controls and accessories are improved with a rear-drive
engine as compared to a front drive. A rear-drive power turbine
also tends to provide a better engine configuration from the stand-
point of a smaller gas generator shaft diameter with the attendant
lower bearing and seal velocities, improved compressor hub tip
ratio, and lighter gas generator turbine disk.

272
 APPENDIX

ENGINE DESCRIPTION

The primary features and requirements of the engine used during this program
are outlined following.

A. GENERAL

The engine shall be designed for a dual-engine military helicopter application

and shall reflect the technology commensurate with formal development begin-
ning In 1977. It shall be of the direct-drive, free-power-turbine type and shall
provide an output power shaft In the rearward direction.
Except as specified herein, the engine shall comply with AV-E-8593 turboshaft
engine specifications unless the requirements are considered to be a hindrance
to the advanced technology desired in this program.

B. CYCLE CHARACTERISTICS (Typical)

1. Turbine Inlet Temperature 12880C

2. Compressor Pressure Ratio 10:1
3. Gas Generator Speed 65,000 rpm
4. Power Turbine Speed 36,000 rpm
5. Air Flowrate 3 lb/sec
6. Horsepower 600 hp

C. OPERATIONAL REQUIREMENTS

1. Environmental

1. Ambient Temperature -540C to 570C

2. Nacelle Maximum Temperature 80oC - Continuous
(150oC - 15 sec)
3. Altitude 20,000 ft
4. Fuel Temperature -540C to +570C

2. Mission Cycle

A typical mission cycle Is shown In Figure A-l and can be used for system
performance trade-off analysis,

3. Contaminated Fuel

1. The engine shall operate on fuel contaminated In accordance

with AV-E-8593.

2. Fuel system components shall be designed for a life of at

least 300 hr when operated on contaminated fuel.

273
 »«»., CRUnf «00 KNOTO. 0M0 «. 00» «MIR

MAXIMUM
rawiR rowiR .
ClIMO MMWUVfR

OROUNO IDLi mouNOim.!

Figure A-l. Typical Mission Cycle

4. Specified Fuel

The engine shall function satisfactorily throughout Its operating envelope for any
steady-state and transient operating condition without requiring adjustment of
the fuel system when using fuels conforming to and havine any of the variations
In characteristics permitted by MIL-T-5624, Grades JP-4 and JP-5, or JP-8,
except that operation shall not be required using fuel having a kinematic viscosity
greater than 12 cs.

5. Lubricants

The engine shall function satisfactorily throughout its operating envelope when
using oil conforming to and having any of the variations in characteristics per-
mitted by M1L-L-23699 or MIL-L-7808, except that operation shall not be re-
quired using oil having a kinematic viscosity greater than 13,000 cs.

6. Attitude

The engine shall be capable of continuous operation In nose-up or nose-down atti-

tudes of up to 45 deg from the horizontal and roll attitudes of up to 20-deg inclina-
tion to either side. The engine shall be capable of satisfactory operation for
30 sec in nose-up or nose-down attitudes of up to 90 deg and roll attitudes of up
to 45-deg Inclination to either side.

7. Inlet Particle Separator

Protection from sand and dust auJ foreign object Ingestion shall be provided by
means of an Inlet particle separator, which shall be self-contained and Integral
with the engine.

274
8, Condition Monitoring

The engine shall incorporate the necessary components to allow monitoring of

the overall engine condition and provide fault Isolation capability for all major
engine modules or subsystems. The condition monitoring system shall consist
of the sensors required for monitoring that will Interface at a suitable location
with a ground or flight diagnostic system. At least the following indications
shall be provided:

1. Gas generator rotor speed

2. Power turbine rotor speed

3. Output shaft torque

4. Measured turbine temperature

5. Oil pressure

6. Oil temperature

7. Magnetic particle detector(s)

8. Oil level

9. Engine start (number) and time (hours) counter similar to

MS17321.

10. Fuel filter bypass (If applicable)

11. Oil filter bypass.

The gas generator and power turbine rotor speed signals shall be an exact func-
tion of the gas generator and power turbine rotor speeds, respectively.

9, Suction Fuel System

An engine fuel pump shall be employed that shall provide fuel required to meet all
engine fuel flow demands whei supplied with fuel at the pump inlet at a pressure
of 1.0 psl above the true vapor pressure of the fuel and with a vapor-to-llquld
ratio of 1,0, The pump shall not be required to have a dry lift capability. All
contaminated fuel requirements set forth In the operational requirements, out-
lined above, shall apply. No fuel reclrculatlon to the vehicle tank shall be
allowed,

10, Engine Starting

This subsystem shall provide for two successive starts or start attempts of the
engine. For purposes of design of this subsystem. It can be assumed that suffi-
cient hydraulic, pneumatic, or electrical energy will be available from air-
frame sources.

275
11. Fuel Control

The fuel control shall not have provisions for an emergency manual mode of con-
trol. The following features shall be incorporated into the fuel control:
1. Isochronous power turbine governing

2. Provision for selected engine operation at power turbine speeds

from 85 to 115% of maximum operating speed at any power
level

3. A direct (closed-loop) temperature limiting control

4. Redundant power turbine overspeed protection

5. Load sharing capability for multiengine installation.

12. Lubricating System

The lubricating system shall satisfactorily lubricate the complete engine, without
change in lubricant, throughout its operating envelope, except tnat operation is
not required if the oil viscosity is greater than 13,000 cs. The oil reservoir and
cooler shall be furnished as component parts of the engine lubricant system and
shall adequately cool either MIL-L-23699 or MIL-L-7808 oil throughout the entire
operating envelope of the engine. The lubricating system shall require no ex-
ternal oil system.

13. Electrical Power

All electrical power required by the engine, including ignition, shall be provided
by an integral system. There is no requirement for generation of electrical
power for other than engine use.

14. Fuel Filter

If a fuel filter with finer than ISOOA« openings is required, it shall be serviceable
and a part of the engine. The capacity shall be sufficient to permit a cumulative
fuel flow equivalent to a minimum of 25-hr engine operation at maximum continuous-
rated static, sea level, standard day power with fuel contamination as specified in
AV-E-8593 without being cleaned. Main flow filter shall be provided with an in-
tegral bypass, which prevents washing of the filter element or collection device,
provisions for attaching instrumentation for remote indication of filter bypassing,
and an accessible visual impending bypass indication.

15. Measured Turbine Temperature Sensing System

A measured turbine temperature sensing system shall be provided. The device

shall be located at a point which will provide a temperature-indicating signal
so related to turbine temperature that a single temperature limit may be set that
will be applicable throughout the operating envelope of the engine. The temperature-
indicating signal shall be suitable for an external readout.

276

.■
16. Torque Sensor

The engine shall provide a signal for operation of a torque Indicator throughout
the complete range of the engine. The accuracy of the torque sensor, from 507r
rated output shaft speed to 12ß% rated output shall speed, shall be:

1. Within plus or minus 2% of the torque at maximum continuous-

rated power from zero output shaft torque to the torque at
maximum continuous-rated power

2. Within plus or minus 2% of the torque being measured from the

torque at maximum continuous-rated power to the transient
(0.2 min) torque limit.

277
 SELECTED BIBLIOGRAPHY

1. Dotseth, W. D., SURVIVABILITY DESIGN GUIDE SMALI^-ARMS

BALLISTIC PROTECTION, North American Rockwell Corporation,
USAAMRDL TR 71-41A, Novembcf 1971, AD 891122 L.

2. Davis, J. V., ADVANCEMENT OF SMALL GAS TURBINE COM-

PONENT TECHNOLOGY, Continental Aviation, USAAVLABSTR
69-10A, December 1969, AD 864281.

3. Davis, J. V., and E. J. Dellert, ADVANCEMENT OF SMALL GAS

TURBINE COMPONENT TECHNOLOGY, Vol. II, Continental Avia-
tion, USAAVLABSTR 69-10B, February 1970, AD 869323.

4. Richardson, H. L., and R. M. Tommasini, ADVANCEMENT OF

SMALL GAS TURBINE ENGINE ACCESSORY TECHNOLOGV, Final
Report, Curtiss-Wright Corporation, USAAVLABS TR 67-13,
April 1967, AD 652896.

5. White, A. H., and D. F. Wills, ADVANCED ENGINE CONTROL

PROGRAM FOR TWO TO FIVE LB/SEC AIRFLOW ENGINES,
Chandler Evans, R-492-6, March 1970, AD 867819.

6. Fuller, L. E., EXPERIMENTAL GAS TURBINE ENGINE STARTER,

Rocketdyne, Army Mobility Equipment Rest Development Center,
R-7541, August 1968, AD 837704 L.

7. McAnally, W. J. Ill, and M. T. Schilling, INVESTIGATION OF

FEASIBILITY OF INTEGRAL GAS TURBINE ENGINE SOLID
PARTICLE INLET SEPARATORS, PHASE II, FEASIBILITY DEMON-
STRATION, FINAL REPORT, Pratt & Whitney Aircraft, USAAVLABS
71-13, FR-4197, April 1971, AD 725593.

8. Colvln, D. C, ADVANCED ACCESSORY DRIVE SYSTEMS, Pratt &

Whitney Aircraft, USAF Systems Command FP 71-109, October 1971.

9. Nlcholls, Bernard H., AUXILIARY POWER SYSTEMS STUDY FOR

1975 FIGHTER AIRCRAFT, Garrett AlResearch Manufacturing
Division, January 1968, AD 825803.

10. Soule, T. L., D. E. Anschutz, and M. L. Perkins, STUDY REPORT

ON AN ADVANCED CONTROL SYSTEM FOR APPLICATION TO
SMAH FREE TURBINE ENGINES, Hamilton Standard, HSER-5300,
June 1969, AD 860679.

11. Johnson, H. T., DESIGN AND EVALUATION OF ADVANCED HIGH

SPEED FUEL PUMPS, Battelle, USAAMRDL TR-71-37, July 1971,
AD 729867.

278
w

SELECTED BIBLIOGRAPHY (Continued)

12. House, T. L., and D. R. Artis, ENVIRONMENTAL EFFECTS ON

ARMY HELICOPTER FLIGHT CONTROLS, American Helicopter
Society, American Institute of Aeronautics and Astronautics and
University of Texas, Joint Symposium on Environmental Effects on
VTOL Designs, Texas, 16 to 18 November 1970, American Heli-
copter Society.

13. Cross, R. H., J. D. Hawkett, and W. G. Lydlard, A COMPARISON

OF POSSIBLE TECHNIQUES FOR DRIVING AIRCRAFT AUXILIARY
SYSTEMS FROM GAS TURBINE ENGINES, Royal Aircraft Establish-
ment, England, RAE-TR-71116, June 1971, AD 890291.

14. A DDC BIBLIOGRAPHY ON HELICOPTER ENGINES AND ROTORS,

Defense Document Center, DDC-TAS-fi8-57, November 1968,
AD 680200.

15. Secunde, R. R., R. P. Macosko, and D. S. Repas, INTEGRATED

ENGINE-GENERATOR CONCEPT FOR AIRCRAFT ELECTRIC
SECONDARY POWER, NASA, TMX2579, June 1972.

16. McAnally, W. J., HI, and M. T. Schilling, INVESTIGATION OF

THE FEASIBILITY OF INTEGRAL GAS TURBINE SOLID PARTICLE
INLFT SEPARATOR, PHASE I- FEASIBILITY STUDY AND DESIGN,
Pratt & Whitney Aircraft, USA A V LABS TR 70-44, FR-3767,
August 1970, AD 875953.

17. Lee, C. H., INVESTIGATION AND DEVELOPMENT OF NEW CON-

CEPTS FOR IMPROVEMENT OF AIRCRAFT ELECTRICAL POWER
SYSTEMS, Garrett AiResearch Manufacturing Division, NASA
CR-86410, NAS 12-659, January 1970.

18. Mardonian, G. H., AIRCRAFT ENGINE VULNERABILITY REDUC-

TION STUDY, Avco-Lycoming, TR-1 05.24.2, August 1968,
AD 394126.

19. Mascianica, F. S., BALLISTIC TECHNOLOGY OF LIGHTWEIGHT

ARMOR, Army Materials and Mechanics Research Center, July 1971,
AD 519-230-L.

20. Evjin, J. M., and L. D. Miller, Jr., OPTIMIZING THE DESIGN OF

BATTERY - STARTER/GENERA TOR SYSTEM, SAE Paper
No. 710392, March 1971.

21. Liff, B., and R. B. Bossier, THE STARTING OF TURBINE EN-

GINES IN HELICOPTERS, 28th Annual National Forum of the Ameri-
can Helicopter Society, May 1972, A72-34509, Preprint 662.

22. Amarel, G. J., SECONDARY POWER SYSTEM FOR ADVANCED

TRANSPORT HELICOPTER, 28th Annual National Forum of the
American Helicopter Society, May 1972, A72-34480, Preprint 664.

279
 SELECTED BIBLIOGRAPHY (Continued)

23. Nicholls, B. H., A SECONDARY POWER SYSTEM STUDY FOR

ADVANCED ROTARY WINGED AIRCRAFT, USAAMRDL TR 72-13,
August 1972, AD 751854.
24. Tracy, J. R., LIGHTWEIGHT FRONTAL AREA ACCESSORY AND
DRIVE SYSTEM PROGRAM, Pratt & Whitney Aircraft, FRDC,
FR-3254, Naval Air System Command, June 1969,

25. Karpis, Levine, Sample, and Smith, SECONDARY POWER SYSTEM

STUDY FOR ADVANCED ROTARY-WING AIRCRAFT, The Boeing
Company, Vertol Division USAAMRDL TR 71-52, November 1971.

672.76
280