c Thorsten Renk 2016 - 2021

For private use only. This text may not be sold, re-distributed or adapted without
explicit permission by the author.
Flightgear Space Shuttle
Flight Manual

Standard Edition
Text version of April 25, 2021
refers to Shuttle milestone 13 / Flightgear 2018.3 and higher
2
Contents

I A first mission 11
1 First steps 13
Using the Shuttle with FG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Shuttle-specific commandline options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Using default FG terrain from orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Simulation options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Save/resume and scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Mission control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
A look around the cockpit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Avionics — the MEDS layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Avionics — navigating the DPS layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Entering items into the DPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
The ADI ball . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
The head-up display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Overhead analog gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Night operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2 Launch 31
General overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Launch using autopilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Manually flying a launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3 Orbital insertion 41
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Attitude control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Propellant dump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
The state vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
The orbital elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Shuttle orbital maneuvering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
MPS shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
The insertion burn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

4 Preparing for orbital operations 49

Thermal management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
DPS systems management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Orbital DAP selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Starting the radiator loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Temperature readings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Automatic attitude control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Antenna operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3
4 CONTENTS

5 Payload handling 59
Launching payloads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
PDRS deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
RMS arm operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
The payload retention system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Using software overrides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
RMS powerdown and jettison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

6 Coming home 65
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Pre de-orbit preparations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Planning the de-orbit maneuver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Terminal area energy management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Final approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Training TAEM from the carrier aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Manually flying an entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

7 Troubleshooting 79
Launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
On orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Approach and touchdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

II Advanced topics 85

8 Nominal procedures 87
Crossfeeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Mounting the COAS frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
IMU alignment using COAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
IMU alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Fuel cell purge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Cabin air EVA preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Extravehicular activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Deploying the air data probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Single engine OMS and RCS burns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
The OMS kit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Launch performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Circuit breakers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

9 Failures 97
Types of failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
The caution and warning system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Software CWS messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
MEDS fault messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

10 Trivia 107
CONTENTS 5

III Orbital maneuvering and Rendezvous 111

11 Orbital maneuvering basics 113
The coordinate system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
The prograde burn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
The normal burn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
The radial burn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
In-plane orbital maneuvering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Changing the orbital plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Finite propellant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Maneuvering using PEG-7 targeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

IV Reference material 119

12 DPS pages 121
GNC functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Systems management functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Software utility functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

13 Mission files 137

Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Launch site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Countdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Ascent trajectory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Abort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Post-MECO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
DAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
RMS auto sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Orbital targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Notepad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

A Glossary of acronyms 145

B Index values of DPS functions 147

C Key bindings 149

6 CONTENTS
Preface

’The world’s greatest all-electric flying machine’ is how not been flown by Atlantis, for instance the tail cone and the
the astronauts aboard Columbia’s maiden flight described air data probe mounted on Enterprise during the early glid-
their ride after coming back. This may no longer hold ing tests or the standard weight external fuel tank used by
true today as complexity of avionics has made tremendous Columbia.
progress since the Shuttle first launched, but the spacecraft
continues to be a remarkably complex piece of machinery,
capable of captivating people’s imagination. Realism
The prime selling argument for simulation software seems
Scope to be realism, yet the number of people who really want
to learn to operate a realistic simulation seems to be much
The Flightgear Space Shuttle aims to provide a simulation smaller than the number who wants to have one. The Flight-
that is as faithful to the capabilities and systems of the orig- gear Space Shuttle simulation is in many aspects very real-
inal as possible. However, Flightgear is by nature a flight istic, which means that it is near impossible to simply start
simulator, not a mission simulator. The simulation takes it up and fly somewhere. It takes months of training to learn
the Shuttle as a flying vehicle with given characteristics and to operate a real Shuttle, and while that involves also deal-
simulates whatever you want to do with it — there is no as- ing with off-nominal situation, don’t expect to master the
sumption hard-coded that the Shuttle needs to launch from Shuttle quickly even for a nominal mission.
a particular site or needs to land on certain runways only. As for the realism of the simulation in particular, aero-
You can test how the ascent performance changes if you try dynamic coefficients are taken from a large body of NASA
La Paz as a launch site where you are outside of most of data — real flight data for the nominal envelope, wind tun-
the atmosphere. Equally well, you can explore whether you nel data for off-nominal attitudes. The modeling thus in-
can land the Shuttle on a moderate sized regional airport in cludes even details like the cross-coupling between con-
your vicinity. On the other hand, there is little material in trols, for instance moving the elevons alters the airstream
terms of re-enacting historic Shuttle missions provided with at the aft fuselage and causes suction forces that give addi-
the simulation — most importantly there is no real mission tional yawing moments.
control center simulated. The aerodynamic model is impossible to fly ’raw’ since it
You can (and should) use the simulation to explore the is yaw-unstable at Mach numbers larger than two — a flight
’What if?’ of space flight. A lot of attention has gone control system (FCS) always needs to manage attitude such
in giving plausible responses for off-nominal situations (in that the dangerous region is not entered. The complete FCS
fact, about 90% of the code deals with off-nominal situa- is developed from scratch for the simulation and not based
tions). Thus, if you are curious as to why a procedure like on NASA data or software, i.e. while it accomplishes the
the OMS fuel dump is done during the preparation for en- same things (or at least tries to), the principles it works on
try, you can drop the procedure and experience later how internally are in general different.
the changed center of gravity affects the handling during at- The same is true for the higher level autopilot and guid-
mospheric flight. You can disconnect a power bus from the ance software. This means that some of the guidance inter-
fuel cells, and you will learn what systems require that bus nal parameters displayed in the real Shuttle are not available
and to what degree redundancy management allows you to in the simulation because they have no counterpart in the
continue the mission. way the simulated guidance works. One notable difference
The simulated Shuttle in particular is OV-104 Atlantis in is also that simulated guidance works a lot faster than real
a late configuration, i.e. with a full glass cockpit and HUDs, guidance (as it runs on 2016 hardware, not on 1970 hard-
guidance software with the contingency abort functionality ware) — essentially solutions converge instantly.
that was added to all Shuttles after the Challenger accident The simulation of on-board systems is done as close as
and super-lightweight external tank — and (in the few cases possible to the real counterpart — for instance there is a
where this matters) hardware specifics of Atlantis. It is how- real radiation and heat transfer model solved to compute
ever possible in-sim to select equipment that has historically temperature distribution across the orbiter giving meaning

7
8 CONTENTS

to operation of the freon loop, tanks are driven by real pres- educated hypotheses as to where problems might have orig-
sure models which reflect leakage or blockage of reservoir inated.
valves, lines can freeze if not heated, components consume In contrast, this flight manual aims to provide a tutorial-
power individually etc. like self-contained introduction to both spaceflight in gen-
The interface to the avionics is also kept as close to the eral and operating a Space Shuttle in particular (and in ad-
real counterpart as feasible, including the whole framework dition covers implementation-specific issues) — at the ex-
of dealing with errors — for instance screens can lose con- pense of not going into details of systems and being some-
nection to the display processor, display processors can lose times sloppy on real procedures for the sake of simplicity.
connection to a data bus etc. It is however consistently not In addition, the flight manual contains chapters with refer-
possible to patch avionics software or display its internal ence material as it is included in the simulation, such as a
data as the real Shuttle software is not present in the simu- list of all implemented avionics pages including which op-
lation. tions are functional.
Ideally, both references should be read side by side.
There is more to the simulation than this manual can rea-
Navigation sonably cover, and anyone interested in the details will find
them in the SCOM. In particular anyone interested in a
Navigation in space in reality is a complex topic which has close-up description of the function of the various systems
historically required e.g. tracking of the Shuttle by ground should read the corresponding sections in the SCOM.
radar stations and uplinking the position corrections ob-
tained that way. The reason is that the Shuttle’s navigational
state vector is essentially derived from integrating the accel- Note on units
eration signals from inertial measurement units which has
some error rate over time. Computations internally (unless in JSBSim native systems)
Due to the considerable unknowns in this process, the ex- are done in SI units (kilometers, kilogramm, Newton,. . . ).
act way information from the inertial and various other atti- On the other hand, the Shuttle’s systems display imperial
tude sensors can be combined into state vector updates via units (miles, pounds, pounds force,. . . ) and so is the Shuttle
Kalman filtering is not fully part of the simulation. How- Crew Operations Manual. The general rule is therefore that
ever, some care has been taken to optionally divorce simu- on any display that really exists aboard the Shuttle quan-
lated ’truth’ from the state used by the various instruments tities are displayed in imperial units, whereas on menu di-
and the guidance routines. alogs or the FG-native HUD layer, quantities are displayed
In particular, attitude drift of the instrumentation in or- in SI units.
bit is simulated when the star tracker is not available, and
precision positioning during rendezvous operations requires
the operation of the radar. The inertial units of the Shut- Technical references
tle need to be calibrated periodically with the help of the
star trackers. Area navigation when approaching the land- The following (incomplete) list of documents has been
ing site is simulated with a simplified model for the Kalman used to model the Shuttle’s aerodynamics:
filter. This gives at least a flavour of the challenges real
space navigation would pose. Aerodynamic design of the Space Shuttle Orbiter,
Bornemann and Surber,
Rockwell International Space Systems Group.
Relation with SCOM
Entry Dynamics of Space Shuttle Orbiter With Longitu-
The Flightgear Space Shuttle is distributed with the original dinal Stability and Control — Uncertainties at Supersonic
NASA Shuttle Crew Operations Manual (SCOM). The sim- and Hypersonic Speeds,
ulation can be (for the most part) operated by following this Stone and Powell,
documentation, and the SCOM has been the main guideline NASA.
for implementing the Shuttle’s simulated systems.
However, the SCOM is not intended as a sole reference or Space shuttle: Determination of the aerodynamic
training course. Astronauts receive extra training for their interference between the space shuttle orbiter, external
mission and are assumed to have a background knowledge tank, and solid rocket booster on a 0.004 scale ascent
in physics, engineering and spaceflight. The SCOM is not configuration,
structured as a tutorial how to operate the Shuttle or to teach Ramsey, Buchholz, Allen and Dehart,
the basics of spaceflight but explains what its systems do NASA.
(in sometimes excruciating detail) such that crews can form
CONTENTS 9

Why the Wings stay on the Space Shuttle Orbiter during Basic vs Deluxe Edition
the First Stage Ascent,
Carl. F. Ehrlich There are two versions of this manual — this Basic Edition,
and a more detailed Deluxe Edition. The general idea is
Aerodynamic Design Data Book Volume 1 - Orbiter that the Basic Edition contains all information that is nec-
Vehicle STS 1, essary to use the Shuttle during a simple regular mission
Rockwell International Space Systems Group while neglecting some details of the operation, i.e. launch,
maneuver in orbit, dock to ISS, handle payload, de-orbit
For modeling the Shuttle’s various systems and their and land.
operation in detail, the following workbooks have been In contrast, the Deluxe Edition is intended to detail all
used for reference: simulated aspects of Shuttle operations and in addition pro-
vide background information on spaceflight in general and
the Shuttle in particular.
Guidance and Control
In particular, the Deluxe Edition of the manual contains
Insertion/Orbit/Deorbit Workbook,
the following additional chapters:
USA006501, United Space Alliance LLC
Rocket engines: An overview of how the various types of
Navigation Aids Workbook, rocket engines work, what their engineering and oper-
USA006055, United Space Alliance LLC ating challenges are and what their controls and moni-
toring devices in the cockpit are.
Entry, TAEM and Approach/Landing Workbook,
USA005512, United Space Alliance LLC Area navigation: This chapter describes how navigation
information is obtained during the last stage of entry,
Entry Digital Autopilot Workbook, how it is cross-checked and finally incorporated into
USA006497, United Space Alliance LLC the state vector and what the procedures to deal with
bad or incomplete sensor information are.
Inertial Measurement Unit Workbook, Aerodynamics: An introduction into the flight character-
USA004488, United Space Alliance LLC istics of the Shuttle, offering background information
on topics like yaw instability, Mach-dependent control
Air Data System Workbook, efficiency, cross-couplings or glide characteristics and
USA006080, United Space Alliance LLC explaining the usage of the NO Y JET entry mode.

Intact Ascents Abort Workbook, Off-nominal procedures: All checklist and techniques to
USA007151, United Space Alliance LLC deal with systems problems such as loss of an electric
bus or display re-configuration after loss of an IDP are
described in this chapter along with the relevant back-
Contingency Aborts,
ground information on the affected systems.
USA005671, United Space Alliance LLC
Communication: This contains a summary of the various
Flight Procedures Handbook Ascent/Aborts, antenna systems aboard the Shuttle, how they work,
JSC-10559, NASA how they need to be configured for various mission-
specific requirements and how to deal with communi-
Star Tracker, Heads-Up Display, Crew Optical Align- cation problems.
ment Sight Workbook,
USA006082, United Space Alliance LLC DPS memory management: In reality, the procedures to
load the mission-phase specific software into the Shut-
tle’s computers can be quite involved. This chapter
Data Processing Workbook,
provides the necessary background information and
USA005350, United Space Alliance LLC
procedures for realistic interfacing with the data pro-
cessing system.
Rendezvous/Proximity Operations Workbook,
RNDZ2102, NASA The Backup Flight System: The BFS is a safety measure
in case the primary software fails. This chapter ex-
Flight Procedures Handbook Entry, plains how to configure and engage the BFS in an
JSC-11542, NASA emergency to fly the Shuttle safely into orbit or to the
ground.
10 CONTENTS

Real orbital maneuvering: The Shuttle does not carry

enough propellant such that orbital maneuvering could
be done based on analytical solutions followed by cor-
rections — hence numerical orbital solvers to do the
targeting need to be used. This chapter describes how
such a solver can be used with the simulation and how
the on-board PEG-4 targeting routines can be used and
how the avionics targeting differs from mission control
targeting.

Rendezvous: Flying the Shuttle to another object in orbit

requires a careful choreography of maneuvers which
already starts from the launch. In addition, it requires
to use on-board sensors to track the rendezvous target
and correct navigational errors during the last stage of
the rendezvous. This chapter provides the background
information and procedures to bring the Shuttle from
launch into the proximity region of a target.
Proximity operations: Close to a rendezvous target, a spe-
cial set of piloting techniques is used to hold a posi-
tion relative to a target, approach to the target and do
braking burns, fly around the target or dock with the
target. The chapter provides the general orbital me-
chanics context to understand these procedures and de-
scribes many useful maneuvers in detail.

Intact launch aborts: For any single engine failure, the

Shuttle can be flown to a safe landing or (dependent on
when the failure occurs) even reach orbit. This chapter
provides background information and tutorials for all
intact abort scenarios.

Contingency aborts: For the loss of two or more engines,

there is generally no window for a safe landing, rather
the focus is on crew survival. The chapter discusses
the general theory behind such contingency aborts and
contains walkthroughs for two rather different cases.
 Part I

A first mission

11
Chapter 1

First steps

Using the Shuttle with FG SpaceShuttle-mission: You find Atlantis in a sta-

ble, circular orbit with all systems configured for or-
This manual assumes that you’re reasonably familiar with bital operations, about two days into the mission. This
Flightgear in general, i.e. that you can start at a selected is the proper mission phase to train orbital operations
airport or geographical location, look at the commandline like docking or payload capture (e.g. via selecting
created by the launcher of your choice, know how to set up a scenario), enjoy the view or learn de-orbit prepara-
your control devices (joystick, mouse, pedals, . . . ), config- tions.
ure graphics quality for the performance of your computer
etc. If this is not the case, please refer to the Flightgear User SpaceShuttle-entry: With all systems prepared,
Manual and the Wiki for these topics. you find yourself at the entry interface, just at the
fringes of the upper atmosphere, on a downward trajec-
Note: Make sure that you don’t have auto coordi- tory. In just a few minutes, the heat shield will have to
nation enabled when using the Shuttle and that your deal with temperatures close to 3000 degrees Fahren-
controls center precisely when you let go. If you heit.
can’t center your controls, you will be unable to sta-
bilize Shuttle attitude in orbit, because unlike an air- SpaceShuttle-TAEM: About 30 kilometers high, sixty
plane a spacecraft has no dynamical stability. You miles to site and a good ten minutes to go, this places
may need to configure deadbands. you right at the terminal area energy management
(TAEM) interface where the Shuttle starts to be flown
Unlike most aircraft, the Shuttle needs to be started in a
like an aircraft. You need to dissipate energy during
mission stage, for instance using
the pattern to reach a good final approach.
--aircraft=SpaceShuttle-launch
or SpaceShuttle-approach: Aimed at the runway
--aircraft=SpaceShuttle-approach. from about 12.000 ft and a good seven miles out, you
This makes sure the Shuttle starts in the correct config- need to fly Atlantis to a save touchdown.
uration. Different from a normal aircraft, weight, aerody-
namics and performance change quite dramatically over the SpaceShuttle-glide: This stage is not part of a nom-
course of a mission, and you’d be hard-pressed to land a inal mission but aims to re-create the gliding tests done
Shuttle with a fully loaded external tank attached! from the Shuttle Carrier Aircraft with Enterprise. The
The most important scenarios are: starting position is from about 36.000 ft on the back
of the Shuttle Carrier Aircraft (SCA) in good distance
SpaceShuttle-launch: Atlantis is on the pad, fully
to the runway. There is the opportunity to fly some S-
fueled with about five minutes to ignition. You will
turns and perhaps five minutes of flight before touch-
need to do the last launch preparations yourself, most
down.
importantly start the auxiliary power units (APUs),
make sure the launch guidance targets are specified
The recommended way of specifying location is differ-
and possibly activate the autopilot.
ent for these options. For TAEM and approach, airport (and
SpaceShuttle-orbit: Some eight and a half minutes optionally runway) where you want to land can be given di-
after launch, you find yourself in a stable elliptic orbit rectly (note that airports are labeled by their ICAO code)
after main engine cutoff. You need to disconect the and you will be placed with a suitable offset from the air-
external tank next, plan insertion into a circular orbit port. For instance
and prepare Atlantis for orbital operations. --airport=KTTS --runway=15

13
14 CHAPTER 1. FIRST STEPS

Figure 1.1: Layout of the Space Shuttle forward cockpit with the names of the various panels indicated

will place you right in the approach onto runway 15 at trying to determine an airspeed when in the vacuum of low
the Shuttle Landing Facility, Kennedy Space Center. orbit).
In the real Space Shuttle, many configurations of the
Note: In order to land on the specified runway in the software are mission-specific (’i-loaded’). In FG, this is
TAEM stage, you need to fly a straight-in approach. supported by mission files. A mission file may contain
If you fly the nominal overhead approach, you will the hardware configuration (payload, type of external tank
land from the direction 180 degrees, i.e. on runway flown, . . . ) as well as orbital targets, landing sites for abort
33 in the above example. scenarios, abort criteria as well as gains and deadbands for
the orbital maneuvering digital autopilot software and or-
In contrast, for the orbit and entry stage, the initial posi- bital parameters of a rendezvous target.
tion needs to be specified in geographical coordinates (the The mission file needs to be stored under
entry interface typically is some 4000 miles from the in- SpaceShuttle/Mission/ The precise file to be used can
tended landing site) and the current orbit is then given by be selected from the commandline via
heading. For instance --prop:/mission/filename=my_mission.xml
--lat=0.0 --lon=-160.0 --heading=45 by default mission.xml is used.
places Atlantis onto a viable entry for a landing at Van-
Note: If you want to edit mission files, install the
denberg AFB in California whereas
Shuttle into a location where you have read/write
--lat=56.0 --lon=0.0 --heading=90
access (typically your home directory) and config-
corresponds to a 56 degree inclination orbit. ure the Aircraft search path to that location in the
Positioning for launch can be handled either way — just launcher. Usually you will not be able to edit files
specifying an airport will place the pad rather crudely onto a in a system-wide installation as a normal user.
runway though (FG thinks the Shuttle is a normal airplane)
— so specify the coordinates of the launch pad for greater Since in orbit you’ll be moving around with Mach 27,
accuracy. In particular you may want to switch off real weather fetching and, prob-
--lat=28.6206 --lon=-80.6133 ably most important, automatic scenery downloading via
is launch complex 39 at Kennedy Space Center. terrasync. Orbital motion will make FG request to down-
In general, passing other options like airspeed and alti- load all of the available scenery in a rather short time, and
tude can not expected to work or might even crash the sim- unless you have an extremely fast internet connection this
ulation (the flight dynamics computation runs into problems may simply not work.
SHUTTLE-SPECIFIC COMMANDLINE OPTIONS 15

Figure 1.2: Layout of the overhead panels with names indicated

The standard FG terrain and weather simulation are not changes. If the view needs to be moved forward/backward,
particularly well suited to provide visuals from orbit. For also press Ctrl together with the center mouse button.
altitudes larger than 30 km, the use of the Earthview orbital Using mouse gestures, every point of the cockpit can be
rendering engine is highly recommended. Earthview can viewed from every angle, i.e. there doesn’t have to be a
be switched on and off automatically at a certain altitude as specific view to operate a particular panel, you can move
specified in the Shuttle simulation options. any view to the right spot.
Of the three FG rendering engines, only the Atmospheric However, to make things easier, the combination
Light Scattering (ALS) framework is designed to provide Ctrl + v brings up the Shuttle view manager (Fig. 1.5)
visuals of the atmosphere from space. Also, the hard light- which is usable from any internal view. This shows a top
ing of the Shuttle created by the absence of atmospheric view of the flightdeck, indicating a number of views (usu-
stray light in orbit and many other effects are only modeled ally to specific panels). Simply clicking on the view moves
in ALS, thus this is the recommended rendering framework the current view to the indicated position. This allows to
for best visuals. quickly access some of the less easily reached panels.
A final note on managing the view: Flightgear has a cou-
ple of pre-defined and craft specific views. While it is pos-
sible to cycle through all views using v , and Shift + v , Shuttle-specific commandline options
all viewpoints can also be moved dynamically. Usually
this is done by depressing center mouse button and mov- There’s a handful of special commandline parameters you
ing mouse up/down or left/right for corresponding view can pass to the Shuttle simulation that are not part of the
16 CHAPTER 1. FIRST STEPS

Figure 1.3: Layout of the aft panels with names indicated

Figure 1.4: Layout of the right aft panels with names indicated
USING DEFAULT FG TERRAIN FROM ORBIT 17

Using default FG terrain from orbit

As stated above, the recommended way to render visuals
during the orbital portion of a Shuttle mission is the Earth-
view engine. It is however possible (and perhaps even de-
sirable for launch abort scenarios where the altitude never
grows too high) to use the default terrain and weather sim-
ulation. This is not forseen in the GUI and requires several
overrides to be set via the commandline.
Since only Advanced Weather tries to adjust the atmo-
sphere visuals with altitude, it is generally advisable to not
use the Basic Weather engine in this situation.

Note: If you use the default terrain and weather sim-

ulation from orbit, you do this at your own risk.
Figure 1.5: The Shuttle view manager dialog. Memory consumption might be too large for your
computer and cause severe performance loss or a
crash. Do not file a bug report if this happens - you’ve
been warned!
general Flightgear set. The format in the commandline or The following properties can be set to make you see ter-
the launcher is always rain and clouds for hundreds of kilometers:
--prop:<property_path>=<value> /local-weather/config/aux-max-vis-range-m
where <property_path> is one of the options below If you’re using Advanced Weather, this is the maximal
and <value> the chosen setting. As discussed above, range to which the engine will allow the visibility to grow
at high altitudes. The parameter is actually the log of the
/mission/filename desired value, i.e. to have 600 km maximal visibility, you
is used to load a specific mission file. For practicing var- need to set it to 13.3046849341983. The Advanced Weather
ious TAEM geometries, engine will then decide upon the actual value of the visibil-
ity based on altitude.
/sim/presets/TAEM-approach-course-deg
/sim/rendering/clouds3d-vis-range
can be used to divorce the selected runway direction from This will allow clouds to be drawn to the specified range
the direction the Shuttle approaches, using this option any (in meters). Since in AW beyond some 80 km distance,
initial point for the TAEM mission phase can be chosen. clouds are represented by impostors anyway, the perfor-
The appearance of the displays can be influenced by mance impact of this is modest.
/sim/config/shuttle/font-selection /sim/rendering/static-lod/bare
This range (again in meters) determines how far out the
which can be set to 0 or 1 (default is 0). Selecting the al-
terrain is loaded. This needs to match the setting of the
ternative option selects a font that is closer to the real Shut-
maximal visibility range for a consistent experience.
tle displays, but unfortunately doesn’t include all symbols
/environment/snow-level-m
that are needed. Finally, if you need a ’lean’ Shuttle which
The need to modify this parameter is perhaps surpris-
starts up faster, you can influence with
ing, but the environment rendering code does not consider
/sim/config/shuttle/mdu-groups Earth’s curvature — as a result, terrain truly far away where
how many of the displays are initialized. The value can the curvature is notable might be drawn in snow. Setting this
range from 0 to 4, and for each step three displays will be to a high value will consistently show unsnowed terrain.
initialized. Thus, if you don’t need the aft displays and After setting these parameter from the commandline,
don’t want to spend resources on updating them, you can make sure the corresponding GUI options/sliders are not
set the value to 3. The parameter defaults to 4. modified, as this will re-set the parameters to the range the
GUI allows.
/sim/presets/ISS-prox-x (y,z)
can be used to force the detailed ISS simulation close to
the Shuttle, wherever the Shuttle and the orbital state of the Simulation options
station specified in the mission file might be. The model is
placed in proximity coordinates (in m), so for instance spec- The simulation options specific for the Shuttle are found in
ifying x to 1500 places ISS 1500 m ahead of the Shuttle. the menu under SpaceShuttle → Simulation Options (see
18 CHAPTER 1. FIRST STEPS

Fig. 1.6). You may need those to access educational aspects by a factor of 10 or 100 to quickly see the effect of
of the simulation, tune performance to your computers ca- thermal management (educational).
pabilities and configure things like payload or equipment.
Navigation: By default, the simulation assumes navigation
is never perfect and has errors which need to be re-
duced by the appropriate procedures (for instance by
operating the star tracker). This option allows to use
the actual simulated state of the Shuttle to be shown
on the instruments, i.e. all navigation readings can be
assumed to be 100% correct regardless of whether nav-
igation equipment is operated at all (educational).

DPS realism: In reality, the Space Shuttle has quite in-

volved procedures managing the memory configura-
tion of its computers. Mastering these takes quite a
while and is definitely too heavy for anyone who is not
deeply interested in Shuttle procedures. When the sim-
plified option is chosen, many computer-related things
will ’just work’, whereas with the realistic option the
real checklist needs to be followed.

MDU update speed: The nine screens in the cockpit,

known as main display units or MDUs, potentially
have to display hundreds of parameters. Fetching
these, processing them and formating them on-screen
causes a sizable performance hit. To reduce it, the
MDUs are updated in a staggered fashion (i.e. not all
in every frame). With this parameter, the update cy-
cle of the MDUs can be influenced from slow to fast.
Note that you can switch off MDUs you do not need to
reduce the performance impact (performance).

ADI ball detail: The 3d motion of the ADI ball on the

MEDS primary flight display is the single most ex-
pensive instrument to render. This option selects what
amount of detail is spent on the ADI. At lowest setting
it is not animated at all, at highest setting it closely
Figure 1.6: The simulation options menu. mirrors the original with almost every detail (perfor-
mance).

Cockpit detail: By default, only the forward panels of the

Some of these options are persistent, but flightdeck are rendered. If this is too heavy to ren-
since FG treats the various selectable scenar- der, the HUD view can be used together with the FG-
ios as ’different’ aircraft, options selected for native HUD ( h ) and two MDUs projected as popup
SpaceShuttle-launch are not remembered in windows — this is the least-performance consuming
e.g. SpaceShuttle-orbit. option. Alternatively, a full and highly detailed cock-
pit mesh is available showing also overhead and rear
The first section deals with the simulation. The various panels which can be accessed with this option (perfor-
options are: mance).

Thermal system computation speed: The simulation Show launch gantry: By default, the Shuttle in launch po-
solves a set of radiation balance and heat diffusion sition is just placed on a pad. If this option is selected,
equation to get the temperatures across several parts the full service gantry model is also shown next to the
of the orbiter. In reality, the Shuttle has a high termal pad. Since the model is fairly detailed, select this op-
inertia and it takes hours to reach an equilibrium state. tion only if you have the performance to spare (perfor-
This option allows to speed up thermal computations mance).
SIMULATION OPTIONS 19

SRBs shake view: The initial part of the launch is de- failure: This option will produce a callout when a failure
scribed as very rough by the astronauts. To convey this occurs and identify the failure. Experienced users will
feeling, this option shakes the view around while the be able to tell from the instrumentation what has hap-
SRBs are burning (for about two minutes after liftoff). pened.
Note that this makes cockpit controls hard to use, so
the option is best used only when launching under au- essential: The group of essential callouts contains infor-
topilot. mation that can not be communicated in a different
way because it has no real counterpart. For instance,
Enhanced displays: This option selects a special graphics in the real Shuttle the pilot knows which control stick
effect for the displays which will simulate lighting bet- he is touching, but a simulation user usually only has
ter — at night the display will be hard to read if set too one, and so the information to which simulated con-
bright, and during the day it will be hard to read when troller his joystick is currently connected is given via
brightness is set too low. In addition, a dust layer can callout. Essential callouts should not be disabled un-
be rendered onto the display (performance). less for video recording or similar.

Velcro visuals: If you want to see more realistic visuals of The next section allows to select the configuration of the
the ubiquitous velcro strips in the cockpit, choose this Shuttle. The options are:
option — it will apply a fur shader effect to the strips.
This may be rather expensive on lower-end graphics ET type: Historically, the external tank has undergone an
cards (performance). evolution to ever lighter variants with correspondingly
better ascent performance. This option selects the type
Display dust level: This slider is only active when the en- of ET to be flown.
hanced displays option is chosen. It determines how
much dust is seen on the screens (visuals). Tail cone: For early gliding tests, Enterprise was fitted
with a tail cone to reduce drag. This option allows
SM realism: The real Shuttle systems management dis- to use this configuration which improves the best glide
plays show out-of-limit information for the values they ratio from some 4.5 to 7 (only in gliding scenarios).
display. Values out of their normal range are marked
Payload: This option allows to select a payload carried in-
bright and commented with up (over limit) or down
side the payload bay. Some payloads can be handled
(below limit) arrows. By default this information is off
with the RMS arm.
as it makes the displays computationally more expen-
sive to render, but it can be switched on here (perfor- OMS kit: The OMS kit is an extra set of tanks which pro-
mance). vide more propellant which was planned for the Shut-
tle but have never been flown in reality. Three different
The second section defines how various voice callouts are variants with different quantity are available.
managed. Voice callouts are messages printed on-screen as
well (if the option is set) read aloud by the FG text to speech COAS: The crew optical alignment sight is a device a bit
engine. There are different kinds of messages that can be like a HUD which projects a targeting recticle onto a
separately enabled or disabled. combiner. It can be mounted in forward and rear po-
sition and is useful for both precision alignment using
help : Messages of this class give advice to a new user on stars as well as rendezvous operations.
what to do. They’re typically reminders that a proce-
dure needs to be performed now. allow/stow loose items: This checkbox allows to initialize
a simulation of loose items (currently a flight manual
info : Such messages report that a certain mission stage has folder) in the cockpit to visualize the effect of zero
been reached. An experienced user will be able to see gravity. These items will slowly drift around, respond
the same information from the instrumentation in-sim, to thruster firings and bounce off the cabin walls. Note
but it might be helpful to be reminded. that this costs some extra performance — and remem-
ber to stow all loose items before entry!
real: Standardized voice callouts which are used in the real
Shuttle cockpit are classified by this group. The last section configures the Earthview orbital renderer
and the behavior of co-orbiting objects. By checking the
limits: When this option is active, the limit system will box it is possible to start Earthview automatically when
produce warnings whenever the Shuttle reaches an the Shuttle passes a certain altitude and end it again once
aerodynamic, thermal or other limit that action need the Shuttle drops below that altitude where the slider se-
to be taken soon. lects that transition altitude. The orbital target simulation
20 CHAPTER 1. FIRST STEPS

distance governs at which point the simulation will change

from treating an orbital target as a mere coordinate position
to a full 3d model in the scene which is simulated in more
detail.

Note: The Shuttle has with some margin the most

complex systems and avionics simulation in FG, in
addition to having a highly detailed cockpit mesh
with several hundreds of animated switches and talk-
backs. Expect the need to drastically reduce details
and options on lower-end hardware.
Figure 1.7: A selectable scenario.

Save/resume and scenarios

create an interesting situation. In addition, it is possible to
Since a realistic mission duration is several days and it also add a user-defined scenario.
might not be feasible to keep the simulation running that The third option, Failure training , allows to create a
long, there are two options designed to work around this is- random set of failures at a specific difficulty level (easy,
sue - the save/resume capability and the scenario selection, medium or hard). This is intended to train failure proce-
both available in the dropdown menu. dures without actual knowledge (as it would be in a sce-
These options complement each other and are not de- nario) what kind of failure is going to occur. Currently fail-
signed to be mixed (this may or may not work). ure training is available for launch only, and is best selected
Save/Resume writes the current state of the simulation prior to liftoff.
to file such that the simulation can be ended and resumed
some other time. This allows to experience a longer mission
without having to be in front of the computer for the whole Mission control
time.
To properly resume, the same mission stage and the There’s a very simple implementation of a mission control
same mission definition file as used for the original state center (MCC) available in the simulation. At this point, the
that was saved should be used to resume. For in- function can be used to query the simulation for things that
stance, if the simulation of a rendezvous was started can’t be obtained via the Shuttle’s avionics. One example
using SpaceShuttle-mission and the mission file is the current orbiter weight which needs to be provided to
myMission.xml, ending the simulation and starting with the navigation systems for accurate OMS burns. In reality,
SpaceShuttle-launch, then resuming the state might mission control would keep track of this by knowing the
not work properly. orbiter weight at the beginning of the mission and by sub-
There are five different slots available which allow to tracting propellant usage, water dumps or payload released
type a short description (a timestamp is generated automat- into space. Using the Mission control functionality, this
ically). Each state can be resumed. Expect this to be imper- question can be posed.
fect, in particular not every system state and every switch Another very useful function is the deorbit solution re-
will be remembered. In particular, resuming a rendezvous quest. If the Shuttle groundtrack will pass reasonably close
target in close range is currently not supported. to a landing site, MCC can advise the crew on what burn
In contrast, the scenario concept is designed to quickly parameters to program.
access an interesting mission phase for training pur- However, mission control queries require that a com-
poses. Scenarios are specific for the mission stage munication channel to the ground is open, thus in order
that has been chosen, for instance the high or low to use the function, the antenna and communications sys-
energy atmospheric entry scenarios are only selectable tems of the Shuttle need to be functional and operated cor-
when SpaceShuttle-entry has been chosen, docking rectly. The current state of the comlink and the data rates
or payload capture scenarios are only selectable when for telemetry can be accessed in the menu as well.
SpaceShuttle-mission has been chosen before (see In addition, dialogs provide mission-control type infor-
Fig. 1.7) and so on. mation. For instance the trajectory map (Fig. 1.8) shows
A selected scenario may spawn an orbital target, place the current position of the Shuttle and the rendezvous target,
the Shuttle at a certain location, configure the Shuttle’s sys- the selected landing site and, dependent on type, alternative
tems in a certain way or cause specific failures for you to landing sites or ground communication stations as well as
deal with, or even sets a time at which a failure will occur. the current visual horizon. This makes it possible to quickly
As such, the concept is quite flexible and allows to quickly plan an emergency de-orbit to an upcoming site or find out
A LOOK AROUND THE COCKPIT 21

when the next S-band communication with the ground will If the cockpit detail option is set to ’high’, you will
be possible. In addition to the future orbits, the dialog also see many more panels in addition — then the complete
shows the location of programmed OMS burns. flightdeck including overhead (see Fig. 1.2) and rear panels
(Fig. 1.3 and 1.4) is shown. If the cockpit detail is instead
on ’low’ you will just see the lights and their associated
controls.
The additional panels contain many usable controls, but
generally they do not include controls for flight-critical sys-
tems but rather for on-orbit management tasks, thus you do
not need them during the phases where the Shuttle is actu-
ally piloted.

Figure 1.8: The trajectory map.

Another useful tool is the rendezvous-planning dialog,

which provides information on the relative position of the
Shuttle and a rendezvous target. This dialog changes de-
pendent on the distance to the target and as such provides
also information not readily available to MCC, such as the
relative state of the vehicles during docking. (The latter
function is provided since due to the rather rough 3d model
of the docking area, proper visual cues are difficult to ob-
tain).
Finally, the notepad is a feature to keep record of
mission-specific information. For instance, it registers the
attachment point of the currently selected payload, it keeps
Figure 1.9: A three position switch with a cover.
a record of the last five answers that MCC has transmitted
and it shows specific information that a selected mission or
scenario file wants to provide, for instance burn parameters
There are several types of control elements on the vari-
that should be entered or specific instructions.
ous panels. Most often you’ll probably encounter switches.
In general, the mission control feature should be consid-
Usually they are between guards to protect them against an
ered experimental. While in reality MCC has much more
astronaut accidentially bumping into them in zero g. The
computing power available than the Shuttle has aboard and
more critical ones also have covers which need to be re-
can determine e.g. burn parameters for a maneuver much
moved before the switch can be operated (see Fig. 1.9 for
more accurately, the simulated MCC has no access to such
an example).
features, and for any detailed planning or orbital maneuver-
ing, offline computation tools need to be used. Most switches have two positions, but don’t get compla-
cent, a fair amount has three. For two position switches
the labeling is usually top and bottom, for three position
A look around the cockpit switches the left/right side labels indicate what the switch
does in the middle position. Most are equipped with tooltip
Let’s start by getting familiar with the controls in the cock- context — if you hover with the mouse pointer over a
pit. Start FG with the Shuttle on the launchpad (or in orbit switch, you will get information on what the switch is and
— both give you time to look around as they don’t require what its current position is if you have tooltips activated in
any immediate action. the View Options.
By default, you find yourself in the Commander’s (CDR) Switches (and switch covers) in FG are moved through
seat on the left, whereas the pilot (PLT) is placed on the their valid positions by left and middle mousebutton clicks.
right. The general layout of the panels is shown in Fig. 1.1 Alternatively you can use the mousewheel up and down.
along with the names of the panels — try to memorize Another common element are pushbuttons. They usually
those, as we will later use them to describe where switches have a small light on them, a so-called talkback which tells
are located. You can press Ctrl + c to quickly see what their current state is. If you look at Fig. 1.10, they
which switches and buttons are functional and what their come in two different kinds. Grouped pushbuttons refer to
clickspots are. the same system. For instance, in the example above the
22 CHAPTER 1. FIRST STEPS

tigate them soon in some detail.

Figure 1.10: Two types of pushbuttons.

buttons on the right select between AUTO and CSS (manual

flight) in the pitch (and yaw/roll) channel. If you push the
pitch AUTO button, the CSS talkback light will go out and
the AUTO talkback light will come on to indicate that the Figure 1.12: A rotary switch and status lights.
pitch channel is now in automatic mode.
On the other hand, toggle pushbuttons like the SPD BK
There are also few rotary switches (Fig. 1.12). They work
(speedbrake) button on the left have two talkback lights —
by turning the white indicator into the desired position and
the upper will go extinct and the lower will illuminate when
are operated like switches with either clicks or the mouse-
you push the button, indicating that speedbrake control has
wheel. The figure also has an example for indicator lights
changed from automatic to manual.
(in this case for RCS jet firings) which are similar to talk-
backs, except they don’t illuminate in response to a pushbut-
ton or a switch but reflect the state of the Shuttle’s systems.
A rather common (but operationally rarely used) element
are circuit breakers (Fig. 1.13). They can be toggled from
pushed in (on) to pulled out (off) with left mouse button
clicks.

Figure 1.13: A circuit breaker.

Figure 1.11: A multifunction display unit (MDU). Finally, on panel C2, there are two keyboards. They con-
tain alphanumerics as well as command keys and are used
to communicate with the avionics. A key is used by simply
Pushbuttons are generally operated by left mouse clicks. clicking it with the left mouse button.
Just as switches, many are equipped with tooltips. Many critical systems are implemented in an arm-fire
The edgekeys of the multifunction displays (MDUs) scheme — switches or pushbuttons don’t do anything be-
(Fig. 1.11) are a special kind of pushbutton insofar as they fore the system is armed. Given that there are controls in the
don’t have talkback lights. Rather, the menu shown in light cockpit to detonate explosives to separate the RMS arm or
blue in the lower portion of the display changes directly in the antenna or to lower gear (which can never be retracted),
response to pressing an edgekey. it would be quite dangerous to set those systems off by ac-
The MDUs in addition have on/off knobs and brightness cidentially pushing the wrong button — the design of the
knobs. They are operated like switches, i.e. using left and cockpit controls tries to make very sure this happens only if
middle mouse clicks or the mousewheel. MDU brightness this is the intention of the crew.
can be changed in steps. Since the MDUs are the main win- Finding the right clickspot in the cockpit while you’re
dow into the function of the Shuttle systems, we will inves- trying to pilot the Shuttle through a critical situation can be
AVIONICS — THE MEDS LAYER 23

a challenge — which is why many major systems (includ- the attitude director indicator (ADI) ball in the center, an
ing the keyboards) are duplicated as GUI windows where accelerometer in the lower left, the horizontal situation in-
it’s obvious what needs to be clicked and all time-critical dicator (HSI) in the lower center and several mission-stage
operations have key bindings in addition. For instance, the specific instruments in the lower right. Particularly impor-
Shuttle gear is nominally lowered a few seconds prior to tant for piloting manually are the magenta elements — they
touchdown — at a time when you’re quite possibly busy are error needles, if they deviate from their zero-position,
initiating a flare. Clicking the cockpit buttons to arm and you need to maneuver as to ’fly towards’ them to bring the
lower gear works fine, but you can do it more quickly by Shuttle into the attitude suggested by guidance.
using the keys Ctrl + g to arm gear and Shift + g to The light blue part below the instruments is the MEDS
lower it. menu, identified by a text (above ’FLIGHT INSTRUMENT
MENU’). To the far left and right is technical information
which is not important for the moment, below are the menu
Avionics — the MEDS layer options, each above an edgekey that is associated with the
corresponding option. The MEDS is navigated by choosing
The vast majority of the communication with the Shuttle’s options via the edgekey.
systems is done via the avionics, there are very few analog Let’s try this — press the left-most edgekey (’UP’), then
instruments left in the cockpit, so before doing any space- ’SUBSYS’ and ’SPI’ — you should see the display of
flight, we need to discuss the basics of how to access the Fig. 1.15.
systems.
The front panels F6, F7 and F8 (see Fig. 1.1) are dom-
inated by the MDU screens. These are part of the multi-
function electronic display system (MEDS), to which also
three keyboards (only two accessible from the front seats)
and four integrated display processors (IDPs) belong.
The MEDS provides the primary instrumentation by
which you can fly the Shuttle and monitor the most impor-
tant systems. For that purpose, the IDPs have access to data
buses which feed sensor data from throughout the Shuttle
and convert that into a visual representation.
Let’s take a look around the MEDS. In front of you, you
should see the PFD (Fig. 1.14) somewhere (if you’re still
sitting on the launchpad, it doesn’t look like this yet — it
only changes upon engine ignition).

Figure 1.15: The MEDS surface position indicator (SPI).

In particular, note that the menu option has changed color

to white to indicate that it is currently selected. You’re now
looking at the Shuttle’s surface position indicator (SPI),
which shows what the four elevons, the body flap, the rud-
der and the speedbrake are doing. This display is most use-
ful to gauge trim in aerodynamic flight.
Change now to OMS — this brings up the status dis-
play of the orbital maneuvering system (OMS) and the main
propulsion system (MPS) shown in Fig. 1.16. This is used
during engine burns to verify that each engine reaches the
proper thrust and is safe to operate. You can also look at the
APU display which shows the status of the auxiliary power
Figure 1.14: The MEDS primary flight display (PFD). units and the hydraulic system.
Now go back to the main menu (’UP’) and select ’CON-
FIG’ — you should be looking at the display of Fig. 1.17.
This shows four tapes with (from left to right) Mach num- This screen shows the current configuration of the MEDS
ber, angle of attack, altitude and vertical speed, as well as system. In blue you see the IDP that is currently responsible
24 CHAPTER 1. FIRST STEPS

Figure 1.16: The MEDS OMS/MPS status display. Figure 1.18: A DPS display.

that looks positively antique and resembles a monochrome

cathode ray tube (CRT) screen (Fig.1.18). Which it actually
is — unlike the MEDS which was introduced when a glass
cockpit was fitted to the Shuttles, the DPS screens were part
of the Shuttle fleet from the beginning.
The DPS screens form a different layer inside the MEDS
display (the content of the old CRT screens is just ren-
dered into the upper portion of the screen). As a result,
DPS screens are not affected by the edgekeys and cannot be
changed by the MEDS. Instead, they’re configured by key-
board commands (and we’ll learn how to do that in the next
section).
You can take some more time to explore the MEDS
layer — though you have seen the most important screens
by now. Several menu options set the configuration (e.g.
Figure 1.17: The MEDS configuration display. PORT SEL), others are responsible for fault message han-
dling (MSG ACK and MSG RST) and don’t do anything
unless a fault occurs, yet others refer to options which are
not implemented because they make no sense in the context
for generating the graphics of the screen you’re looking at of the simulation (CST or MEMORY — the latter is used
in the middle row, and above all displays that are also un- to patch the IDP memory, which doesn’t exist in-sim as the
der the control of that IDP. You can also see the names for IDP doesn’t run on the original software in an emulator).
the displays — for instance the left-most display is called
CDR1, the center display is CRT3 and so on. Further down
are the analog-digital converters (ADC) responsible for pro- Avionics — navigating the DPS layer
cessing the sensor data for the IDPs.
It’s very clear from the display that there is a separation
Note: In the following, it is assumed that the simpli-
— not every screen uses the same processor, but rather in-
fied DPS simulation option is chosen. With the de-
dependent processors drive different screens. The reason is
tailed option, several additional procedures may be
redundancy — in case any of the three IDPs fails, the cock-
required.
pit remains functional (the same is true for electricity —
all MDUs and IDPs are distributed evenly across the three The data processing system (DPS) is really where most
main power buses such that a loss of power on one bus never of the interaction with the Shuttle’s systems happens. The
disables the whole avionics). avionics software runs on five general-purpose computers
Now let’s go back to the main menu and try the option (GPCs), and there are two versions of it, coded by indepen-
’DPS’ (data processing system). This leads to something dent teams to make sure no bug can completely disable all
AVIONICS — NAVIGATING THE DPS LAYER 25

avionics functions. The Primary Avionics Software System To understand that, look again at Fig. 1.17. Each IDP
(PASS) is what normally controls the Shuttle, the Backup controls various screens. A keyboard talks to one of the
Flight System (BFS) runs in the background, standing by to IDPs. So anything you type will affect every screen that
take over in the case of a severe problem in PASS. shows DPS and is controlled by the IDP the keyboard is
To avoid a hardware malfuction disabling the avionics, linked to.
the PASS is usually run on several GPCs simultaneously That’s hard to keep track of — which is why there’s a
(in a ’redundant set’) and the results compared. If one GPC reminder. Look at the botton of the DPS screen — there’s
yields wrong the results, the others will over-vote it and as a a little box with a number in it. The number is the IDP
result it will be isolated. A nominal configuration has PASS controlling the screen. If the left hand keyboard is currently
on four GPCs and BFS on the fifth. talking to the IDP, there’s a red bar extending left of that
Now, while BFS is written to offer limited functionality box. If the right keyboard is talking to the IDP, there’s a
but fit completely into GPC memory, PASS offers more op- yellow bar extending right (the two keyboards can actually
tions but only part of it can be run at any one time. Thus, talk to the same IDP). So if you look at a DPS screen and it
the software memory configuration needs to be changed for shows a red bar at the bottom, you know that whatever you
every Shuttle mission phase. type on the left keyboard will affect that screen.
PASS software is divided into Major Functions — the Now — how to change that to be able to control a partic-
most important ones are Guidance, Navigation and Control ular screen from a keyboard?
(GNC) and Systems Management (SM) (in addition there is
in theory payload (PL) which however is not really used).
In addition, GNC is subdivided into Operational Sequences
(OPS). Here, OPS 1 manages ascent, OPS 2 on-orbit oper-
ations, OPS 3 entry and landing and OPS 6 supports launch
aborts (other OPS are used for testing and ground check-
out and are of no concern here). A GPC can run only one
OPS at any given time (except OPS 1 and 6 which reside in
memory simultaneously).
The DPS screens provide the interface to the currently
running software, and they’re further subdivided into Ma-
jor Modes (MM). For instance MM 101 is active pre-liftoff
on the pad, MM 102 is for the first stage to SRB separa-
tion, MM 103 is the second stage ascent, MM 104 is used
for a possible first OMS-1 burn to target apogee and so on.
Figure 1.19: IDP select switches on panel C2.
The sequence of major mode screens typically mirrors the
sequence of events during the various mission stages.
In each OPS, major modes are augmented by additional Look at the center part of panel C2 (Fig. 1.19). There
interface pages — special functions (SPEC) and displays are two switches labeled LEFT CRT SEL and RIGHT CRT
(DISP) — the difference being that displays only show in- SEL — the left one allows to switch the left keyboard from
formation, special functions allow to enter commands in ad- IDP 3 to IDP 1 and back, the right one to switch the right
dition. These can be called ’on top’ of the current major keyboard from IDP 3 to IDP 2 and back (remember, in the
mode page which can be resumed. old Shuttle cockpit there were CRT screens instead of the
Take a look again at Fig. 1.18. In the upper left corner, MEDS and the IDPs — they just kept the label). So you
you can see the number 1011. This indicates that this is can’t access every screen from every keyboard, but you have
a major mode display for MM 101 (the trailing 1 is irrel- some limited control.
evant). Right to it is the name of the page — LAUNCH (For ease of use, Data Processing System → Keyboard
TRAJ in this case. To the far right is another 1, followed provides a virtual keyboard pop-up window, and via
by a time in days, hours, minutes and seconds. The time is, Data Processing System → Display Processor that can be
dependent on setting, either UTC or mission elapsed time switched to every IDP).
(MET). The number stands for the GPC that generates the Now pick an MDU of your choice, use the edgekeys to
display (remember, there are usually several running PASS, bring up the DPS and use the selectors to connect a key-
and only one of them produces a given DPS page). board to it — verify by the appearance of the bar. We’re set
To interact with the DPS, we need to type keyboard com- to communicate with the DPS.
mands. But what keyboard — there’s two keyboards on C2, The main way to do that is to type command sequences.
but nine screens in front of us — which keyboard refers to For instance, we can bring up special function page 50 by
which screen? typing
26 CHAPTER 1. FIRST STEPS

SPEC 50 PRO At this point, your head might start to spin. Probably you
(’special function 50 proceed’). While you type, you see can’t readily make sense of what’s visible on, say. GNC
the commands appear on every DPS screen connected with SYS SUMM 2, in order to call up a page you somehow have
the keyboard on the so-called scratch line. The command to remember its number (there’s several dozens of them),
isn’t actually executed till you press the closing key (PRO but you can’t always call all pages because you might be in
in this case), so you can go back and delete parts of it if you the wrong software mode, and before anything can be typed
spot a mistake using the CLEAR key. Only after closing you have to verify that your keyboard is talking to a partic-
the command, the page changes. ular screen? And then you’re supposed to pay attention to
You get to a page named HORIZ SIT and you will note what screen is connected to what computer? Seriously?
that the number in the upper left changed to 1011/050 — You’re in good company. Astronaut John Young, Com-
which means that you’re still in major mode 101 and that mander of Columbia’s maiden flight actually said: ’What
the corresponding page is ’underneath’ the special function we have in the Shuttle is a disaster. We are not making
you’re currently looking at. You can get back by pressing computers do what we want. (...) We end up working for
the key RESUME (or RSM on the virtual keyboard). the computer, rather than the computer working for us’
Now type Unfortunately that’s exactly how the Shuttle DPS works.
SPEC 18 PRO The DPS pages are not well structured in that they would
this calls up a page named GNC SYS SUMM 1 and the separate relevant from irrelevant information, or that the
number in the upper left changes to 1011/ /018. The same symbols would usually denote the same things — they
reason this does not read 1011/018 is that you’re looking at just present a complete information overload and you have
a display page — SPEC is used both for calling up special to know them by heart to know where to look at.
function pages and display pages. Like a SPEC, you can It’s how reality was. You have to bite the bullet and learn
resume a DISP with a single keystroke command. how to navigate the system, and we’ll cover the content of
Now open first SPEC 50, then SPEC 18 without resum- the DPS pages one by one.
ing SPEC 50 first — the number in the upper left reads
1011/050/018 — meaning that both SPEC 50 and OPS 101
are underneath your current page and you can go sequen-
Entering items into the DPS
cially back by pressing RESUME. If you haven’t done so, go back to OPS 101 as appropriate
There are some display pages you might want to call for a Shuttle on the launchpad (you may have to re-start
up rather frequently, among them system and fault sum- your session). Then call up SPEC 50 again. If you look
maries. The single keystroke command SYS SUMM cy- closely, you can see that some entries are associated with
cles through the two summary displays for the current major a number, for instance there is ALTM 9 29.92 in the upper
function (there’s two summary pages for GNC and two for line left of the center, or 41 TAL SITE 6.
SM) whereas FAULT SUMM brings up the fault summary These are how the information is referenced and changed
page — so you don’t need to reference them by number if — one number is an item, the other is its value. The value
you need to. Technically they’re all DISP pages, so you can of item 9 is 29.92 (that’s the pressure at sea level for the
select them just like any others. altimeter). The value of item 40 is 6 (that’s the index for
If you want to change Major Mode or OPS sequence, all RAF Fairford airbase as an abort landing site).
you need to do is press OPS instead of SPEC and key in the The command sequence to change the value of an item is
number of the major mode. If you need to call a different ITEM (number) (delimiter) (value) EXEC where ’+’ and
OPS sequence, you always reference it by the first Major ’-’ act as delimiters (and at the same time set the sign of
Mode page. the value). Thus, if you want to change air pressure for the
Let’s try this (note that in reality we’d never do this on altimeter to 30.01, you need to type
the launchpad!). Say we want to bring up the second page ITEM 9 + 30.01 EXEC
of OPS 2 (on-orbit operations) — that’d be OPS 202. Type (again, if you type something wrong, you can always go
OPS 202 PRO. back the last keystrokes using CLEAR). If you want to se-
You’ll get an ’Illegal entry’ message as a reward for your lect Banjul as abort landing site, you type
trouble (you can clear it using the MSG RESET key). ITEM 41 + 7 EXEC
Why? Because we are in OPS 1, the requested page is in — of course this assumes you know that the code for
OPS 2, so we first need to get to OPS 2. Thus, we first need Banjul is 7, so there’s that. We already agreed that this is a
to do mess, so let’s not talk about it further.
OPS 201 PRO It you try to set an item that doesn’t exist or that isn’t
to load the OPS 2 software, and then use implemented in the FG simulation, you will get an ’Illegal
OPS 202 PRO entry’ message — same as if you try an item that isn’t avail-
to get to the page we actually want. able in the current OPS sequence. Don’t try to set random
THE ADI BALL 27

values in order to see what they do, only access items where The first of these allows to select the coordinate system
you know what they’re for. the ADI displays. The switch selects between REF, LVLH
There’s a different kind of items as well — they just have and INRTL. The first one is implemented during launch and
a number associated with them but no value. They usually entry to be pretty much what systems aboard an aircraft
either execute a command based on parameters other items would show — the ADI shows heading, pitch and roll when
set or toggle a function. On SPEC 50, for instance items used in this mode with the horizon line clearly marked. On
34 and 35 (ABS and DELTA, lower right portion of the dis- orbit, this is not useful, so while the Shuttle is in OPS 2, this
play) are of the toggle type. You’ll see a star next to item 35 mode shows instead so-called starball coordinates which
indicating that this setting is active, after you toggle using correspond to right ascension and declination, i.e. the sys-
ITEM 34 EXEC tem is suitable e.g. look up stars by their coordinates.
you see the star move over, indicating that the other set- The LVLH (’local vertical, local horizon’) option is al-
ting is active. ways fairly similar to what aircraft show, except that instead
Item 16 (next to LOAD) on the right is a way to execute of heading it displays sideslip angle β — thus during pow-
a command — to load all the settings above. ered flight or during entry when the spacecraft body axis
These are the basic techniques to talk to the DPS, and points along the velocity axis, it always shows zero.
we’ll utilize them later in concrete examples when describ- For a spacecraft in orbit, orientation with respect to the
ing how to fly an actual mission. local horizon may not be all that useful, and thus the last
switch position option selects inertial attitude (which is con-
ceptually similar to starball). In this coordinate system, the
The ADI ball Z-axis is defined by Earth’s rotation axis. You’ll notice that
this as well as the starball coordinate systems do not rotate
Now, take a look again at the PFD and in particular the ADI the ADI ball if the spacecraft is freely drifting with rotation
in the upper center. This is a fairly special instrument, be- rates nulled.
cause it displays rotations in three axes. You might be used The currently selected coordinate system is shown on the
to the horizon display of an airplane. The Shuttle’s ADI upper right of the PFD next to the text ATT.
ball is something similar, but unlike an artificial horizon, The second switch has to do with the pink error needles
it really displays three axis rotations (the reason for this is — it selects how many degrees errors are represented by
that aircraft cannot usually sustain large sideslip angles, i.e. the full needle deflection (the number is shown close to the
body axis needs to coincide with the velocity — this is dif- right scale. The switch allows to select between 10, 5 and
ferent for a spacecraft). 1 degrees. This is important when you’re trying to pilot
manually following the ADI needles.
Finally. the last switch has to do with the rate indicators
(the three scales with light blue triangles framing the ADI
ball). They show the current pitch, yaw and roll rates, and
the switch allows to select what rate should be represented
by a full arrow deflection. The selectable values are again
10, 5 and 1 degrees per second and the currently selected
option is displayed next to the scales.
Given the limited view angle from the cockpit windows
and the lack of any reference cues in space, the ADI is quite
an important instrument and understanding how to use it
properly is crucial for flying the Shuttle.

The head-up display

Figure 1.20: The ADI ball in action. One of the most prominent displays in the cockpit is the
head-up display (HUD) overlaying the forward view from
the cockpit. Its main function is to aid piloting the Shuttle
The ADI is also special in that there are hardware during the late TAEM and approach/landing phase which is
switches on panels F6 and F8 which influence its function. always flown manually. In reality it has little operational
On the Commander’s side, they’re below the CDR 2 MDU use during other phases of the flight, but of course it still
under the heading ADI and read ATTITUDE, ERROR and displays limited earth-relative attitude information during
RATE. ascent and in orbit.
28 CHAPTER 1. FIRST STEPS

place. Usually times specified to the avionics are Mis-

sion Elapsed Time (MET), which is counted from zero at
liftoff. Alternatively, timers may show Greenwich mean
time (GMT) — for a spacecraft, there’s usually little point
in showing local time, though that can be accomplished.

Figure 1.21: The head-up display and its controls.

When used in any other mission phase than ap-

proach/landing or when it does not have guidance infor-
mation, it displays TAEM symbology but with all specific
strings blanked and the guidance diamond non-functional
(see chapter 6 for an explanation how the HUD is opera- Figure 1.22: Event timer controls on panel C2 and timer on
tionally used). The only really useful information it dis- F7.
plays in orbit is part of LVLH attitude (unlike the ADI ball,
it only displays pitch and roll, not yaw). During launch and All DPS pages show time in the upper right corner —
entry, it will also display airspeed and altitude as long as they default to GMT, though it can be changed to MET us-
they are within the display ranges. ing the time utility display on SPEC 2 (which, however, is
The HUD is switched on and off via the power switch only available in OPS 2). In addition, the DPS pages show
on the lower right. Brightness can be set either via the au- the CRT timer below the current time — this can be set by
tomatic switch (right) or the knob in the center below the the software and asked to count down to zero or up to some
HUD. The most important control is the mode switch on value and also throw an alarm when the timer has expired.
the left hand side. Each time the switch is pushed down, However, this also requires SPEC 2.
it increases the de-clutter level. At level 0, all symbol- When the CRT timer is not available, there is also the
ogy is shown, at level 3 only the boresight, after which event timer on panel F7 with controls on C2 (see Fig. 1.22).
the next declutter attempt cycles back to full display. Since The timer is set by dialing a time with the buttons in
de-cluttering is part of approach operations, there’s the keythe lowest row, followed by flicking the ’set’ switch (right).
binding shift + h to do the operation in a hurry. This copies the set time to the digital counter on F7. Timer
During the last phase of the flight, it is a valuable help for
control ’start’ initiates the counting, ’stop’ ends it. With the
the crew, as it can for instance display a virtual runway in mode switch, it is possible to choose whether the display
case clouds obscure the view by the time the final approach should count down from the set time to zero or count up
is begun. In addition, it displays visual aim points and guide
from zero to the set time. The same timer is also displayed
triangles which help to get onto the right glideslope and to on A4 in the aft section of the cockpit (however, note it can’t
initiate the pullup at the right time. be set there).
The software timers work in principle the same way, ex-
cept that rather than with switches, they are operated by
Timers item entries on the keypad.
MET is also available on two timers in the cock-
For many purposes, it’s important to know the current time pit — one located overhead above the pilot’s con-
or specify to the avionics at what time an event should take sole, one located in the aft section on panel A4 (see
OVERHEAD ANALOG GAUGES 29

The gauges and readouts need to be studied with care, be-

cause they are all multi-function devices. Below each group
of gauges (and graphically connected via a white bracket) is
a switch or rotary switch which determines what the gauges
show.
For instance, the digital readouts can be set to display re-
action control system (RCS) propellant, RCS oxidizer, or-
bital maneuvering system(OMS) propellant, OMS oxidizer
or the lowest value of all available. Failing to realize that
the knob is set to display OMS while trying to e.g. assess
whether there’s sufficient propellant in the RCS tanks for
entry or whether cross-feeding needs to be done can have
fatal consequences.
As with the avionics, the way the information is pre-
Figure 1.23: Event and mission timer on panel A4. sented is confusing and prone to mistakes, but unfortunately
this is how the real Shuttle works.

Figs. 1.23 and 1.24). Alternatively you can get MET from
Mission control → MCC status . Night operations
With an orbit taking 90 minutes, half of which is during
Overhead analog gauges darkness, it’s as good as impossible to avoid operating the
Shuttle during the night, although launch and touchdown
Located on the overhead panels directly above the front can of course be planned during the day. There are multiple
windows, there is a prominent set of analog gauges as well lighting options for the fligth deck available, as well as the
as a few digital readouts (see Fig. 1.24). The gauges pri- payload bay floodlights.
marily deal with the environment system and show temper-
atures in the cabin and storage tanks, gas flows or storage
tank pressures.
While the correct functioning of the environment system
is of crucial concern for a real Shuttle crew, in the context
of the simulation it can be to some degree neglected, espe-
cially on the first few flights (by and large, it will run fine
by itself).
Of greater importance however are the three digital read-
outs displaying the remaining percentages of RCS and OMS
oxidizer and propellant for each of the three modules (for-
ward, aft left and aft right).

Figure 1.25: The overhead lighting controls.

The cockpit lights are controlled from overhead panels

(Fig. 1.25).
There are panel backlights, making the labels of the pan-
els visible at night, instrument lights providing illumina-
tion to the analog instrument faces (there aren’t that many
left in the glass cockpit variant of the Shuttle though), the
glareshield flood lights providing illumination specifically
Figure 1.24: Gauges on the overhead panels. for the lower panels and the cabin floodlights mounted in
the ceiling, providing illumination for the whole cabin.
30 CHAPTER 1. FIRST STEPS

All of these lights can be dimmed individually, but for in-

stance the glareshield light switches are three-way switches
allowing to set the light to full bright and so by-pass the dim
function.

Figure 1.26: Glareshield and panel backlighting for night

operations.

In addition to the lights, also all MDUs and the HUDs

have a brightness selector, thus if the cockpit is kept largely
dark during the night portion, screens can be dimmed ac-
cordingly.
Unlike the incandescent or fluorescent flightdeck lamps,
the payload bay floodlights are metal halide lamps, i.e. they
take a while to ramp up to full brightness after switching
them on, and after switching them off, they should be al-
lowed a ten minute period to cool off before re-activating
them.
The Shuttle does not have external landing or taxi lights
(they would need to be mounted in awkward positions as the
leading wing edge receives the most fierce heat load during
entry). However, at selected landing sites, exterior night
lighting is implemented, i.e. the touchdown zone will be
illuminated by powerful floodlights. Nevertheless, expect
night landings to be more difficult since many visual cues
which can be used during the day are not available at night.
Now, enough looking around — it’s time to get this
spacecraft off the pad!
Chapter 2

Launch

General overview from the center of Earth increases and eventually, when or-
bital speed of about Mach 27 is reached, reaches the same
In reality, a spacecraft is usually launched into an orbit to magnitude as gravity. At this point, the engines can be
do something — rendezvous with another orbiting object switched off and since no net forces act, the Shuttle floats
for instance, or launch a satellite. Thus, it matters into in weightlessness.
which orbit the Shuttle is launched, and for this not only The ideal launch thus brings the Shuttle out of the at-
the course flown during launch but also the precise timing mosphere quickly and then accelerates it to orbital speed.
of the launch is crucial (and a missed launch window usu- The longer the launch trajectory passes through the atmo-
ally means a long delay before the next occasion presents sphere, the longer drag can act, leading to loss of perfor-
itself). mance. Thus spacecraft launch near-vertical to get out of
We will discuss orbital mechanics in some detail later, for the dense atmosphere as quickly as possible but then pitch
the moment let’s deal with a simpler problem: We’d like to down once they get higher. There is no additional bonus for
launch the Shuttle into an orbit with a given inclination such getting higher once out of the atmosphere — in fact getting
that the apogee reaches to the altitude at which we’d like to too high during launch is just a waste of propellant, because
circularize the orbit. once at orbital speed, it’s very cheap to add altitude. In
general, there is very little in terms of propellant reserves
Inclination is one of the orbital elements. It’s the an- available — too much wasted, and you won’t make a stable
gle under which the orbit crosses the equator, and orbit.
also the maximal latitude the orbit will ever reach.
From the latter, it follows that the inclination of the Once the launching spacecraft is outside the atmosphere
orbit must always be greater than the latitude of the and pitches down to build up speed, some component of
launch site. thrust still needs to be vectored downward to prevent it from
losing altitude. Again, the longer the race to orbital speed
One can understand most of what happens during launch takes and the more thrust needs to be vectored down, the
from a balance of forces. The first of these is gravity — to a more propellant is wasted. To visualize this, picture an ex-
good approximation, gravity accelerates everything down- treme case in which the Shuttle just vectors thrust down-
ward towards the center of Earth. In order for the Shuttle to wards, hanging at 150 km altitude, such that upward accel-
not fall from the sky, gravity needs to be balanced by other eration just about compensates gravity. While under thrust,
forces. Thrust from the rocket engines is the most important the Shuttle won’t fall down but it will use up all its propel-
one for the first phases of the launch. Since the acceleration lant (the situation might seem contrieved at this point, but
of the combined main engines (SSMEs) and solid rocket is actually pretty close what happens during contingency
boosters (SRBs) is about three times larger than the gravita- aborts following a two engine failure). Thus, a real launch
tional acceleration, the launch stack lifts speedily from the trajectory is a compromise between getting out of the at-
pad and accelerates rapidly upward. mosphere fast (prefering a vertical ascent) and getting up to
Aerodynamic lift and drag are also important forces. As speed fast (preferring a more shallow trajectory).
the Shuttle pushes through the lower dense atmosphere, There are yet more things to consider. One of them is the
friction with the air tries to slow it down. At the same performance of the craft. The Space Shuttle is capable of
time, the wings generate lift, however since the orbiter is strong acceleration up to almost three g initially as long as
mounted vertically on the launch stack, the lift doesn’t help the solid rocket boosters (SRBs) are attached, but once they
to counter gravity but tends to bend the trajectory. disconnect, the main engines reach not even one g (i.e. the
Finally there’s the centrifugal force — as the Shuttle Shuttle can’t even accelerate upwards at this point!) — only
moves faster and faster, the centrifugal force pushing away when the propellant reserves deplete and the external tank

31
32 CHAPTER 2. LAUNCH

Figure 2.1: Schematic picture of the forces acing during the first stage of a launch

mass reduces, the thrust reaches in excess of 3 g towards the Fig. 2.1 summarizes the above discussion for the early
end. stage of the launch. The dominating forces are gravity
This pronounced drop in acceleration needs to be fac- and thrust, most of the thrust is pointed upward, but some
tored into launch planning — at SRB separation, there smaller component is already accelerating the Shuttle into
needs to be enough vertical speed reserves that the Shut- the direction of the launch azimuth. Lift and drag are minor
tle does not fall back into the atmosphere till SSME thrust forces, acting to bend the Shuttle out of the trajectory and
grows strong enough to hold the altitude. Indeed, for more to reduce thrust.
than two minutes after SRB separation the Shuttle follows Let’s take a break here to explain important concepts re-
a vertical near-ballistic path, like a stone that is thrown con- lated to what we might mean by ’velocity’. Prior to liftoff,
tinuing upward yet is constantly losing vertical speed to the the Shuttle sits motionless on the pad, i.e. it has zero earth-
pull of gravity. relative velocity. However, Earth itself rotates eastward -
The final consideration are aerodynamic forces. In terms and at the latitude of Kennedy Space Center, that corre-
of ascent performance, they are perhaps not a major fac- sponds to an eastward motion of about 400 m/s.
tor, but the stresses caused by drag forces during launch are An orbiting body in turn knows nothing about Earth’s ro-
more than twice as high as during a nominal entry. In fact, tation, it just feels Earth’s mass. That means that orbital
the peak forces are almost strong enough to rip the wings speed is measured in a non-rotating frame, i.e. it is an in-
off the Shuttle. It is therefore crucial to steer the Shuttle in ertial velocity. Thus, the Shuttle sitting on the pad actually
the right attitude through the region of maximal dynamical has already a good 400 m/s of inertial velocity for free al-
pressure (max q) — the ascent must be flown in heads-down ready — which we can utilize by launching eastward. Or,
attitude — and even then the main engines need to be throt- if we would launch westward, we would have to accelerate
tled down to 67% of rated power for half a minute to limit to more than Mach 1 to even reach zero inertial velocity!
the stresses. Dynamical pressure rises with airspeed and Which is why launches are always done eastward to utilize
with air density — as the Shuttle climbs, density eventu- Earth’s rotation and save propellant.
ally wins out and q drops to zero. In practice the dangerous Now, the atmosphere is rotating along with the planet
region is from about 30 to about 60 seconds into the flight. (otherwise the surface would be blasted by a super-storm),
LAUNCH USING AUTOPILOT 33

Figure 2.2: Schematic picture of the forces acing during the second stage of a launch

thus while the Shuttle is ascending through the atmosphere, thrust component.
ground- (and air-) relative velocity is more important than It follows that during the second stage burn, as cen-
inertial velocity, whereas above the atmosphere only iner- trifugal force ramps up thrust is gradually vectored more
tial velocity matters. Consequently many instruments show and more horizontal and towards the main engine cutoff
ground-relative velocity prior to SRB separation and switch (MECO), the nose of the Shuttle goes below the hori-
to inertial velocity afterwards. zon when in inverted flight to accomplish that. MECO is
Let’s now take a look at a later stage of the launch, some reached when the centrifugal force is strong enough to carry
two minutes after SRB separation (Fig. 2.2). The spacecraft the Shuttle without any thrust.
is now well out of the atmosphere, and centrifugal force is Control during the launch is provided by direct thrust
building up. Altitude is maintained by vectoring thrust such vector control (TVC). Both SRB and SSME exhausts can
that the combination of centrifugal force and upward thrust be vectored around their mounting axis, allowing to cre-
component balances gravity, and the horizontal component ate strong pitch, yaw and roll moments (in fact, thrust vec-
of the thrust is accelerating the spacecraft. toring is so powerful that the resulting radial accelerations
Note that the SSMEs are actually mounted at an an- for rapid maneuvers are capable of overstressing the launch
gle. The reason is that the thrust axis needs to be vectored stack). The engine gimbaling is driven by rate controllers,
through the Shuttle’s center of gravity — and with the ex- so the handling characteristics in manual flight are excellent
ternal tank mounted, that is about in the forward part of the — fast and crisp responses to any control movement can be
tank (the much heavier oxygen is foward in the ET, the hy- expected.
drogen towards the rear, which makes the ET a top-heavy
contraption — if the order of the tanks were reversed, it’d
be impossible to vector thrust through the center of grav- Launch using autopilot
ity). This is important to remember — if you’re flying heads
down with the nose pointed perfectly horizontal, you have Let’s now see it all come together and launch the Shuttle un-
an upward thrust component, conversely if you’re flying der autopilot (AP) guidance. Start at Kennedy Space Center
heads up with the nose at zero pitch, you have a downward (KSC).
34 CHAPTER 2. LAUNCH

To present the above in somewhat more detail: Gravity exerts a force proportional to the current mass m(t) of the launch
stack, i.e.
FG = m(t)g(R)
where g(R) = GM/R2 is the gravitational acceleration at the distance R from the center of Earth with G the gravitational
constant and M the mass of Earth. At the altitude the Shuttle can reach, it is always close to 9.8 m/s2 . The centrifugal
force is proportional to the square of the inertial velocity vI , from which follows that it grows rather rapidly towards the
end of the burn:
FC = m(t)vI2 /R
The thrust of a rocket engine is proportional to the exhaust velocity and the rate at which propellant is expelled and comes
out as
dm
FT = −vE
dt
Finally, the magnitude of the lift and drag forces is
1 2
FL(D) = ρv ACL(D)
2 A
where vA is airspeed, A a reference area, ρ the air density and CL the lift coefficient. The combination q = 12 ρvA
2
is known
as dynamic pressure. Since air density drops with altitude as ρ = ρ0 exp(−h/7.5km), it is clear from this formula why
during a launch dynamic pressure first increases and then decreases. The vector components of the forces are arranged as
shown in the figures. Note that the resulting accelerations are a = F/m(t), thus centrifugal and gravitational acceleration
are independent of mass, whereas lift/drag and thrust accelerations grow as the ET mass depletes, hence the need to
throttle back engines close to MECO.
Orbital speed (for a perfectly circular orbit) is reached when FG = FC , i.e.
r
p GM
vI = g(R) · R =
R

--lat=28.6206 --lon=-80.6133 --heading=270 ment in targeting a launch into a given orbit is timing — but
will get you to the approximate position of Launch Com- we won’t bother with this right now.
plex 39. The first thing we need to do is make sure the Next, verify that the autopilot is engaged. Take a look
AP knows what we’re trying to do. Of course in reality all at the pushbuttons on panel F2 (see Fig. 1.10). PITCH,
this would be i-loaded into the Shuttle’s computers during ROLL/YAW and SPD BK/THROT should all have their
the preparation of the mission (and indeed, a set of instruc- talkbacks on AUTO — if not, push the buttons till the talk-
tions is automatically loaded with each mission file). Open backs change. Verify that the upper left corner of the PFD
SpaceShuttle → Launch Guidance in the menu. shows DAP: Auto and Throt: Auto. The Shuttle is now
under automatic controls. If you ever want to control pitch,
Note: If you choose a different launch site, you need yaw/roll or throttle manually, press the CSS pushbuttons
to make sure to orient the Shuttle on the pad such (CSS stands for control stick steering) and make sure the
that the tail fin points into the eastern hemisphere, throttle button talkback shows MAN.
i.e. pass a heading parameter westward — otherwise
To recap what we just did — guidance tells you where to
launch guidance might not acquire the correct launch
go, the autopilot steers the vehicle to where guidance says it
azimuth.
should be. You can have guidance active but fly manually,
Dependent on what your mission file is, it might read ac- then you will get indicators where to fly to. If you have no
tive already — if so, cancel it, then set inclination to the guidance active, the AP does not engage and you will not
lowest possible value (that will be the launch site latitude, get any fly-to indicators.
some 28 degree for KSC) to get a due east launch, leave The next task is to start the auxiliary power units (APUs).
target apoapsis at what it is, press activate again and verify They are used to supply pressure to the hydraulic system,
that the label reads active. Launch azimuth should read 90 which is needed to gimbal the engines and move the engine
degrees (i.e. eastward). Guidance now knows where you valves, so we can’t fly without them running (the engines
want to go. As mentioned above, the second crucial ele- simply won’t ignite). However, the APUs consume a lot
LAUNCH USING AUTOPILOT 35

Figure 2.3: The MEDS APU/HYD status display.

of fuel, there’s hydrazine aboard to keep them running for

perhaps 90 minutes in total, and we’re going to need them
during entry, so we start them only a few minutes prior to
launch and need to switch them off soon after MECO. Also,
they produce lots of heat and need to be cooled, so when- Figure 2.4: The APU startup sequence on panel R2.
ever we use the APUs, we also need to activate the water
spray boilers which take care of the excess heat or they will
overheat and die.
First, let’s bring the APU/HYD MEDS display on one of (4) Open the boiler N2 valves to allow pressurized nitro-
the MDUs to monitor what is happening (Fig. 2.3). This gen to push cooling water into them
display shows (from left to right and top to bottom) fuel (5) Check the talkback — it should now be grey rather
quantity, fuel pressure, cooling water quantity, APU tem- than stripe-pattern (’barberpole’), indicating that the
perature, hydraulic fluid quantity and hydraulic fluid pres- APU is ready to start. If the APU is not ready, check
sure. the previous steps and verify that each switch is where
Starting the APUs is a moderately complex procedure — it should be
but Flightgear offers in-sim checklists to help you remem-
ber it. Open Help → Aircraft Checklists and select ’Pre- (6) Put the start switches into start/run position. Look
Launch’ — you’ll see the tasks not ready in red, and what- at the APU MEDS screen and observe how hydraulic
ever is ready in green. pressure climbs to 690 (pounds per square feet) as the
Change to the pilot position ( Shift + v in case you turbine reaches operating RPM.
don’t remember how to cycle through views) or to panel
(7) Once the turbine has reached full RPM, engage the hy-
view, then look at panel R2 (Fig. 2.4).
draulic pump fully by switching main pump pressure
Now do the following (for all three switches of a group, to ’NORM’. This puts more load onto the turbine (and
each controlling one APU): hence the turbine won’t start unless the pump is set to
low). Verify on the MEDS display that pressure now
(1) Switch APU controller power on — this connects the reads around 3000.
turbines with power and activates their digital control
unit Both the checklist items and all bars on APU/HYD
should now be green — Atlantis is ready for liftoff.
(2) Open the fuel valves — this allows hydrazine to flow Now would be a good time to arrange the displays to your
from tanks to turbines needs — having the PFD for ascent/entry, the APU/HYD,
the OMS/MPS MEDS page and the DPS LAUNCH TRAJ1
(3) Switch the water spray boiler controller power on — (OPS 101) is what you need at this point.
this connects the cooling units to electrical power and In reality, the crew can’t initiate main engine ignition,
activates their digital controller this is done by launch control. Since the simulation does
36 CHAPTER 2. LAUNCH

Figure 2.5: The important instruments during the first stage of the launch.

not include a detailed pre-launch simulation, there is a key sounds like a velocity measure, for a spacecraft it actually
binding to initialize ignition: Press Shift + i to ignite all hardly ever is. Instead, you can think of it as a measure
three main engines. for aerodynamic forces — it really reflects the dynamical
If all is well and they ramp up to at least 90% of pressure of the air streaming by. Somewhere beyond 450
rated power during the first three seconds (you can mon- KEAS, the Shuttle gets into a dangerous region where the
itor the current power, or rather chamber pressure, on the forces might rip the connector struts to the ET off. Thus,
OMS/MPS display), the solid rocket boosters will automat- keeping an eye on the KEAS box is a good idea in the at-
ically ignite as well, at which point liftoff happens. mosphere.
Alternatively it is also possible to program an automated On the lower HSI, pointers of current inertial and earth-
countdown and ignition in the mission file (see appendeix relative speed are shown in a compass. For a due-east
13). If that is done, the simulation provides countdown mes- launch, they should be very similar and the inclination error
sages and automatically ignites the engines at the correct (∆inc) very small — for high inclination launches they’re
point (zero of the countdown corresponds to liftoff), how- initially rather off and only come together later in the as-
ever this poses a time-constraint: The Shuttle needs to be cent.
fully ready for liftoff, in particular APUs need to be run- A more visual picture of the launch progress is provided
ning — and if that is not the case, a launchpad abort will by the ASCENT TRAJ displays. Note that they’re not
occur. Thus, this option is not suitable for beginners who showing altitude against range but altitude against velocity
need some time to learn procedures. (ground-relative on ASCENT TRAJ 1 and inertial on AS-
If liftoff is successful, the Shuttle will rise till it has CENT TRAJ 2). So while going up means you get higher,
cleared the pad, then rotate till the tailfin points into the going right means you go faster.
direction of the launch azimuth, then pull back into an in- In addition to the current state (a triangle), the display
verted flight attitude and pitch down to about 75-80 degrees. also shows predictor circles — assuming you do not change
There’s a lot of things to monitor now and the informa- pitch or thrust, they’re where the Shuttle will be 30 (and 60)
tion is very dense. Most of your attention should be focused seconds into the future. Since in reality you do change pitch
on the PFD and the DPS ASCENT TRAJ 1 (Fig. 2.5). and throttle, they’re never exact but nevertheless give you a
Let’s start with the PFD. On the ADI ball, you can see good idea about what is happening. Also note that current
current vehicle attitude. During launch, it’s preferable to throttle setting is displayed on the right — though you can
have the ADI coordinates either in REF or LVLH, inertial see engine status better from the OMS/MPS display.
attitude isn’t really that useful because we pitch relative to The idea of a launch is then to fly the Shuttle close to the
the horizon. The two tapes to the left show current altitude shown trajectory which is propellant-efficient and satisfies
with changing scales, first in feet, then thousands of feet, all structural constraints. The AP will do a reasonable job
then finally miles) and vertical speed — but in fact it’s easier with this — it won’t usually follow the trajectory exactly
to monitor altitude and the vertical speed trend on the DPS everywhere, but it will stay reasonably close to it. There’s
display. hence no need to be alarmed if you see some minor devia-
During the passage through the atmosphere, it’s crucial tion.
to monitor KEAS (knots equivalent airspeed). While this So, after the Shuttle has acquired launch course, watch
LAUNCH USING AUTOPILOT 37

If specified in the mission file (this is not the default),

there is one significant event that will take place at around
Mach 14 — the Shuttle will roll to heads-up attitude. This
takes about 30 seconds and is done to allow an early com-
munication link to the TDRS satellite network. Since the
thrust vector of the SSMEs is not the same as the roll axis,
the roll is a moderately cumbersome affair and might cause
some transient motion as guidance re-aligns the Shuttle with
the desired trajectory. Following the roll, a steady climb to-
wards orbital speed follows just like without the roll.
This is in fact a good opportunity to focus on engine sta-
tus — look at the OMS/MPS display (Fig. 2.6). In the lower
right is the current chamber pressure of the engines with
100% corresponding to full rated power (it is possible to
throttle the engines higher in emergency situations). Above
is helium reservoir and regulator pressure, with the reser-
voir slowly depleting as the flight progresses.
Now - what do we need helium for?
The Shuttle’s main engines use a pre-burner to drive
Figure 2.6: Towards MECO, engines are throttled back. the turbo-pumps which bring the fuel into the combustion
chamber. But that means that liquid oxygen and hydrogen
are on the same turbine shaft at some point - if they ever get
in contact, the engine explodes. To avoid that, the gap is
KEAS climb towards ∼ 450, then reduce again as you clear
purged with high-pressure helium while the engine is run-
the dense atmosphere. Note how we pitch down further
ning. So the depletion of helium tells how long the engine
to increase speed, close to the limit permitted by the wing
has been running, and if the regulator pressure ever drops
bending moments. Follow the triangle climb out to SRB
into the red, the engine needs to be off immediately or you
separation.
might lose all three. Thus, if a helium pressure ever goes
After separation, the DPS display switches automatically
red, you need to act quickly! But at the moment, you should
to ASCENT TRAJ 2 and a few minor changed occur on
see green all across the board.
the PFD. Technically, there has been a major mode transi-
Eventually the ET will be so empty that acceleration
tion to MM 103. Look at the accelerometer — while we
reaches close to 3 g. At this point, you can see that the
had close to 3 g with the SRBs burning, SSME acceleration
engines are throttled back automatically to avoid overstress-
is much lower with the ET still mostly full. But vertical
ing the attachment points between orbiter and ET. Roughly
speed is high at this point — watch how the Shuttle climbs
at this point, the predictors will also start to move upward,
ballistically while SSME thrust is vectored horizontally to
indicating that if you do nothing, centrifugal force will push
increase velocity, and only when vertical speed has dropped
you higher — in reality we’ll pitch down of course and
to near-zero, thrust is vectored downward again to stabilize
throttle back continuously.
altitude.
If the ’real’ message stream is activated, you will get to Note: If the throttle is not set to AUTO, throttling
hear on-screen messages and voice callouts like Two en- back will not occur automatically and the resulting
gine TAL or Negative return — these are the so-called abort accelerations if you do not throttle back manually
boundary callouts which tell you what to do in the case of will destroy the Shuttle.
an engine failure. At the moment, you can just ignore them,
launch aborts are a very advanced topic. Finally, the triangle at the top of ASCENT TRAJ 2 will
The next few minutes are a more relaxed affair — pitch begin to move along the scale till it reaches CO (cutoff)
is gradually decreased as centrifugal force ramps up, always — once that happens, the AP will switch all engines off.
to keep vertical speed rougly close to zero. Usually there’s a Congratulations — you’re in orbit!
small sink at this point. Mach number increases steadily and You can look at the propellant reserves now — AS-
acceleration ramps up as the ET empties. Also, ASCENT CENT TRAJ 2 shows them — likely they’ll be fairly large,
TRAJ 2 will now show a counter that indicates when guid- some 5%, if you’ve followed the tutorial. That’s because
ance expects MECO to happen - it’s based on a prediction we picked the most efficient inclination possible (high-
for the trajectory and thus only accurate up to a few seconds inclination is more costly) and, most importantly, we didn’t
at this stage, but you can monitor it to get an idea how much carry a payload. Once you try to lug a full payload into a
longer the ascent will take. high inclination orbit, there’ll be less propellant available
38 CHAPTER 2. LAUNCH

Figure 2.7: The important instruments during the second stage of the launch.

(there’ll always be some, because it would damage the en- gushes out and reaches the engine flames, an explosion is
gines to let them run dry, and since they’re supposed to be the most likely outcome.
re-used, main-engine cutoff (MECO) occurs always before Thus, in order to train a manual ascent, it might be a
that happens unless there is an engine failure). good idea to disable such scenarios. This can be done via
Space Shuttle → Limits and Failures by setting the simu-
lation mode to ’soft’. This means limit violations will re-
Manually flying a launch sult in a warning on-screen, but no actual damage will be
simulated. The simulation will also not include any com-
Piloting a launch manually (by ’control stick steering’ or ponent failures (engine flameouts, power losses, hydraulic
CSS) is not how the Shuttle is usually operated. There’s a failures,. . . ) when not instructed to do so, so with limit sim-
couple of reasons for that, for instance accurately moving ulation to ’soft’ you’re safe from nasty surprises.
controls under a sustained 3 g acceleration is not that easy, Also, keeping all on-screen callout types active is a good
and especially while the SRBs are burning astronauts de- idea during the first few launches because they provide
scribe the ride like ’going down a rough road fast without helpful information on what is about to happen or what
any suspension’. It is however possible to fly an ascent in you are supposed to do. Later, when you are more familiar
CSS — and much more rewarding than to let the AP do it. with the procedures, you can gradually disable the ’help’
It isn’t even that difficult. The digital autopilot for as- and ’info’ streams to get closer to reality.
cent is a rate-control scheme which will hold vehicle at- In general, control the Shuttle with smooth motions dur-
titude steady if you do nothing and any movement of the ing ascent, don’t try to yank it around or accelerations will
controls will command the vehicle to assume a certain turn- grow too large. Keep rates reasonably small — below six
ing rate. You never need to bother with fine-print like trim degrees per second, there’s no hurry to any maneuver.
or make minute adjustments to the controls — the vehicle You can fly all maneuvers by using the pink needles on
moves like on rails. Also thrust vectoring is a powerful con- the PFD — they’re ’fly-to’ indicators linked to launch guid-
trol method, so you can expect near-instantaneous response ance. For instance, if the roll needle moves left, roll left
in all situations and no sluggishness or lag. The pilot’s task till the needle is centered again, if the pitch needle moves
is mainly in planning the attitude the vehicle should assume down, pitch down till it is centered again. However, you
(and guidance gives multiple aids to do that). will inevitably lag behind guidance that way, so it pays off
The greatest danger is that the safe ascent path through to also monitor the overall launch progress on LAUNCH
the atmosphere is comparatively narrow — the Shuttle is TRAJ 1 and 2. You often need to plan ahead — so use the
pushed close to its structural limits near max. q. Like all predictor circles to aim at the trajectory.
rockets, it is in essence a flying bomb, and with the weight
of all elements (including the ET structure) trimmed down • The most difficult part is to roll to the correct launch
to a minimum, breakage of a wing or an attachment strut azimuth after liftoff — mostly because you’re look-
of the orbiter to the ET likely means the wreckage punches ing right into the Euler singularity can can’t easily get
through the ET hull and rips it open — once propellant (and an attitude reading. You can cheat by orienting your
oxygen from the umbilical connecting the Shuttle and tank) launchpad using
MANUALLY FLYING A LAUNCH 39

--heading=XXX Note: Remember that if throttle is on manual,

where XXX is 180 degrees to launch azimuth (remem- MECO must be done manually using the cutoff but-
ber, FG assumes you are an airplane, so the Shut- tons and will not be commanded by guidance.
tle gets initialized 180 degree to launch heading, then
pitched up 90 degrees onto the pad — so if you just Congratulations on your first CSS ascent!
pull back you get right onto the right course). Alterna-
tively it’s a good idea to pitch back to about 85 degrees
and then use a combination of yaw and roll to get onto
the right course. Practice makes perfect!
• Pitch to about 78 degrees when on launch course and
wait for the callout to throttle down — monitor the
OMS/MPS page to verify you’re not going below 67%
of rated power. Don’t delay throttling down too much
or you will exceed pressure limits.
• Wait for the callout to throttle back up, when that
has happened gently pitch down to 45 degrees (still in
heads-down attitude) and wait till SRB separation.
• After SRB separation, pitch down to about -10 degrees
and monitor vertical speed. Once vertical speed drops
to zero, pitch up to vector thrust down and hold vertical
speed around zero — use ASCENT TRAJ 2 to aim the
predictors at the nominal trajectory line and stay there.
• If you feel adventurous, you can roll to heads-up at-
titude around Mach 14. Anticipate that you need to
change pitch and do yaw corrections because the en-
gines are mounted at an angle — the attitude that vec-
tors enough thrust down in inverted flight is not the
same as in upright flight.
• Keep an eye on Mach number — once you pass Mach
18, you can use gentle yaw corrections (don’t yaw out
of your course for more than a few degrees) to null
any inclination errors if you’re aiming at a particular
inclination. Don’t do this much before, because due to
the velocity you got from Earth’s rotation it is natural
to have inclination errors until inertial speed is large.
• Once acceleration reaches close to 3 g, start throttling
back the engines to keep acceleration below 3 g —
there will be warning calls if you don’t.
• The throttle is mapped to the range 67% to max. cho-
sen value, so you can’t ’idle’ the engine like in a plane
by pushing throttle into lowest position and neither can
you completely cut engine thrust with the throttle.
• Shortly before reaching orbital velocity, there’ll be
a callout to prepare you. Monitor the small tri-
angle move towards the CO marker and cut thrust
once it’s there by depressing the engine cutoff but-
tons Ctrl + q , Ctrl + w and Ctrl + e in short
sequence.
40 CHAPTER 2. LAUNCH
Chapter 3

Orbital insertion

Overview produce an on-screen message to tell you which you have

just taken if the ’essential’ callout stream is active).
Let’s pick up a Shuttle mission right after MECO. For a
nominal ascent, we’re not on a stable orbit yet, rather At-
lantis should be on a rising trajectory towards the desired
apogee (highest point of the orbit), but the perigee (lowest
point) should be inside the upper reaches of the atmosphere.
The reason is that we don’t want the ET to remain in space
— it’s supposed to lose energy due to atmospheric friction,
enter the low atmosphere and burn up.
Thus, at our apogee, we will ignite the orbital maneuver-
ing engines (OMS) to circularize the orbit and leave the tank
behind. That point is a good 45 minutes away — but there’s
plenty to do during that time — we need to switch off sys-
tems we only need during launch, yet on the other hand ac-
tivate systems we need for orbital operations. There’s some
7000 lb of propellant and oxidizer left in the engine mani-
folds and feedlines — they need to go. The fuel lines need
to be vacuum inerted. And of course the OMS burn itself
needs to be programmed. Figure 3.1: Orbital DAP pushbuttons on panel C3.

The transition DAP supports both automatic and manual

Attitude control operation of the RCS thrusters — see the DAP pushbuttons
on panel C3 (Fig. 3.1). Right now only the INRTL talkback
If so specified in the mission file, the external tank should in the upper row should be illuminated, indicating that the
disconnect automatically, followed by a transition to major Shuttle is under manual control with inertial attitude hold
mode 104. If not, you can disconnect the tank manually mode engaged.
using the switch on panel C3 or the key d . What does that mean? It means the DAP commands the
Now that thrust vectoring of the main engines is no Shuttle to have zero rotation rates if you don’t move the
longer available, the Shuttle has to rely on the reaction con- stick — if you move the stick, a fixed rotation rate around
trol system (RCS) for attitude control and small-scale ma- an axis is commanded. You will notice if you look out that
neuvering. A digital autopilot (DAP) converts inputs from there’s some 0.07 deg per second rotation relative to Earth
the control devices into thruster firing commands. While we though — that’s because a fixed inertial attitude is not fixed
are still in OPS 1, only a limited set of options is supplied relative to the horizon — to fix attitude relative to a horizon
by the Transition DAP, after transiting to OPS 2, the Orbital (as an airplane does), we need to engage the LVLH mode,
DAP selection will be much more sophisticated. but that is not available in the transition DAP.
There’s actually two control devices in the real Shuttle Gently move the stick and see how the Shuttle responds.
cockpit - the translational hand controller (THC) and the ro- If you pitch down, you ought to see the ET which was dis-
tational hand controller (RHC). Chances are you have only connected earlier. On a real mission, it’d be photographed at
one joystick plugged into your computer, so the simulation this point to determine whether foam has come off during
uses m to switch from RHC to THC and back (this will launch and might have damaged the Shuttle’s heat shield.

41
42 CHAPTER 3. ORBITAL INSERTION

The rate controlled DAP actually makes flying rather pleas-

ant — you don’t normally notice that it all runs on thrusters,
it feels like handling a plane. If you ever find yourself tum-
bling out of control during a maneuver in space, just switch
the DAP to INRTL and center controls — it’ll end any spin
on its own! But now, let’s get to work!

Propellant dump
During ascent, propellant and oxidizer flow from the ET
through the umbilical lines into the aft-section of the Shut-
tle where they are distributed to the engines. At MECO,
the flow to the engines is cut, but there is still propellant
and oxidizer in the feedlines and distribution manifolds. In
fact, there is quite a lot, some 7000 lb of it, and if left there
it’s not only hazardous due to potential pressure buildup but
also extra weight that alters the center of gravity. The Shut- Figure 3.2: Location of the O2 manifold pressurization (1)
tle can’t be trimmed for aerodynamic flight with so much and the H2 and O2 fill/drain valve (2) controls.
weight to the rear.
Thus, it has to go (as a side remark, note that out of the
two, liquid oxygen is in fact the heavier and hence more
problematic for aerodynamics than the hydrogen). Both
fuel and oxidizer are expelled using high pressure helium
fed into the manifolds. It would be a bad idea to allow
oxygen and hydrogen to mix during the process (remem-
ber, this is rocket fuel we’re talking about!), so the oxygen
goes through the engines while the hydrogen goes through
the fill/drain opening to the side.
If the mission file specifies it, the propellant
dump will be done automatically and controlled
by the avionics, otherwise you have to do it man-
ually. Open Help → Aircraft Checklists and
bring up the Post-MECO checklist. Also open
the Space Shuttle → Main Propulsion System and
Space Shuttle → Propellant Fill/Drain dialogs. Figure 3.3: A propellant dump in outside view.
Alternatively you can operate the controls from the 3d
cockpit — the panel is probably the most difficult to lo-
calize, you can only see it from the pilot’s seat, moving
a bit forward and then looking backward (Fig. 3.2). The we have limited fuel for them and we’re going to need them
good news is that it’s not used very often — nominally only during entry). Everything is done on panel R2.
the hydraulics shutdown of the main engines later uses it as First, take the operate switches from ’run’ to ’stop’ po-
well. sition. Once the RPM indicators show that the turbines are
Click to open the manifold press LO2 valve (allowing no longer running, switch off the controller power and close
the oxygen to be expelled through the engines) on the the fuel valves (Fig. 2.4 may be of help). Finally, locate the
Main Propulsion System dialog, then open the LH2 inboard HYD CIRC PUMP switches (right above the fuel valves)
and outboard fill/drain valves (Propellant Fill/Drain dialog). and switch all three to the GPC middle position. This will
Verify that the checklist shows green, then initiate the dump keep the hydraulic fluid circulating through the system and
— either via the dialog button or the switch on the top of prevent blockages from forming while it isn’t actively used.
panel R2. You can switch to outside view and observe the Okay — now the tasks which need immediate attention
leftover propellant vent. are taken care of, it’s time to think of orbital insertion. In the
Once the process has finished (about two minutes later), early days of the Shuttle programs, MECO was done prior
close the LO2 manifold pressurization valve again. to reaching orbital speed so that the tank would drop into
Now it’s time to switch off the APUs again (remember, the atmosphere quickly. Thus, a few minutes after MECO,
THE STATE VECTOR 43

a burn by the orbital maneuvering system (OMS) was re- unique equations telling us where the spacecraft will be at
quired (OMS-1) to raise the apogee to a target value and the any future time.
Shuttle would leave the tank behind right them. This six number combination of position and velocity
In the later days, MECO occurred somewhat later, saving vector (usually along with a time) is called the Shuttle’s
OMS propellant, and Shuttle and ET would coast close-by state vector, and maintaining a good state vector is primar-
towards apogee, and only then go separate ways when the ily what orbital navigation is about. If you ever want to look
Shuttle circularized orbit. at it, you can find it on SPEC 34 (Fig. 3.4) in the lower right
In order to understand just what needs to be planned and corner under ORBITER STATE.
executed next, let’s talk a bit about orbital maneuvering in However, while the state vector is very useful for avion-
general. ics computations, unless you’re a math genius, you’re un-
likely to make much of the six numbers and their incessant
change. On the other hand, we’ve argued above that an or-
The state vector bit is basically simple — so isn’t there a way to express the
properties of it in six other numbers which show that it’s
Once in orbit, the Shuttle is in an (approximately) friction- simple?
less environment. Due to the lack of atmosphere, there are There is — one can use six different numbers, the so-
no dissipative forces which would change its energy as long called orbital elements, out of which five are unchanged in
as no thrusters are ignited. All orbiting bodies move on el- time and show the characteristics of the orbit, whereas the
liptical trajectories around the center of gravity (the center last shows the motion of the orbiting body in time.
of Earth), and as the energy of the orbiting body is con-
served, so is its angular momentum.
The orbital elements
Each closed orbit is an ellipse (in inertial coordinates, that
is), with one of the focal points of the ellipse at the center
of the gravitating mass. For a perfectly spherical mass dis-
tribution (which Earth is almost, but not quite), all the mass
distribution is equivalent to a point mass at the center of the
sphere, so in a way, the Shuttle can be thought of to be on
an orbit immediately after clearing the atmosphere — that
orbit would just intersect with the surface of Earth. Gen-
erally we only think of proper orbits only once they don’t
intersect the atmosphere any more.
For low orbits, it is fairly natural to describe the shape of
the ellipse by the closest and farthest distance to the point
mass — the apoapsis and periapsis (or, in the case of Earth,
apogee and perigee). Note that for the orbits the Shuttle
Figure 3.4: The SPEC 34 display showing the Shuttle’s can actually reach, it makes more sense to quote apogee
state vector. and perigee above the surface of Earth (i.e. they become
the ’altitudes’ between the Shuttle is flying, but in order to
convert them to the ellipse shape, the mean Earth radius of
In order to describe where the Shuttle is at any given mo- RE ∼ 6370 km needs to be added
ment, we have to know the current time and position. The For higher orbits, alternatively often semi-major axis and
position in turn corresponds to three numbers in a given co- eccentricity are used. The semi-major axis is the mean of
ordinate system. We could describe orbits by supplying a apoapsis and periapsis, whereas the eccentricity is the ra-
huge table of position coordinates dependent on time — but tio of the difference between apo- and periapsis and their
it would be a very awkward description. sum. Since the Shuttle can only reach orbits of max. ∼ 600
Rather, knowing the position at some time and in addi- km above the surface and the atmospheric influence extends
tion the three components of the velocity vector is enough to about 150 km above the surface, the difference between
to know everything about the orbit. From the position in apo- and periapsis can be a few hundred kilometers at most,
the gravitational field and the velocity, we can compute po- so orbital eccentricity is always small and semi-major axis
tential and kinetic energy and add them up to find total or- is aways a few hundred km more than RE .
bital energy. Likewise, from position and velocity we can The orbit always moves in a plane — the next two ele-
compute the angular momentum vector. Making use of the ments have to do with the location of that plane with respect
fact that both are conserved along the orbit, we can derive to a reference plane — Earth’s equatorial plane.
44 CHAPTER 3. ORBITAL INSERTION

Figure 3.5: Groundtracks of a low inclination (Shuttle) and a high inclination (ISS) orbit

A Shuttle orbit may be tilted with respect to the equa- stant, as the shape of the orbit never alters. They are hence
tor by an angle — this angle is known as the inclination. an efficient way to see the relevant properties of an orbit
The point where the orbiting object crosses the equatorial (but somewhat awkward to use in computations).
plane from the lower hemisphere into the upper hemisphere Actually, in reality this is not quite true — since Earth
is known as the ascending node, and the longitude of the is not exactly a pointmass, there are minute changes over
ascending node is the next orbital element. time in all the other orbital elements, but for all practical
Due to Earth rotation, the ascending node doesn’t remain purposes their mean values are still very useful.
at the same longitude in an earth-relative coordinate system One can understand a lot of orbital mechanics from the
though, it has meaning only in an inertial coordinate sys- conservation of orbital energy and angular momentum. The
tem. Nevertheless, from the groundtrack of orbits (Fig. 3.5) angular momentum vector is conserved both in direction
the meaning of inclination and ascending node longitude and magnitude. Conservation of direction means that the
is rather apparent: Spherical trigonometry implies that the trajectory will always stay in the same orbital plane — this
groundtrack crosses the equator always at an angle equal to fixes inclination and longitude of the ascending node.
the inclination and does not reach higher in latitude than the Conservation of magnitude means that at the points
inclination. where the velocity vector is perpendicular to the radius vec-
In the absence of rotation, groundtracks would be closed tor, i.e. apogee and perigee, the product of (inertial) speed
sinusoidals — due to the rotation, they do not close, each as- and orbital radius is constant. Thus, if you know radius at
cending node moves westward from orbit to orbit in Earth- apogee RA and perigee Rp and orbital speed at perigee vp ,
relative coordinates. Two orbits with the same inclination it follows that the speed at apogee va = Rp /Ra · vp — in
but different longitude of the ascending node cross the equa- other words, speed drops as the Shuttle climbs to apogee
tor at different longitudes, but under the same angle and and rises again as it descends to perigee.
reach to the same latitudes. This corresponds to a tradeoff between kinetic and po-
Now, the first two elements describe the shape of the el- tential energy — moving higher in the gravitational poten-
lipse, the next two the plane in which the ellipse lies – the tial needs part of the kinetic energy which is returned when
fifth then connects the two. The argument of the periapsis is the Shuttle drops lower. Looking at the energy conserva-
the angle between ascending node and periapsis (you would tion allows to compute the speed for every altitude if it is
not see it directly on the groundtrack, but it tells you where known at one altitude. The total orbital energy and the rela-
on the orbital motion the lowest point is reached). tive balance between kinetic and potential energy determine
With these five elements, the complete shape of an orbit semi-major axis and eccentricity, or apogee and perigee.
is set. The last element, the true anomaly, finally describes By applying thrust, the Shuttle can alter its velocity vec-
where on the ellipse the orbiting body currently is (usually tor to some degree. If the thrust is applied along the current
in terms of an angle relative to the periapsis). vector (prograde) or against it (retrograde) the direction of
The beauty of a description of orbits in terms of elements the velocity vector is unchanged but its magnitude changes.
rather than in terms of a state vector is that while all six Such burns alter the kinetic energy and therefore apogee and
parameters of the state vector change in time, only the true perigee of an orbit.
anomaly changes in time, all other elements remain con- If thrust is applied perpendicular to both the orbital plane
SHUTTLE ORBITAL MANEUVERING 45

and the velocity vector (normal or anti-normal — pointed After MECO, HA should be quite a bit larger than HP. To
towards the horizon 90 degree to the direction of the tra- the left, slightly higher, is an entry TTA with a time indica-
jectory), it alters only direction of the velocity vector, but tor — that’s the time to the next apsis. Since after MECO
not magnitude, i.e. kinetic energy remains the same. Hence we’re heading upward to the apoapsis, it reads TTA and will
such burns alter the orbital plane, i.e. inclination and longi- revert to TTP once we passed apoapsis.
tude of the ascending node. Surely the important information is not displayed in an
Finally, a burn upward or downward (radial or anti- overly prominent fashion. . . but this is the Shuttle, remem-
radial) alters direction of the velocity vector only but also ber? And in reality, of course Mission Control always
moves the Shuttle higher or lower in the gravitational poten- knows.
tial, i.e. general alters the orbit in a way that is not simple. Thus, after MECO apogee should read about whatever
value was programmed in the mission file or the launch
guidance dialog. If it is lower, we’d need an OMS-1 pro-
Shuttle orbital maneuvering grade burn, otherwise we need an OMS-2 burn at the apoap-
sis (you can read off TTA how much more time to get there)
In principle, a transfer from any orbit into any other orbit to circularize orbit.
can be de-composed into a sequence of properly times pro- How long a burn? In reality mission control would com-
(retro-)grade and normal burns, there is no need to ever do pute that for us, or we could use targeting software aboard
radial burns. the Shuttle, but for the moment, we can get a good estimate
Changes of the orbital plane close to a planet however with a simple rule of thumb:
are very propellant-costly, so realistically the Shuttle can’t
change inclination by more than a few tenth of a degree 1 percent of OMS propellant
once in orbit — MECO has to occur essentially in the target = 6 feet per second velocity change
orbital plane, or the target orbit can never be recovered later. = 3 nautical miles apsis change
Only small correction nudges are possible.
In practice, most orbital maneuvering of the Shuttle that Thus, for every mile we need to raise periapsis during
is not for rendezvous purposes hence deals with controlling circularization, we’re going to need a velocity change ∆v of
apoapsis, periapsis and the argument of the periapsis, i.e. 2 ft/sec. This can be directly programmed into the Shuttle’s
consist of timed prograde and retrograde burns. guidance as a so-called PEG (powered explicit guidance)-7
Now, where aboard is the information on the current orbit target.
available? Let’s try this. First, if we want to do an OMS-2 burn only,
change to OPS 105 (the display is otherwise identical). In
the lower left corner, there is an entry mask under TGT PEG
7 — that’s where we can specify ∆v values. The entries are
done along the orbital coordinate systems - x is prograde, y
is normal and z is radial.
To raise periapsis by 100 miles, we’re going to add ∆vx
of 200 ft/s, so let’s enter
ITEM 19 + 200 EXEC
and verify that the value is shown. We also need to
tell guidance when to do the burn — that’s ignition time
(TIG) and done with items 10 - 13 (enter subsequently days,
hours, minutes and seconds in mission elapsed time).
If everything is entered, let’s load it
ITEM 22 EXEC
and observe how some values change. Most notably, over
at the apsis display there’s now values of TGT HA and TGT
Figure 3.6: Orbital maneuvering on OPS 104. HP which indicate what guidance estimates the burn will
do. If it gives you the desired orbit, leave as is, otherwise ex-
ecute item 22 again to unload the solution, enter a new ∆vx
Take a look at the DPS screen OPS 104 (Fig. 3.6) — it or check whether the ignition time is at the apoapsis (TTA
ought to open automatically after ET separation, but if it away from current MET) and iterate until you’re happy.
does not, use OPS 104 PRO). About in the middle of the (Unlike the real Shuttle, the FG version does not require
right column, you can see a table with columns HA and HP to enter an ignition time, in which case the solution is com-
and a row CUR — the meaning of the entries is the current puted for a burn right now and the computation is repeated
height of apoapsis (in miles, HA) and of the periapsis (HP). in every avionics update cycle— so you don’t need to fiddle
46 CHAPTER 3. ORBITAL INSERTION

with entering time and figuring out what MET currently is,
you can simply wait till TTA is about five minutes and enter
your burn parameters then).

Predicting the effect of a future burn is one of the few

tasks of the avionics which actually take an appre-
ciable time — this is because internally a numerical
fast-forward of the state vector is computed. So don’t
be overly surprised if the solution doesn’t appear im-
mediately. Also, note that the accuracy deteriorates
if you try to compute several orbits ahead.

If you’re happy with the solution you can use

ITEM 23 EXEC
to activate the CRT timer (upper right of all DPS pages)
to count to the ignition time.The burn is now programmed.
Since there’s probably some time still to the apogee, we
can work through the post-OMS-1 checklist next and take
down a few more systems we no longer need.

Figure 3.7: Engine controller power and helium isolation

MPS shutdown valve switches on panel R2.

Let’s start with the engines. The engine controller units

which manage the SSMEs are still powered at this point
— time to take them down, we won’t need them any more The insertion burn
and they suck a lot of power from the fuel cells. Each con-
troller is redundantly powered by two AC buses, so there’s
six switches in total on panel R2 that need to be off. At this point, you should have programmed a timed OMS
Below are the switches for the valves connecting the high burn and watch the countdown timer to roughly agree with
pressure helium supply with the engines — there’s eight the TTA (time to apoapsis) indicator on OPS 105. If you
isolation valves in total which all need to be closed (see haven’t done so, you can still program the burn some five
Fig. 3.7). minutes to apsis without specifying an ignition time and
There’s still traces of propellant left in the fuel lines even start it manually by looking at TTA.
after the propellant dump, probably by now evaporated.
We’ll now open the LO2 and LH2 inboard and outboard
fill-drain valves (on the menu dialog) for two minutes, then
close them again. This should empty the manifolds of any
residuals.
Finally, the umbilical door in the heat shield through
which the ET feeds into the engine manifolds is still open
— it needs to be closed as it is a huge gap in the thermal
protection. The switches to do that are located on the bot-
tom part of panel R2 (see Fig. 3.8). Start with the center line
latch keeping the umbilical door open, wait for the talkback
to change to STO (stowed), then operate the left and right
door closing switches, wait for the talkbacks to show CL
and finally latch left and right door firmly — wait for the
talkbacks to indicate LAT.
Then take a breath to enjoy the scenery. We need to wait Figure 3.8: The ET umbilical door switches on panel R2.
for apogee to insert into a circular orbit.
THE INSERTION BURN 47

If you programmed a burn earlier, the timing com- on OPS 105 (the next right of it should read AUTO to
pared with TTA might not be exact — this is partially indicate the current DAP setting). The Shuttle now starts to
due to the accuracy of the prediction and partially due maneuver into attitude, and a timer below MNVR 27 will
to the fact that every time you maneuver in any way, give an estimate when the correct attitude will be reached.
you change TTA a bit. Usually being a bit off is no Once the timer string vanishes, that’s the case (the burn atti-
reason for concern. tude is prograde aligned, heads down with the nose pointed
a bit below zero pitch — also the OMS engines are mounted
at an angle.
Note: If the text next to item 22 reads LOAD, you If you have not specified an ignition time, an EXEC mes-
have not actually loaded the burn into the guidance sage will start flashing in the upper right corner as soon as
computer yet — just modified the parameters. The you are in attitude, otherwise it will flash once the timer has
display can be a bit deceptive, so make sure the text fifteen seconds left to go. Press
reads CNCL. EXEC
Also, on the upper right, DVTOT should show the on the key pad to start the burn. The OMS/MPS MEDS
planned velocity change in ft/sec and TGO should incicate display will show how chamber pressure climbs up as the
the duration of the burn. The various components VGO X, engines burn. At the same time VGO X, Y and Z on OPS
Y and Z show what the expected acceleration in body co- 105 will decrease with the acceleration sensed along At-
ordinates is — since the Shuttle body isn’t aligned with the lantis’ body axes — since the burn is vectored through the
OMS thrust axis, generally non-zero components are ex- CoG, expect VX to be about six times VZ and VY to be
pected even for a strictly prograde burn. close to zero for a two engine burn.

Note: Only the DPS pages designated with MNVR

Note: If you operate the Shuttle with realistic naviga-
EXEC (MM 104, 105, 202 and 302) allow to ignite
tion, you need to make sure that item 9 shows the cur-
engines — the ones designated as MNVR COAST
rent weight of the orbiter (or at least a good approx-
(MM 106, 301 and 303) are for planning purpose
imation of it) including payload. The burn computer
only.
will use the number entered there to compute burn
duration, and if you specify the wrong weight, the
acceleration will be wrong and the burn won’t match
the target. If you use perfect navigation, the com-
puter will automatically pick the real current weight.

The first step is to test engine gimbal functionality (it’s

bad to discover during the burn that the gimbal motors have
failed). Do
ITEM 34 EXEC
and see if any error messages are produced (there should
be none). Now would be a good time to arm the engines. On
the upper left part of panel C3 are two covered switches la-
beled OMS ENG. Switch them to ARM/PRESS (the middle
position) — that arms the engines and supplies Nitrogen for
the purge of the lines afterwards. Operationally it’s a good
idea to disarm the engines whenever you don’t want to use
them — but don’t forget to switch them on again! Figure 3.9: A two engine OMS burn seen from outside.

Should you never need to end a burn early (for in-

stance because you discover during the burn that the Look at current apses and verify that they approach closer
weight is entered wrongly and the orbital change to targeted apses as the burn progresses. During the burn,
would be too large), taking the engine switches out the AP switches to OMS thrust vector control (TVC) to
of ARM/PRESS will cut the engines immediately. maintain burn attitude, after the OMS engines shut down
again, control is automatically returned to the RCS (if you
We need to get into prograde burn attitude next. The want to have manual control, depress the INRTL pushbut-
Shuttle can do that automatically — press the AUTO but- ton instead of the AUTO pushbutton).
ton in the control row on the orbital DAP pushbutton panel If you’ve done everything correctly, you should now be
on C3, then do in a near-circular orbit, with the ET vanishing behind you
ITEM 27 EXEC (likely you’re already in the night portion of your orbit and
48 CHAPTER 3. ORBITAL INSERTION

can’t see much outside anyway). Generally VGO X, Y and

Z won’t be exactly zero at the end of the burn — you can
switch to RCS in translational mode and null them manually
for increased precision as long as you are on a page that
reads MNVR EXEC (acceleration sensing will end as soon
as the corresponding DPS page is left).

You can also fly the OMS burn manually — if you

engage ITEM 27, the avionics will compute errors
shown on the ADI, but if you leave the orbital DAP
in INRTL, you have manual control. Simply use the
error lines in the ADI as ’fly to’ indicators to keep
the correct orientation during the burn.

We’re ready to prepare on-orbit operations.

Chapter 4

Preparing for orbital operations

This chapter describes a Shuttle mission right after the ing up and coolant loops to take the heat away from the
orbital insertion burn. You should be on a nearly circular avionics — as well as passive techniques — attitude con-
orbit, or at least on an orbit that no longer intersects with trol to manage the amount of sunlight that hits the Shuttle.
the upper atmosphere. From now on, there is little hurry to To better understand how this all ties together, let’s go
do anything — you can take your time and go through all through a short summary of how temperatures in space
steps carefully. arise. Technically the Shuttle is a well-isolated system, so
To make the Shuttle ready for extended orbital opera- its thermal energy changes only due to radiation. In the
tions, thermal management becomes crucial, so we need to absence of any internal heating, sunlight heats every sur-
learn about the radiators, the freon cooling loops as well as face it hits (with dark surfaces absorbing more energy and
heating elements. To operate the radiators, we need to open hence getting hotter than white or mirror-bright surfaces).
the payload bay doors. The infrared (IR) radiation from Earth likewise heats every
Likewise, with the software transition to OPS 2, many earthward-pointing surface about equally. As a result, the
new functions for attitude control become available. Usu- surfaces warm. Heat conduction can carry part of the en-
ally one GPC is now configured to run systems management ergy to shaded parts of the Shuttle — this effect is stronger
DPS pages, so we also have much more access to Shuttle in- if the temperature difference gets large.
ternals. However, every surface of the Shuttle does not only ab-
sorb radiation, it also radiates itself — proportional to the
fourth power of the temperature, as it gets hotter, the energy
Thermal management radiated off increases dramatically. Eventually the energy
flux carried away by radiation and conduction will equal
Atlantis is a complicated bit of machinery, and many com- the incoming energy, at which point the surface has reached
ponents have their preferred operating temperature. Hy- its equilibrium temperature.
draulic lines should not freeze and block. Cryo tanks should Running Atlantis’ systems produces some 12-14 kW of
remain cold enough for their contents to remain liquid, but power output which ultimately ends up heating the interior.
hot enough such that there is vapour pressure forcing the Since the cabin is well insulated (otherwise temperatures
contents out when needed. Finally, computers and other during entry would rise high enough to fry everyone inside),
pieces of the avionics should not overheat or they get de- temperatures inside the cabin would increase to intolerable
stroyed, and the interior of the cabin should remain at a levels for both crew and avionics well before an equilibrium
comfortable temperature. between heating and conduction to the outside is reached.
Yet outside is the vacuum of space, allowing unfiltered Coolant loops take care of this problem. The main avion-
sunlight to hit one side of the Shuttle during the day, with no ics bay has two redundant water loops for cooling, several
atmosphere keeping any warmth during the night. The tem- other pieces of equipment sit on cold plates. Both water
perature differences between a sunlit surface and a shaded loops and cold plates connect to the freon cooling loops
surface can be huge — and materials expand and contract in via heat exchangers. The freon loop carries heat from the
unexpected ways under such temperature variations — ten- avionics to the radiator panels on the inside of the payload
sions resulting from uneven temperatures may for instance bay doors from where it is radiated into space. The panels
prevent the payload bay door from moving. are large enough to radiate some 18 kW of power off, thus
Thermal management is a collection of techniques to that is sufficient to deal with the internal heating.
make sure that all systems remain at their operating tem- When the radiator panels are not available because the
peratures at all times. This involves active techniques — payload bay door is closed, the freon loop is cooled by flash
heaters to prevent thrusters and hydraulic lines from freez- evaporators which use the on-board water storage boiling

49
50 CHAPTER 4. PREPARING FOR ORBITAL OPERATIONS

off to take away heat. This doesn’t work all the way down For instance a new nominal bus assignment table (NBAT)
into the atmosphere, so during entry eventually ammonia is loaded changing the way the GPCs read out the various
boilers take over the task. Loss of all cooling options of data buses, and a new software distribution is used — while
the freon loop usually means overheating and failing avion- during launch four GPCs were running the primary avionics
ics half an hour later, i.e. should be treated as a mission- guidance, navigation and control (GNC) functionality, now
relevant problem. only three run this in a redundant set and the fourth runs
Another kind of issue arises when the Shuttle assumes systems management (SM) — you can check the current
the same inertial attitude for a prolonged time. Then one assignment on DISP 6 at all times.
side always faces the sun during the day portion of the orbit Now, how to access SM? That’s what the major function
(and heats quite a lot), but the opposite side never receives switches labeled CRT 1-3 on panel C2 are for — if you
any sunlight, and part of the Shuttle even face away from move a switch from GNC (upper position) into SM (middle
Earth’s IR radiation, so they get very cold. Cold enough position), the corresponding IDP will now show SM rather
in fact to jam the hydraulics lines or let propellant freeze in than GNC software on all screens it controls (the switches
thruster feedlines. There are heaters installed to prevent that are in fact 3-way switches, allowing to access also payload
from happening, but whenever feasible, a more resource- (PL) software — that however never got installed aboard
friendly solution is to slowly let the Shuttle rotate to make the Shuttle, so the PL position really doesn’t do anything).
sure sunlight equally heats all sides. If you do it, you’ll end up with the OPS 201 SM page
Now, let’s prepare the avionics for orbital operations and which is the antenna control utility (which at this point
get used to see some thermal management. doesn’t do much as the antenna isn’t deployed yet).

DPS systems management

Load the OPS 2 software by typing
OPS 201 PRO
(remember, OPS transitions are done by calling up the
first major mode). The immediate result will be to bring up
the Universal Pointing utility page.
Note: In the following, it is assumed that the simpli-
fied DPS simulation option is chosen. With the de-
tailed option, this won’t work and several additional
procedures are required.

Figure 4.2: The antenna management page.

The DPS SM page tree works the same way as the GNC
tree — there’s SPEC and DISP which can lie ’over’ an OPS
page, you can go back by the resume key, there’s systems
summary pages which can be accessed and toggled with key
commands. . . — there’s just no key command to go from
SM to GNC or back, you always need to use the IDP switch.
Note that there can be overlapping page designations —
OPS 201 in GNC is universal pointing, OPS 201 in SM
however leads you to antenna management. OPS 202 in
GNC is orbital maneuvering, whereas OPS 202 in SM is
payload bay door control. So don’t get confused (and if
Figure 4.1: The OPS 201 page.
you get frustrated, remember John Young and that you’re in
good company).
However, more is happening underneath (in fact, the tran- Operationally it’s a good idea to have one screen devoted
sition in FG happens unrealistically fast — in the real Shut- to SM and leave the rest to display the MEDS pages and the
tle, there would be an extended wait for an OPS transition). GNC DPS functionality, but you can do as you like.
ORBITAL DAP SELECTION 51

Radiative equilibrium temperatures can be computed from the balance between incoming and outgoing energy flux of a
surface. The incoming radiation flux from the Sun at the location of Earth is jin = 1.36 kW/m2 , most of this at optical
wavelengths. However, much of this gets reflected by the surfaces. The reflection factor is called albedo A, and only the
fraction 1 − A of the incoming energy flux actually heats a surface. For a pristine white or mirror-bright surface, A can
be as high as 0.9. If the surface is at an angle with the sun, an additional factor sin(α) of the sine of the incident angle
appears — at incident light parallel to the surface, no absorption can happen.
Outgoing radiation is to a good approximation thermal blackbody radiation in the infrared with a total flux dependent on
temperature T as

jout = σT 4
with the Stefan-Boltzmann constant σ = 5.67 · 10−8 Wm−2 K−4 . Thus, radiated energy grows quickly with temperature.
Equating jin = jout so that the surface neither absorbs additional energy nor radiates it off, the temperature follows as
p
Teq = 4 jin (1 − A) sin(α)/σ
(note since that both incoming and outgoing radiation is proportional to the area, it doesn’t matter for the temperature how
large a surface element is).
The above expression would give zero temperature for a shaded surface. This of course is not realistic — in reality, it takes
a while to reach equilibrium temperature (materials have thermal inertia proportional to their mass) and heat conduction
from lit to shaded surface also influences temperatures. But the simple equation above shows the main factors driving
equilibrium temperatures — the hottest temperatures will be reached for near-black (A = 0) surfaces with full exposure
to the sun.

Orbital DAP selection type of jets the DAP uses to control attitude. PRI and ALT
select the more powerful primary RCS thrusters (in outside
So far, we’ve been controlling the Shuttle via the transition view they cause a visible flame), VERN selects the weak
DAP which had pretty limited functionality — now we have Vernier thrusters which don’t allow very crisp maneuvering
a much richer set of tools at our disposal, and you’ll notice but also don’t use much propellant.
that there’s now many more talkbaks illuminated on the or- Under translation, there is the low-Z and high-Z option
bital DAP pushbutton panel. (Z is the coordinate upward seen from the payload bay).
Under SELECT, you now have a choice between A and Low-z is used for docking or to separate from a payload —
B. These are pre-defined configurations for rates, dead- since a translation burn to separate from a payload or reduce
bands, pulse durations etc. which can be edited in orbit approach velocity normally points the exhaust plumes right
— the pushbutton just allows to quickly change from one to at the payload (which isn’t good), low-Z instead utilizes the
the other configuration without having to enter lots of val- angled forward and backward pointing thrusters to create
ues into the keyboard. a net Z-axis translation. That’s just 1/12 of the thrust of a
Under CONTROL, there are the familiar choices AUTO nominal Z-translation, but the payload survives.
and INRTL available, but in addition also LVLH and FREE The rest of the buttons change between continuous or
can be selected. Like inertial attitude hold, LVLH is also an pulsed burns (translation) or stick controls rates and pulsed
attitude hold mode, but it holds ’local vertical, local hori- rotation burns. Pulses are generated every time the stick is
zon’ — which means using this function makes the Shuttle moved out of neutral, so to make use of pulsed modes, you
behave a bit like an airplane. It will keep a fixed attitude need to flick the stick a few times into the desired direction
relative to the horizon and the velocity vector — since an and return it to zero rather than hold it there.
orbit goes all around Earth, that’s only possible by creating Pulsed modes can be selected per-channel, so it’s possi-
a small rotation rate of about 0.07 deg/s. Finally, FREE puts ble to fly a rate controlled DAP in pitch and yaw but maneu-
the Shuttle into free drift — the DAP won’t try to hold any ver with pulsed burns in the roll axis. Putting all rotational
attitude or control thrusters to a fixed turn rate, rather stick channels in PULSE mode is equivalent to selecting controls
movements will cause angular acceleration. FREE.
Below, under MANUAL MODE, are buttons for transla- Take some time to explore what the various modes do
tion and rotation separately. Note that if you have selected (remember, m switches between rotation and translation).
control AUTO, most of these buttons won’t do anything. Having grown accustomed to INRTL DAP A with pri-
The exception are PRI, ALT and VERN. These select the mary thrusters which makes you easily forget that you’re
52 CHAPTER 4. PREPARING FOR ORBITAL OPERATIONS

in space where every motion has to be canceled explicitly

as there’s no friction, you probably wonder why the other
modes are there at all.
The reason is that moving the Shuttle around with high
turn rates and tight deadbands on attitude and rates is nice,
but costs a lot of propellant. And on a normal mission pro-
pellant needs to last for more than a week. So usually it’s
sufficient to turn the Shuttle slowly and to bring it approxi-
mately into the right attitude. Using Vernier thrusters in free
drift mode will accomplish that for you at a small fraction
of propellant usage. Same if you want to hold a fixed atti-
tude — Vernier thrusters can manage the task more cheaply Figure 4.3: The OMS and RCS heater controls on A14.
and more exactly than primary thrusters.
If you’ve been alert, you’ll also have noticed something
else: The free drift or pulsed modes are difficult to fly, be-
cause if you try to roll, there’s also pitching and yawing mo-
tion created. That’s called cross couplings, and the reason
is that the RCS thrusters, mounted at the nose and the two
OMS pods, are not symmetrical. Sideward firing thrusters
for instance are much lower at the nose than at the rear —
thus there’s a net rolling moment generated. They’re also
not equally far from the center of gravity — causing a yaw-
ing moment.
In general, compensating thrusters are fired to mitigate
these effects (the attitude-holding modes automatically use
all channels to maneuver anyway), so you get the worst
cross coupling from the Vernier thrusters where often no
such compensation is available. But control cross couplings
Figure 4.4: Ku-antenna circuit breakers on R14.
are a fact of life during Shuttle operations, and pilots learn
to anticipate them and live with them.

to activate only the group (1), the two breakers labeled (2)
Starting the radiator loop power the jettison mechanism.
Now it’s time to start the thermal management. Open the Finally, close the MPS TVC isolation valves —
post-insertion checklist. The first thing we need to do is this will disconnect the engines from the hydraulic
to switch heaters on (or rather to AUTO — heaters are system while on-orbit (you’ll find the dialog under
not always on but controlled by thermostats) to keep the Main Propulsion System or on the ’hidden’ panel behind
RCS and OMS at operating temperatures. Do this from the the pilot’s seat, see Fig. 3.2).
dialog Space Shuttle → RCS/OMS thermal management . Now turn left and have a look at the thermal controls on
You might notice that some are labeled as A and B — that’s panel L1 (Fig. 4.5).
the designation of the main DC power bus these systems Currently water coolant loop and freon pumps are run-
connect to. Critical systems are always redundently pow- ning already, just the freon is cooled by the flash evapora-
ered by more than one bus, such that they can still be oper- tor system (FES) and doesn’t go through the radiator. We
ated if part of the Shuttle’s power grid goes. start by switching both radiator loop controllers on (1). That
In case you want to use the 3d cockpit switches instead, should bring the talkbacks below from barberpole to BYP,
they’re located low on the aft section of the flightdeck on indicating that the radiator loop is operational, but currently
panel A14 (see Fig.4.3) — the panel also has controls for by-passed by the freon. Now we close the by-pass valves
the explosive separation of equipment, so don’t randomly (2) to bring the freon through the radiator.
try switches!). Of course, the radiator won’t work just yet because the
Next do the circuit breakers of the heaters for payload bay door is still closed, so all we could do is radiate
the Ku-antenna (you’ll find them in the dialog under into the payload bay, which wouldn’t be very useful. So we
Mechanical Systems ). leave the flash evaporator on for the moment.
In the 3d cockpit, the circuit breakers are located on the By the way, if you were alert, you could have noticed a
lower right side on panel R14 (see Fig. 4.4) — be careful tiny exhaust plume at the left aft section of the Shuttle in
STARTING THE RADIATOR LOOP 53

Figure 4.5: Starting up the radiator loop.

outside views — that’s the FES exhaust. And if you put item 14 to activate software by-passing of the mechanical
orbital DAP into FREE, you’ll also note that the exhaust switch, then item 15 to command the doors open and watch
is propulsive, i.e. exerts a tiny yawing force. An attitudethe show.
hold DAP cancels that out, but the force still requires regu- On OPS 202 you can follow how latches are retracted
lar thruster firings. in sequence and finally the doors moved, and you can also
So to make the radiator actually work, we need to open take a look through the payload specialist station window
the payload bay doors. There’s a switch, but we can also or admire the sequence in outside view. It takes a good two
use the software to do it for a change — bring up SM OPS minutes with all actuators intact to open the door. Once
202 (Fig. 4.6). that’s done, bring the software switch to stop (item 16),
de-activate software bypass (item 14) and switch power off
(item 1).
If you want to use the payload bay door controls in the
cockpit rather than the software, you can find them on panel
R13L on the right side of the aft section of the flightdeck,
see Fig. 4.7 (you need to have the detailed cockpit option
selected). The same panel also contains the Ku antenna de-
ploy switch.
Note the very similar switches which provide power to
the payload bay door (upper left) and which provide power
to the payload bay mechanical systems (upper right) —
SCOM contains a specific warning not to confuse them!
To open or close the door, one of the systems needs to be
powered on, and then the payload bay door switch (center
left) will work. The talkback above the main switch indi-
Figure 4.6: The payload bay door controls page. cates the current state of the door as OP (open), CL (closed)
or barberpole (in transit).
Verify in the checklist that the radiator is operating, then
Select item 1 to power the electromechanical actuators, switch the FES off on panel L1 (3) and also de-activate HI-
then item 3 to select automatic door opening mode. Select load mode. We now should have proper cooling — but how
54 CHAPTER 4. PREPARING FOR ORBITAL OPERATIONS

Figure 4.8: The SM systems summary 2 display.

Figure 4.7: The payload bay door controls on R13L.

not working), ’H’ means the parameter is off-scale on the

to verify this? high end, i.e. the true value is higher than displayed but the
sensor can’t determine it properly, ’L’ means the parameter
is off-scale on the low end, ’↑’ indicates a parameter that
Temperature readings is higher than it should be, ↓ a parameter that is lower than
it should be. In this way, problems can be spotted more
The simple way to find out whether the radia- quickly.
tor loop is properly cooling is to use the dialog In the lower right, you can see the freon temperature af-
Space Shuttle → Orbiter Temperature Distribution . ter passing the evaporator. This should be below 65 F when
This gives a quick summary of temperatures across the in orbit (it can go higher during ascent and insertion). To
orbiter, including freon temperatures (in K). Note that tem- the left of it are three avionics bay temperature readings.
peratures there hardly ever seem to change. The reason is Above you can see the APU temperature readings next to
that the timescale to reach equilibrium temperatures is not TEMP EGT — they are very high when the APUs are actu-
seconds but rather hours. The Shuttle has a lot of mass, ally running, but once they’re cooled off, they represent the
and that means lots of energy needs to be radiated off or temperature in the rear part of the Shuttle fuselage where
absorbed to change its temperature significantly. So half an the APUs are located.
hour after liftoff, you can not expect to see anything near The temperature readings in the upper part of the display
equilibrium temperatures. however are readouts from cryogenic storage tanks and do
In order to see the effect of thermal management on not represent ambient temperature in the orbiter.
the Shuttle more quickly (for educational purposes), it You can take some time to increase thermal system com-
is possible to clock up the simulation speed of ther- putation speed and see how the various temperatures change
mal physics by a factor hundred, bringing equilibration when you e.g. switch water cooling off, or activate two wa-
times into the range of minutes. This is done on the ter cooling loops, or de-activate the high load evaporator, or
Space Shuttle → Simulation Options dialog. A sure sign alternatively cool via radiator or via the evaporators (how-
that the radiator is cooling is a difference between freon in ever, remember that you will fry the avionics if you manage
and freon out temperature. to get avionics bay temperature larger than 130 F or 328 K,
If you want to be realistic stay within the Shuttle avion- and that will be the end of all controlled flight).
ics, there’s temperature readings available in many places. If you want to check on the action of the OMS and RCS
For instance, bring up SM SYS SUMM 2 — you can either heaters, you can do
toggle it by pressing the SYS SUMM key twice or type SPEC 89 PRO
SPEC 79 PRO to get a summary of temperature readings for the rear
directly on a screen that runs systems management. OMS pods, the forward RCS module and the crossfeed lines
If the simulation option ’SM realism’ is set to enhanced, (Fig. 4.9).
the display does not only show the alphanumerics but also The temperatures are supposed to be in the range of 55-
indicates limit status indicators in overbright (as in Fig. 4.8). 75 F for proper operation — if they get much colder, fuel
An ’M’ behind means that data is missing (the sensor is might freeze in the feedlines and block them, disabling the
AUTOMATIC ATTITUDE CONTROL 55

hydraulic fluid through a heat exchanger with the warm part

of the freon loop.
In the simulation, it is quite possible that the freon tem-
perature trips a low temperature alarm once or twice, es-
pecially in a transient state when the radiator has just been
switched on. This usually is no reason for concern and can
be left to sort itself out on its own. However, you actually
need to check whether the alarm was caused by a low tem-
perature condition, because for instance if power-hungry
equipment like the engine controllers are still running, the
radiator loop is unable to cool the freon sufficiently, and the
resulting high temperature condition will eventually cause
equipment failures.
After you’ve studied all temperature readings, this might
be a good time to go through the remaining checklist. With
Figure 4.9: OMS and RCS temperature readouts. the payload bay door now open, you can deploy the Ku-
band antenna on Space Shuttle → Mechanical systems
(you should have activated the four heater switches before).
You can also use the in-cockpit deploy switch on panel
engines (the individual sampling points refer to the location R13L.
of the various temperature sensors, but that’s too detailed Next, flick the three spray boiler duct heaters to prevent
— just focus on the overall range of temperatures for each water vapour from condensing and blocking the ducts —
system). we’re going to need the APUs in a few days when we de-
The thermal state of the hydraulic fluid can be accessed orbit and non-operational steam vents would be bad. You
using can find the switches on R2 right above the boiler power
SPEC 87 PRO switches.
Finally, again under Mechanical systems , switch power
for the star tracker doors on and command both doors to
open (or use the star tracker controls on the overhead pan-
els).

Automatic attitude control

Now that Atlantis is prepared for orbital operations, we
could get used to life in space. In reality, the crew would
now change from the ’pumpkin’ suits worn during ascent
into more comfortable wear, stow the seats in the middle
deck and in general start to make the Shuttle their home.
At the same time, the actual mission program might start.
During this orientation flight, we don’t really have one, but
we’re going to explore the various options attitude control
Figure 4.10: Hydraulic temperature readouts. offers to assist a mission.
In weightlessness, any attitude works equally well —
whether we ’fly’ upside down, on the tail, or backward
If the hydraulic circulation pumps are running, the tem- doesn’t really matter. But it might matter for the mission.
peratures shown should be roughly equal for all sensors For instance, if we have Earth-observation equipment in the
on the right part of the display and the accumulator pres- payload bay, we might want to keep the payload bay pointed
sure (ACCUM P) for all systems should be above 2500 towards Earth — the LHLV pushbutton would accomplish
psi. When the pump switches are in GPC position (as they this for us. If we have sensitive equipment in the payload
should be if you’ve been following this manual) then the bay, we might not want that to get too cold — thus we might
pumps are automatically switched on periodically to keep want to keep the payload bay pointed at the Sun so that it
the fluid moving and temperatures equilibrated. The hy- gets warmed during the day portion of the orbit. We might
draulic system is passively heated by pumping fluid through want to heat the Shuttle equally to prevent thermal stresses
warm parts of the Shuttle and more actively by passing the and thus slowly rotate it around an axis. Or we might want
56 CHAPTER 4. PREPARING FOR ORBITAL OPERATIONS

to observe a particular spot on the planet and keep the pay- but you can change to the backward pointing or the upward
load bay pointed at a ground coordinate. pointing axis. Item 19 allows to enter a tracking mode in
All these things are possible and supported by automatic which the chosen body axis is pointed towards a tracking
maneuvers — in fact, that’s what the universal pointing dis- target — this can for instance be a location on Earth (en-
play (OPS 201) is for. tered via items 11 to 13) — you can use this routine to point
First, switch the ADI coordinate system to inertial (and observation instruments in the payload bay at a specific lo-
using the edgekeys the PFD to ORBIT) — this will help cation on the planet as the orbit crosses nearby. Or you can
to understand what happens next. Press the orbital DAP make the Shuttle track a given star in the sky. Details of
AUTO pushbutton, then do these modes are however a bit beyond the scope of this in-
ITEM 20 EXEC troduction.
on GNC OPS 201 — it’s an option designated ROT.
You’ll observe that the Shuttle immediately starts to roll.
It’s an automatic routine to hold the current inertial attitude Antenna operations
and roll around the body X axis. If you’re alert, you’ll see
Let’s talk a bit about communications with the ground (al-
quite a bit of thruster firings, and the rotation is pretty fast.
though there’s not so much mission control simulated).
Now switch to DAP-B — the rotation will slow down and
Usually this is done via the TDRS satellite network — the
thruster firings will decrease notably, but small drifts out of
Ku-band antenna is automatically directed to track one of
attitude will happen.
the satellite positions, and the satellite in geosync orbit re-
What sets the difference? Each DAP-A and B are char- lays the signal to the ground bases. Given the spread of
acterized by lots of definitions of rates, deadbands, pulse the TDRS satellites, this means that there’s communications
durations etc. DAP-A will rotate at 2.5 deg per second be- coverage for over 90% of an orbit.
cause that is the rotation rate specified, DAP-B will do at The Ku-band antenna also transmits at a fairly high band-
0.5 deg/s, if you switch to Vernier thrusters the rates will be width, so there’s telemetry and video image up- and down-
much lower. In fact, you can access and change all these link available while it operates.
values during a mission — try For situations when the Ku-antenna or the TDRS network
SPEC 20 PRO is not available, the Shuttle also has S-band antennas. These
and you’ll see the DAP configuration utility where you operate at a much lower bandwidth and require direct line-
can see and alter the specified rates and deadbands for each of-sight to a ground station. Using the S-band typically lim-
DAP. Feel free to tinker with this some other time, for now its telemetry and restricts contact with the ground to voice
we resume back to universal pointing (OPS 201). Select communication — and dependent on orbit and orbital alti-
item 21 to cancel the rotation routine. tude, communication is only available for irregular intervals
A more advanced option is automatic maneuver to an at- of perhaps five minutes whenever a ground station happens
titude. You can specify inertial yaw, pitch and roll with to be in reach.
items 5-7 (the ADI shows this if you put the coordinate sys- If you have deployed the antenna, take a look at SM OPS
tem switch on panel F6 into INRTL). Attitude and attitude 201 again. The central portion of the display is (perhaps
targets and errors are also shown on the lower right of the not easily recognizable) a schematic drawing of the Shuttle
universal pointing utility. and its payload bay, seen from the antenna position. The
Once you have specified a target attitude, item 18 exe- pointing vector of the antenna is shown as a square, and
cutes maneuvering to it. whenever the square gets below the line, the Shuttle fuse-
You might have noticed that unlike the orbital maneuver- lage blocks line of sight for the antenna.
ing utility, there is no display of the currently chosen DAP Pointing angles are shown to the left of the display. Be-
on this display. Instead, if you try to start a maneuver when low the graphical portion is a list of the TDRS positions,
not in AUTO mode, you will get a flashing error message which of these are currently in view and an indication which
above the scratch line. Acknowledge it using ACK, then one the antenna is actually tracking. You may find that at
switch to AUTO if this happens, then delete the message any given time, quite many satellites can be in view.
using MSG RESET on the keypad and depress IO RESET The right part of the display is dedicated to the S-band
to allow the message to re-appear if the condition recurs. antenna. There you can see a station indicator string when-
As with the rotation mode, you’ll observe that too tight ever the antenna finds a ground base in range and a display
deadbands have a way of never converging and produce which quadrant of the antenna is transmitting.
constant thruster fire — if you want to avoid this, switch During nominal operations, there’s not overly much the
to a DAP that has less stringent attitude deadbands. crew needs to do with the antenna. The only exception is
There are yet more advanced options, Using item 14, you that during proximity operations with a rendezvous target,
can change the body axis that gets pointed into the chosen the Ku-antenna can be used as a ranging device. We’ll cover
attitude. By default, that’s the forward-pointing +X axis, such details of the antenna system later.
ANTENNA OPERATIONS 57

However, let’s discuss another important aspect of a

Space Shuttle mission next — the handling of payload.
58 CHAPTER 4. PREPARING FOR ORBITAL OPERATIONS
Chapter 5

Payload handling

Launching payloads payload on the launch pad and fly it into space). Looking
through the payload specialist station window into the bay,
Bringing a payload into orbit (and possibly back) is an im- you should see the satellite sitting in the bay. It’s a simple
portant part of Shuttle missions — be it that scientific equip- mockup of a paylod, easy to grab — a real satellite would
ment (SpaceLab) is flown into space, or be it that satellites not be transported with a deployed antenna and would be
or modules for the space station are carried — by design, cradled by a pallet. We’re going to grab it with the RMS
Atlantis is a heavy lifter. arm, lift it out of the payload bay and release it into space.
All payloads are carried in the payload bay where they
are mounted on pallets and held in place by the payload re-
tention system. The whole set of equipment used to handle
them in orbit is known as the payload deployment and re-
trieval system (PDRS). Beyond the retention system, also
the remote manipulator system (RMS) — the ’Canadarm’
— and the manipulator positioning mechanism (MPM) are
part of PDRS.
In the simulation, payloads can be explicit or im-
plicit. Explicit payloads have a 3d model associ-
ated and can be moved using the RMS arm, im-
plicit payloads are just a correction to the liftoff
weight, they change flight characteristics but do noth-
ing else. Space Shuttle → Simulation Options offers
the choice of all implemented explicit payloads whereas
Equipment → Propellant and payload allows to select im- Figure 5.1: The TDRS demo payload sitting in the payload
plicit payload weight. The same choices can be specified in bay.
the mission file.
Shuttle ascent (and to a degree entry and flight) character-
istics change with a payload — accelerations are noticeably First, we need to activate the RMS arm. This can be
smaller and more thrust component needs to be vectored done from the GUI, or if you have full cockpit detail also
downward to achieve the same trajectory. The automatic from the aft panels. Let’s cover the GUI first. Open
launch controller can compensate for the presence of a pay- Payload Operations → RMS arm deployment . First we
load — but take care to not feed an impossible task, with a need to power the RMS — shift the switch to ’primary’
full load, not all inclinations can be reached. When flying (you can also connect it to backup power, but usually we’d
manually you have to compensate for the effects of reduced use primary). During launch, there are strong accelerations
acceleration yourself. acting, whereas the RMS arm is a relatively frail structure
designed to operate in weightlessness. To keep it from be-
ing ripped off in powered flight, it is secured by a shoulder
PDRS deployment brace. This needs to go before the arm can move. This is
a one-time operation — once the brace is removed, it stays
Start this chapter in a stable orbit, select the ’TDRS Demo’ removed. Move the switch to ’port’ (that’s where the RMS
payload in Space Shuttle → Simulation Options (you can arm is installed — originally the Shuttle had plans for two
do that at any time — if you feel this is cheating, select the such arms, but they never materialized, as quite a few other

59
60 CHAPTER 5. PAYLOAD HANDLING

things).
The arm is now still securely held in three locations by
the manipulator positioning mechanism. Unlatch the MPM,
then choose ’deploy’ to release the arm. Note the three
ready to latch indicators at the lower part of the dialog —
they show when the arm is in a position to be secured by the
MPM — only when they show ready is it possibly to safely
stow the arm away.
The in-cockpit controls for the MPM are located on panel
A8L (Fig. 5.2). Power is a three way switch, bring it into
upper position.

Figure 5.3: The SM SPEC 94 page.

By choosing appropriate angles for all joints, the end

effector can be moved into a wide range of positions
and attitudes. That is done by an assortment of driving
modes, ranging from simple manual via augmented manual
and operator-commanded automatical to fully automatical
modes.
In the simplest modes, referred to as ’SINGLE’ or ’DI-
Figure 5.2: The RMS/MPM deployment controls on panel RECT’ on Payload Operations → RMS arm operation ,
A8L. the operator controls joint angles one by one. As you will
realize quickly when spending more than a few minutes in
this mode, this is extremely cumbersome.
The talkbacks to the far right indicate ready to latch sta-
Note that the dialog has three selectors (cooresponding
tus, in latched position the arm is of course technically
to rotary knobs on the panel). The first one selects what
ready. The remaining talkbacks are MPM status — stowed
is to be displayed on the alphanumerics above — besides
(STO) and deployed (DPY) — and RMS arm latch status,
’TEST’, the choices can be ’POSITION X/Y/Z’, ’ATTI-
latched (LAT) or released (REL). The MPM switch deploys
TUDE P/Y/R’ or ’JOINT ANGLE’. The second selector
the system, the latch switch unlatches the arm. There are
chooses the joint to which the first selector and command
also heater switches below — bring one system to AUTO
input refers to — so if you want to know the current angle
to keep the RMS arm under thermal control.
of the wrist pitch joint, you need to put the first selector to
You can also follow the progress of the MPM latch mi- ’JOINT ANGLE’ and the second to ’WRIST PITCH’ and
croswitch indicators as the arm is unlatched on SM SPEC you can read off the number. If you want to know where
94. the end effector currently is, the first selector goes to ’PO-
SITION X/Y/Z’ and the second to ’END EFF’.
The last selector is for the drive mode. Next to it, you can
RMS arm operation see a button ’enter’ and a talkback ’software stop’. What’s
that? To prevent you from inadvertedly switching on a
This would be a good time to talk some about the RMS arm. mode you do not want to drive, you need to confirm all
It’s a multi-jointed device made to resemble a human arm selections by pressing ’enter’.
— thus it has a shoulder part, an elbow part and a wrist
part. The shoulder has two joints which allow yaw and Note: No drive mode selection by menu or rotary
pitch movements (in the Shuttle coordinate system). The switch is valid unless it is confirmed with the button!
elbow has a single joint allowing pitch movements, and the Thus, looking at the currently selected mode can be
wrist consists of three joints allowing pitch, yaw and roll deceiving as the mode may be selected but not active.
movements. At the end of the arm is the end effector, cor-
responding to the human hand — that’s where the arm can The software stop is a safety measure. The joints have
grab. (obviously) mechanical limits, and it would be a bad idea to
RMS ARM OPERATION 61

drive the arm too quickly into them. So the operating soft- Note: Fully automatic modes need to be i-loaded to
ware prevents automatic modes from doing that by cutting work, they are not available by default.
commands for joint movements a bit before that happens —
that’s the software stop. Between soft stop and hard stop, Thus, for the fine work, we’ll use the so-called manual
the arm can only be driven joint by joint, no more sophisti- augmented modes.
cated mode is available. Note: When manual augmented RMS drive modes
(Note that the soft stop can be disabled by item 8 on are selected, your control input is assumed to refer
SPEC 94 — but do this only if you’re sure you know what to the RMS arm, i.e. you can’t control the Shuttle’s
you are doing!) attitude manually. Thus, put the orbital DAP into an
As it happens, the stowing position is just between hard attitude hold mode or an automatic mode before en-
and soft stop for shoulder and elbow pitch joints, so we need gaging any of these modes.
to drive these two joints out of that region before we can
operate more comfortably. Switch to ’ORB UNL X/Y/Z’ (orbiter unloaded transla-
Start by putting the drive selector to ’SINGLE’ and the tional mode). The mode is called ’unloaded’ because the
display selector to ’JOINT ANGLE’, then choose the shoul- reference point that is moved around is the end effector (for
der pitch joint. The switch on the dialog is used to decrease loaded modes, the reference point can be the center of grav-
or increase angle. Put the switch to the right and observe the ity of the payload, but such modes are currently not avail-
arm rise out of the stow position as the shoulder joint oper- able in the simulation).
ates — do this for perhaps 10 degrees, then put the switch You can now control the RMS arm with a translational
into middle position to stop the movement. hand controller with a speed proportional to the controller
Next select the elbow pitch joint and put the switch to the deflection and all the joints will flex automatically to ac-
right to drive the forward portion of the arm down by 10 complish that. What’s best, the end effector keeps its atti-
degrees. You should see the software stop talkback vanish. tude while you do that. The ’ORB UNL P/Y/R’ mode is
Now put the display selector to attitude and select the end fairly similar in that it allows you to change end effector at-
effector. You probably see that effector pitch isn’t exactly titude without changing position (and again, all joints move
zero. Choose the wrist pitch joint, and you’ll discover that automatically to accomplish that).
it shows the same attitude difference — drive it till it’s close Display end effector position coordinates and drive the
to zero. arm carefully to (11.5, 2.0, -1.8) — that is the attachment
Now, that’s how the single or direct drive modes work, point for this payload. It’s a bit hard to see how you’re
but they’re painstaking and tedious, so they’re not used doing out of the payload bay window, but that’s what the
much except where necessary (though it is quite possible to numerical position indicator is for!
operate the arm to grab a payload if really needed). We’re Once you have the arm in the correct position, verify that
now ready to switch to a more sophisticated mode. We will effector pitch, yaw and roll are all zero (if not, drive them
let the computer drive the arm into the vicinity of the TDRS to zero) and then click GRAB. You should get an on-screen
attachment point and then do the fine work manually. Turn message that the payload is successfully attached.
to SM SPEC 94. Items 18 to 20 allow to set a target po-
sition, items 21 - 23 a target attitude. Enter 10, 2 and 1 as
target z, y and z respectively and leave target pitch, yaw and
roll at zero. Then execute item 25 to do a reach limit check
for the entered target — this should come up as ’GOOD’.
With the target defined, put the drive mode switch into
AUTO OPR CMD and confirm with enter. You’ll see that
the arm starts moving. Select the position indication for
the end effector to see when it reaches the target, then de-
select the mode again. Neat, isn’t it? Although the software
doesn’t compute safe trajectories — so you could still bump
the arm into something by just entering coordinates.
For the sake of completeness, there are also fully auto-
mated sequences available where the arm flies through a se-
ries of pre-defined points. Sequences can be assigned to one
Figure 5.4: Moving the RMS arm towards the TDRS.
of four slots via items 13-16 on SPEC 94 and then called up
via the AUTO 1 to AUTO 4 selections. However, unless
we have a sequence we actually i-loaded in the mission file The in-cockpit controls for the RMS arm are located on
definition, this won’t help us much right now. panel A8U (Fig. 5.5).
62 CHAPTER 5. PAYLOAD HANDLING

mechanical components and for the logical controllers. Me-

chanical power is activated by the PL BAY MECH PWR
switches on panel R13L on the right side of the aft flight-
deck. The logic is switched via the retention system con-
trols on panel A6.
First power the system, then operate all five release
switches (also on panel A6) in the dialog and watch for the
indicators to change from ’LAT’ to ’OPN’ (you can also
follow the progress on DISP 97). Once all latches are open,
you can start to move the payload. You will notice that as
soon as you move it out of the correct position, the ’ready
to latch’ indicators will change.
If your mission ever is to catch and secure a payload,
you need to move it into the correct position such that the
’ready to latch’ indicators show ’RDY’, then you can close
the latches to secure the payload, then release the RMS arm.
For the time being, lift the satellite slowly out of the bay
and high up.
Figure 5.5: The RMS controls on panel A8U.

Displayed parameter selection, joint selection and drive

mode selection are rotary switches which need to be turned
into the respective position. In addition, the drive mode
requires to depress the ’ENTER’ button below to take an
effect. The drive mode talkback lights above indicate which
mode is currently active.
Note that the mode selection is fairly intricate because
there are software override options selectable on the SPEC
95 DPS page. This page again duplicates both the in-
cockpit and the menu controls, and if the software override
option is chosen, only software is allowed to set the RMS
control mode — neither GUI nor in-cockpit controls will Figure 5.6: Lifting the TDRS out of the payload bay.
have an effect.

Once you have it stretched out with reasonable clearance

The payload retention system from the Shuttle, you can release it and move the RMS arm
away. It will just float there, orbiting alongside with the
You might be tempted to immediately lift the satellite out Shuttle.
of the payload bay now that the RMS arm is attached, but Let’s gently push away from it — select a low-Z trans-
of course that doesn’t work. The TDRS isn’t free-floating lational DAP from the orbital DAP selection pushbuttons,
inside the bay (otherwise we’d have had a huge problem then ignite thrusters to move Atlantis to a safe distance.
during launch), it’s held in place by the payload retention Having a co-orbiting satellite is a good occasion to prac-
system — there’s five groups of latches which attach it. tice translational RCS use — you can try to bring the TDRS
Open DISP 97 on a DPS screen. This shows the cur- into your front window view — or even to catch it again
rent state of the latches. There’s three payload posi- with the RMS arm (which is hard to do).
tions on the real Shuttle, only one of them is supported
in the simulation. You can unlatch the payload from
Payload Operations → Payload retention (where you can Using software overrides
see ’ready to latch’ and ’latch status’ indicators. The status
shows that the latches are currently closed, and the ready to In case of a switch failure, the RMS arm can still be oper-
latch indicators show that the payload is in a position where ated, though a bit more awkward. This is done via software
the latches can grab it. commands, and the DPS page is SPEC 95, called PDRS
The payload retention system needs power both for its (payload deployment and retrieval system) OVERRIDE.
RMS POWERDOWN AND JETTISON 63

Payload Operations → RMS Arm deployment . Usually

the arm would only be completely powered down at the end
of a mission.
Now, let’s talk about something different — what if
there’s a real problem and the RMS arm gets stuck? In this
case, the payload bay doors can’t be closed at the end of a
mission, and it doesn’t take a genius to figure out that the
Shuttle can’t enter the atmosphere that way.
If that ever happens, we need to emergency jettison the
RMS arm. Open Payload Operations → RMS Pyro —
you see a dialog with an ’ARM’ switch and the switches
for the various explosive charges that can separate the arm
if needed. Alternatively the controls can be found on panel
A14 (see Fig. 4.3).

Figure 5.7: Software overrides for RMS controls. Note: Explosive charges are one-time devices —
once a piece of equipment is jettisoned that way, it’s
gone. Don’t explore these switches unless you know
you want to jettison the part for good.
Let’s try this and bring the page on screen (like all RMS-
related pages, this is an SM function, so the IDP must be set So, if the RMS arm ever gets stuck — arm the explosives,
accordingly) - you should see something like Fig. 5.7. blow the separation charges and watch the arm tumble off
The various blocks pretty much duplicate the controls in into space — you should be able to close the payload bay
the cockpit. Assume for instance the drive mode switch is door then. Note that the same is true for the Ku-band an-
no longer working and we need to change from AUTO 1 tenna which can also be jettisoned, should it ever block the
to SINGLE mode. To do so, we first activate the software doors from closing.
override for the switch via
ITEM 1 EXEC.
This puts a star behind the item to show that the override
is active. The column below now duplicates as the switch
selection position and the indicator light. To select SINGLE
mode, do
ITEM 3 EXEC
– this corresponds to turning the selector switch into the
correct position. You should see one of the stars change to
the SEL column. Now do
ITEM 13 EXEC
to do the equivalent of the enter pushbutton — and you
should see also the star in the IND column change, showing
that the mode is now properly selected. Now you should be
able to drive the RMS in SINGLE mode regardless of what
the selection switch is set to.
The other software options work in a similar way — to
activate e.g. the vernier rates which allow low movement
speeds, you need to select items 26 and 27 in sequence.

RMS powerdown and jettison

If you no longer need the RMS arm, it should be
driven back to the cradle (i.e. drive all joint angles
to zero — you can only do the final steps of that
in ’SINGLE’ or ’DIRECT’ mode due to the software
stop. Once the MPM status indicators show that the
arm is ready to be latched, re-latch the MPM using
64 CHAPTER 5. PAYLOAD HANDLING
Chapter 6

Coming home

Overview Nz holding pullout in which the Shuttle is in free-fall and

vertical speed increases (up to several hundred meters per
Eventually, it’s time to bring Atlantis home. The beauty of second) till the lower atmosphere is hit.
the Space Shuttle concept is that we can fly the spacecraft In order to reach the runway, we need to do ranging. This
to a precise touchdown point on the chosen runway — no works as follows — the remaining distance to the landing
need for a recovery fleet, helicopters or parachutes like for site tells us how quickly we need to decelerate — if we’re
the Apollo program. This is because the Shuttle is a winged coming in short, we need to decelerate less and lengthen the
vehicle — it generates aerodynamic lift in an atmosphere entry path, if we’re coming long we need to brake harder.
and can use it to control the entry trajectory. Deceleration rate is controlled by controlling drag. Drag is
However, let’s not rush things — the first part of the at- a function of airspeed, angle of attack (AoA) and air den-
mospheric trajectory is more concerned with decelerating sity — for given airspeed and AoA, there’s always some
and surviving the fierce heat generated by the shockwave density which leads to the desired drag. By controlling ver-
resulting from hitting air at Mach 27. tical velocity through lift, we can maneuver to the desired
There are several ways a body can enter the atmosphere, altitude where the air has the right density and hence adjust
and the Shuttle is equipped to survive two of them — the deceleration.
regular entry, and the Nz holding pullout procedure during There is, however, a catch. In an airplane, we would con-
aborts. trol lift with the elevator by adjusting AoA (simply put, if
The basic idea of the regular entry is to gradually trade you point the nose down, the plane goes down). In the Shut-
centrifugal force against aerodynamic lift. Consider an orbit tle, this doesn’t work.
with a perigee of 60 km — if there were no air friction, The reason is heating. Hitting the atmosphere at Mach 27
the Shuttle would descend rapidly from apogee, then slow generates a strong shockwave, and this transforms the air
down as perigee comes close, reach zero vertical velocity at into a superheated plasma. In the early days of spaceflight,
perigee and then rise again. is was believed that a sleek, streamlined object stands the
If there’s sufficient drag on the way down, the velocity best chance of reducing heat load as the friction is much re-
will be reduced, and perigee will gradually drop all the time. duced for such a body — but it was quickly realized that this
Eventually, the orbital trajectory will intersect the surface of isn’t actually true. It’s the opposite kind of shape which is
Earth, at which point the Shuttle is certainly no longer or- optimal — a blunt body creates a stronger shockwave, but it
biting. In fact, it can be considered gently falling — gently is detatched from the fuselage, and unlike for a streamlined
because there’s still significant centrifugal force. However, body, most of the heat is just carried away by the surround-
if at that point there’s also sufficient lift available, we can ing air.
change a vertically accelerating motion into one with an ap- This is why the Shuttle needs to enter the atmosphere
proximately constant descent rate. At some point further like a blunt body — at 40 deg AoA, technically in what
along, the centrifugal force responsible for orbital motion would be a stall region for almost any aircraft. In fact, ther-
won’t matter at all — but by then lift will be strong enough mal considerations pretty much regulate the target AoA as
by far to manage the descent rate. a function of Mach number all the way down to Mach 3.
Thus, the characteristics of a regular entry are that first But a consequence of that is that we can not adjust lift
vertical velocity increases (as part of orbital motion and via AoA like an airplane does during entry — AoA has to
some friction), then, as lift builds up, that increase stops be fixed at the target value given by Mach number. The only
(vertical acceleration crosses zero) and from there on de- thing we can control during entry is bank angle. Thus, lift
scent rate is controlled and roughly constant — expect some needs to be managed by bank angle. Flying at zero bank
70 m/s or 14.000 ft/min. This is in striking contrast to an angle gives maximum lift, rolling to large angles (up to 60+

65
66 CHAPTER 6. COMING HOME

of the landing site — weather at the site has to be predicted

good some time in advance. The Shuttle can only reach a
site when the future orbital groundtrack crosses sufficiently
close-by — the nominal cross-range of the Shuttle is above
1000 miles, but there should also be some margin for rang-
ing.
his means that in case the Shuttle needs to de-orbit
quickly, the regular landing sites may not be reachable at
all and a landing at an emergency site would be planned.
There are dozens of such sites all across the world, allow-
ing to de-orbit and land somewhere within an hour or two.
Several of these sites are now part of entry guidance.
In real life, mission control would determine and com-
pute the landing site — in the simulation, the groundtrack
map can be used instead.
Figure 6.1: Shuttle early atmospheric entry. Once chosen, the landing site is entered into the DPS on
SPEC 50 as item 41 according to Tab. 6.1.

code site remark

deg) drastically reduces lift.
Of course rolling to some non-zero bank angle means the 1 Kennedy Space Center regular
trajectory over ground curves — very shallow at first, more 2 Vandenberg Air Force Base planned regular
strongly later. This means we can’t fly a straight line to 3 Edwards Air Force Base regular
the landing site. Which doesn’t matter, because we can’t 4 White Sands Space Harbour regular, lakebed
do that anyway — Coriolis forces will prevent that. So in 5 Zaragoza Airbase TAL
practice we reverse the sign of the bank angle whenever we 6 RAF Fairford TAL
have reached a 10 deg difference in azimuth to the landing 7 Banjul International Airport TAL
site, and during a typical entry there will be several such 8 Moron Airbase TAL
roll reversals. 9 Le Tube TAL
11 Bermuda Intl. emergency site
There’s one final ingredient guidance utilizes — it isn’t
12 Halifax emergency site
quite true that AoA is fixed by thermal considerations,
13 Wilmington emergency site
there’s a ±3 deg variation possible. This corresponds to
14 Atlantic City emergency site
a ±10% change in drag. This is useful because unlike the
15 Myrtle Beach emergency site
drag change by going to a different density altitude, this
16 Gander emergency site
change is instantaneous, so it can be used to quickly do
17 Pease emergency site
small corrections to the trajectory.
18 Oceana Naval Air Station emergency site
The combination of high AoA for thermal protection,
30 Easter Island TAL
bank angle to control lift, vertical speed, altitude and ul-
32 Diego Garcia emergency site
timately drag, roll reversals and fast AoA changes is what
33 Honolulu Intl. emergency site
happens during the hypersonic phase of entry. The whole
34 Keflavik Airbase emergency site
dynamics is very different from normal flight and takes
35 Andersen Airbase Guam emergency site
some time to get used to — but the whole maneuver can
36 Amilcar Cabral Intl. emergency site
be flown (partially) in CSS without a huge workload, and
37 Ascension Island RAF emergency site
this is definitely an interesting experience.
38 Wake Island emergency site
But before we actually do it, we need to prepare the Shut-
39 Lajes Air Base emergency site
tle for entry and plan a de-orbit burn.
Table 6.1: Item codes for Shuttle landing sites implemented
in FG.
Pre de-orbit preparations
The list comprises actually used regular landing sites as
Preparations for de-orbit in a real mission begins well in well as planned sites (which never were used in reality),
advance in terms of winding down and stowing experiments transoceanic abort landing (TAL) sites and emergency sites.
and mounting seats for the crew in flight and middle deck In principle the Shuttle can land at any major airport in the
which have been removed during on-orbit preparations. world, however entry guidance is only available for the sites
The most significant bit for the simulation is the choice that can be entered into the avionics. Also, nighttime illumi-
PLANNING THE DE-ORBIT MANEUVER 67

nation of the runway will only be available at the supported short of actually starting an APU. Some 40 minutes before
landing sites. the planned de-orbit burn, the transition to OPS 3 is made
and the landing site entered as described above.
Note: In the following, it is assumed that the sim- Now we’re ready to hit the atmosphere — which we’ll do
plified DPS simulation option is chosen. With the after a de-orbit burn. For this, change to OPS 302. But what
detailed option, several additional procedures are re- burn to use?
quired.

Since SPEC 50 is only available in OPS 3 and OPS 3 has Planning the de-orbit maneuver
only limited orbital DAP as well as no systems manage-
ment software in PASS, we actually enter the landing site The basic problem can be states as follows: We want to hit
relatively late by the atmosphere at the entry interface (EI) with 400.000 ft
OPS 301 PRO altitude at a distance of 4000 miles to landing site. At the
followed by entry interface, we should have sufficient vertical velocity
SPEC 50 PRO downward to ensure atmosperic capture. That places a con-
and straint on the perigee — the EI has to be placed before the
ITEM 41 + 1 perigee, so the de-orbit burn happens less than half an orbit
(to select Kennedy Space Center (KSC) as landing site). before reaching landing site (a schematic view of the situa-
Once the site is selected, SPEC 50 also shows the currently tion is shown in Fig. 6.2).
chosen runway as well as the approach pattern — we leave To get a good vertical velocity at entry interface, we want
those as they are for the time being. to place perigee to some 30 miles. Using the rule of thumb
Some four hours before de-orbit, the freon in the radia- for de-orbit burns and the perigee predictor on DEORB
tor loop should be cold-soaked. During entry, cooling will MNVR EXEC (OPS 302), we can specify a suitable ∆vx
be provided by the flash evaporator system (FES), but they target. What remains is to adjust the ignition time such that
don’t work in the denser atmosphere, so we aim to have the burn changes the circular orbit we’re in into an elliptic
extra-cold freon which can absorb some 15 minutes worth orbit such that the trajectory intersects 400.000 ft at 4000
of heat buildup. miles to target.
Cold-soaking is accomplished by putting the Shuttle into
a tail-to-sun attitude for some time while selecting OUT Currently the on-board guidance works somewhat
TEMP HI on the radiator controller. Tail to sun means the better to target a 3000 mile range from EI to land-
cabin gets heated very little, thus the freon can grow rather ing site. This makes the entry slightly ’hotter’ and
cold. Freon is cooled that way for a good hour. The time requires a perigee of about 12 miles, but is still well
can be used to look at water stores (DISP 66) which are go- within the Shuttle’s capability.
ing to be used for the FES. Also, this is a good time to end
star tracker operations and close the star tracker doors. There are various tools available offline to do the calcula-
tion, but also the Shuttle has a way — there’s in the middle
Cold soaking in the simulation works a bit different of DEORB MNVR EXEC the innocent-looking entry REI
than in reality. In reality, the freon left in the radi- - that’s the quantity we need. It’s a prediction of the range
ator loop is disconnected from the rest of the loop from entry interface to the chosen landing site.
and allowed to chill, to be mixed into the loop later. So by varying ignition time, letting the avionics do the
In the simulation, only the overall freon temperature prediction and realizing that at orbital speed we’re moving
everywhere in the system can be reduced. some 4.5 miles per second, we can target REI to be about
4000 and the perigee about 30 miles — and that will be a
Once freon is nice and cold, verify that RMS arm is viable de-orbit burn.
stowed, retract Ku-band antenna and stow it properly. Then Of course, in reality this would be more sophisticated,
the payload bay door needs to be closed and the FES but it’ll do — we just need an accuracy of some 500 miles
switched on. to entry interface and the entry autopilot will do the rest.
(We’re intentionally short on describing the tasks, since (There’s also a lazy way to do it which doesn’t work on
the systems have already been explained during the deploy- the real Shuttle — if you just specify burn ∆vx but not ig-
ment of the systems after reaching orbit or prior to launch, nition time, you’ll get to see all parameters updated as if
so just refer to the earlier sections in case there is a ques- you would execute the burn right now — simply watch REI
tion). decrease till it has the right magnitude and start your burn
When everything is ready, final checks of the critical sys- then.)
tems are done. Using the OPS 301 display, OMS gimbal Yet a different way is to ask the mission control center for
checks are performed. The APU prestart procedure is done a de-orbit solution somewhat ahead of the burn (typically
68 CHAPTER 6. COMING HOME

entry interface

landing site

deorbit burn
Figure 6.2: A schematical view of the de-orbit targeting.

some 5 -to 15 minutes. Fig. 6.3 shows how a plan might causes a pitching moment which will turn the Shuttle unless
look like. it is countered by an opposite pitching moment from the air-
foils. However, the elevons have a maximal deflection, and
won’t create any pitching moment higher than some maxi-
mal value. Thus, if the CoG is off too far, the elevons can
no longer control pitch and we tumble helplessly through
the atmosphere.
To prevent that, the CoG has to be between 65 and 67%
of body length (you can check by asking mission control).
This can not be achieved if the OMS tanks are too full at this
point. Usually this should not be the case after a nominal
mission, but if so, the excess fuel needs to be dumped. This
can most conveniently done by adding a ∆vy component to
Figure 6.3: Timeline for getting deorbit burn parameters the burn target (this will by far be insufficient to bend the
from mission control. trajectory in any meaningful way, but waste fuel neverthe-
less).
For an entry into Kennedy Space Center, the burn is done Second, if the airfoils are to work, we need hydraulic
approximately when reaching Australia, and asking for burn power from the APUs. And we need to know whether the
parameters when leaving the African coast gives enough APUs actually work before the de-orbit burn which com-
time to prepare the burn while keeping the accuracy at a mits us to atmospheric entry. On the other hand, the APUs
reasonable level. consume fuel. To be safe, start therefore one APU five min-
utes before the burn but leave pressure to low to minimize
Note: The MCC provided solution is not a very accu- fuel consumption. In this way, we will have one APU run-
rate computation like it would be in reality — it is a ning when we hit the atmosphere no matter what.
quick parametric estimate designed to get you home
Program the de-orbit burn just as you did the insertion
and to give you a feeling of how burn solutions look
burn, then execute it once the time is right, change to OPS
like. Don’t expect it to work in unusual situations.
303 afterwards. Typically the de-orbit burn is 12.000 miles
Now, that we know how to target, let’s talk about two to landing site, that means we still have more than 6000
other important issues — trim and the APUs. miles to go before entry interface.
In orbit, it doesn’t really matter where the center of grav- On the MNVR display, you should now be able to see
ity (CoG) of the orbiter is, the RCS thrusters can orient it in TFF (time to four hundred thousand feet) — you can use
any case. In atmospheric flight, this is different. The wings this to estimate when you will hit the entry interface. Some
will generate lift at the center of lift, which will in gen- ten minutes before that, start the remaining APUs and en-
eral be different from the center of gravity. That difference gage hydraulic pumps to high pressure — verify all three
ENTRY 69

systems showing about 3000 psi. Note: The Aerojet DAP is designed to hold zero
If necessary, the Shuttle can maneuver on the pressure of sideslip, not to acquire it — do not engage Aerojet
a single APU, but this costs maneuverability, so there’s no if you’re not in the correct attitude, or you will regret
reason to push it. You have a few minutes to look through it quickly!
the MEDS APU/HYD display and check that everything is
working in good condition, so do it. The ENTRY TRAJ display is somewhat rich in informa-
tion, so let’s cover it in some detail. In essence, it is a repre-
sentation of the range to go on the x-axis (from right to left)
Note that in OPS 3, you won’t have access to systems vs. Shuttle velocity on the y-axis. A bright Shuttle-shaped
management software any more, so while there is a marker shows where the Shuttle is at the moment. How-
detailed APU and hydraulics monitoring page in the ever, the range to go is not linear, it is compressed towards
DPS, it won’t be usable when you actually need the the right side, so for the same velocity the Shuttle moves
APU. Don’t ask. . . more slowly on the right side of the display.
There are five paths drawn in solid lines from the upper
right to the lower left. The central one is the most important
Entry one, it represents the nominal entry trajectory, i.e. the path
guidance aims to fly. The left-most path represents thermal
and structural limits, left of this line the heat shield can not
be expected to hold up. The rightmost line shows the low
You can skip the de-orbit preparation and targeting
energy limit — right of this line guidance is no longer able
procedures and directly start at the entry interface by
to reach the landing site.
using --aircraft=SpaceShuttle-entry and se-
The Shuttle symbol is trailed by bright triangular mark-
lecting one of the entry scenarios from the menu.
ers which represent the past state of the Shuttle. They can
be used as a rough visual guide to estimate the trend. In
addition, there is a bright box-shaped marker on the same
height, but left or right of the Shuttle. This is the guidance
box, it shows where the Shuttle marker will go if no action
is taken. For converged guidance and an entry on the nom-
inal path, the box should be right on the Shuttle symbol. If
the Shuttle is off the nominal path, the box should be posi-
tioned towards it. The guidance box also trails a series of
bright dots to visually estimate the trend of guidance.
To the left are two scales — α (AOA) and D (drag). For
each of the scales, an arrow indicates what entry guidance
commands as the desired value whereas a triangle indicates
what the actual AOA and drag sensed by the accelerometers
is.
In addition, dotted lines cross the solid path lines.
They’re the so-called drag lines — when crossing e.g. the
15D line, the sensed drag (triangle on the D scale to the
Figure 6.4: The ENTRY TRAJ 1 page. left) should also read 15. Below the graphical portion of the
display are also numbers — they indicate sinkrates, when
the Shuttle marker is on the nominal trajectory with con-
Once Atlantis crosses 400.000 ft, we’re at the entry inter- verged guidance and above a number, the vertical velocity
face, and the atmosphere will make itself felt pretty quickly. in ft/s should roughly agree with the number. These indi-
From here on, we need to fly using the Aerojet DAP since cator lines are mainly for cross-reference in case there is a
orbital DAPs are not designed to cope with aerodynamic problem with entry guidance.
forces. Maneuver into entry attitude — zero sideslip beta, There are several alphanumerical values displayed as
then pitch up by 40 degrees, then request the transition to well — the only ones we will over here are the dynami-
Aerojet by doing cal pressure q and ∆Az, the relative azimuth towards the
OPS 304 PRO landing site. The first provides important cues when the
In response you’ll see the ENTRY TRAJ 1 display Shuttle controls change from jets to airfoils, and the second
(Fig. 6.4). Then switch pitch and roll/yaw channel from determines when a so-called roll reversal needs to be flown.
CSS to AUTO on the eyebrow panel to hand control of the The idea of reading the display is as follows: If the Shut-
Shuttle to the AP. tle marker is left of the nominal trajectory, it is high on en-
70 CHAPTER 6. COMING HOME

Figure 6.5: The important instruments during entry.

ergy and drag needs to be increased to come towards it. The to runway. The accelerometer on the lower left now shows
guidance box tells whether this is already happening — if z-axis acceleration rather than x-axis as during launch —
the guidance box points towards the nominal path, every- it’s a measure of how many g’s the Shuttle currently ’pulls’.
thing is fine and the high energy is already being dissipated. Finally, the vertical acceleration is a new instrument only
If that is not the case, a look on the drag scale might reveal available during entry — using it, we can monitor the tran-
that the required drag (arrow) is in fact higher than the ac- sition from falling down to aerodynamic lift starting to carry
tual drag (triangle) — it may be necessary to do a manual (at this point the value crosses zero).
takeover to correct the condition. On ENTRY TRAJ 1 we have the nominal trajectory with
If on the other hand the Shuttle symbol is right of the the Shuttle symbol showing current range and velocity and
nominal path, the Shuttle is low on energy and drag needs to triangles for past states to estimate the trend. On the far
be decreased, the Shuttle is sinking too fast. If guidance is left is an AoA indicator where the large arrow shows nom-
working correctly, the guidance box should again point to- inal AoA for Mach number and the triangle actual value
wards the center trajectory and the commanded drag lower (the fact that the nominal value is shown makes this dis-
than the actual drag, with the actual drag slowly decreasing. play more useful than the ADI). The upper horizontal scale
shows roll error. Thus, the display can in principle be used
Note: Due to the non-linear mapping of the remain- to fly manually alternatively to the ADI error needles.
ing distance on the display, the Shuttle marker po-
While initially the Shuttle won’t be seen on ENTRY
sition can be deceiving for too large ranges. If the
TRAJ 1 or will be nearly static in the upper right, the first in-
Shuttle marker moves from upper left to lower right,
dication that things are about to get interesting is the glow of
this is an indication that the range is too large to get
plasma building up outside — particularly prominent dur-
a meaningful marker position (it also means that it is
ing a night entry (see Fig. 6.7). This looks fairly spectacular,
highly unlikely that the landing site can be reached at
but is quite normal and no reason for concern. In fact, the
all).
period of maximal heating will occur soon — fairly early
To monitor an entry requires to both observe the PFD and during entry, and well before aerodynamic forces become
ENTRY TRAJ. large. The fundamental reason is that heating scales with
Let’s go over the instruments we need to watch. On the the third power of velocity whereas forces only with the
PFD, we have Mach number on the left tape and AoA next second power.
to it. However, to fly the correct AoA, we’d have to know At a dynamical pressure of 10 psf roll control will be
what the required value given Mach number is, which the taken over from thrusters by the airfoils, and at 40 psf pitch
PFD does not show. The rightmost tape shows vertical control. At this point it may be a good idea to occasionally
speed, and the ADI ball in the center is the main reference look at the MEDS SPI — if the elevons are not on average
for current attitude and provides the error needles in case close to the zero position, we should relieve trim load by
we need to fly manual. using the body flap or speedbrake (but let’s not bother with
Roughly in the center, there is a new box showing ∆Az, this now).
the relative azimuth to target. Whenver this gets too large, Soon after, you can expect the first roll reversal — the
we need to initiate a roll reversal. Below is the current range AP will command a high bank angle, possibly around 60
ENTRY 71

Why does the phase of maximal heating come long before the aerodynamic forces become large? The details of the heat
flux to the Shuttle during entry are difficult to understand and compute, but the essential properties are not. We have seen
that the drag force goes like
1 2
FD = ρvA ACD
2
where ρ is air density, vA the airspeed, A the reference area and CD the drag coefficient. This force does work, and the
basic equation for the work done is
W = FD · s
where s is a distance. The work done by reducing velocity considered per unit time t is the heating power — the thermal
energy flux jth generated by de-celerating the Shuttle. Thus,
1 3
jth = W/t = FD · vA = ρv ACD
2 A
From this we can see that while the force goes with the second power of the velocity, the heat flux generated actually goes
with the third power. Since initially we start with a high velocity, the third power becomes large for rather small densities
already, whereas even a large density can’t easily compensate for the heat flux for a reduced density.
In fact, in real parametric estimates of the entry heat, the heat flux does not scale with the density at all but with the square
root of the density, but there is no easy and compelling way to derive that. However, it magnifies the trend that max.
heating comes quite a bit before max. forces.

Figure 6.6: Instruments during the first roll reversal.

degrees. The sensation is exceedingly weird, because due there will be a roll reversal. Since during the reversal, the
to the high AoA, the motion isn’t really a roll around the full lift acts directly against gravity, the outcome is always
longitudinal axis like it would be for a plane but it has a an error to the vertical speed management which AP has to
strong yawing component. Monitor the maneuver on the reduce subsequently — so expect to see AoA modulation
PFD. and other maneuvering directly following a roll reversal.
Don’t be alarmed if the Shuttle symbol doesn’t immedi- By the time you’ve reached ENTRY TRAJ 4 (and are just
ately capture the nominal trajectory — it’s not easy and er- a few hundred miles to site), the AP should usually have
rors need to be reduced over time. If you look carefully, you managed to capture the nominal trajectory (unless your en-
can sometimes spot how commanded and actual AoA devi- try interface was substantially off).
ate slightly — that’s the AP using the ±3 deg AoA modu- In the late stage of entry, you can expect AoA to decrease
lation to change drag instantaneously. with Mach number. At this point, banking the Shuttle actu-
As you bleed off velocity, the plasma glow will dimin- ally changes course, so ∆Az changes rapidly and roll rever-
ish and eventually vanish. Whenever ∆Az reaches a limit, sals become more frequent, but bank angle less steep.
72 CHAPTER 6. COMING HOME

One thing you need to do before the TAEM phase is to

activate the ammonia (NH3) boilers. The reason is that
the flash evaporator system that cools the Shuttle no longer
functions in dense air. The controls for the NH3 boilers are
on panel L1 on the lower right. Switch at least one of the
systems on when crossing about 100.000 ft in altitude, or a
freon temperature master alarm will remind you later when
the temperature gets too high. If the ’help’ message stream
is on, there will be a reminder at TAEM interface.

Terminal area energy management

The entry phase ends at the so-called terminal area energy
management (TAEM) interface — roughly at this point we
Figure 6.7: Plasma glow during a night entry. should be some 80 miles to site, at an altitude of 85.000 ft
with about Mach 3.2. The issue at hand is that we may be
off in any of these values, but we still want to arrive at the
runway — so we need to manage energy, hence the name
of the mission phase.

You can directly start at the TAEM interface by using

--aircraft=SpaceShuttle-TAEM and setting the
landing site airport. For instance, --airport=KTTS
will give a TEAM into Kennedy Space Center. Re-
member to set landing site and runway in the menu
when doing this or you won’t have guidance!

If we’re high on energy, we need to get rid of it. There’s

a couple of things we can do, starting from liberal use of
the speedbrake via flying a steep high-drag trajectory to ex-
tending the groundtrack by flying S-turns.
Figure 6.8: Shuttle late atmospheric entry. Conversely, if we’re low on energy, the idea is to fly opti-
mum glidepath and to shorten the range to touchdown. We
can do this because the nominal trajectory is planned in a
wasteful way, so by getting rid of that extra range, we ef-
The precise nature of your entry depends on the range and fectively gain in energy.
relative azimuth from entry interface to site. A short range During TAEM, the Shuttle is pretty much flown like an
requires plenty of drag, which means high bank angles to airplane (which makes easier for a pilot to fly it by hand).
kill lift, substantial errors to vertical speed through roll re- However, the glidepath is fairly steep — the orbiter is not a
versals and more violent maneuvering. In contrast, a long very aerodynamically optimized body, the maximum glide
range usually requires an initial relatively shallow bank to ratio is of the order of 4.5 in the subsonic regime, so the
null ∆Az, and then might not require any subsequent bank- vertical speeds are fairly high — expect 7000 to 12.000
ing at all, making the experience very smooth. ft/minute sinkrate to maintain airspeed.
In any case, during the late stage of the entry, you should TAEM is a combination of both horizontal and vertical
be able to see whether you’re coming in with excess energy elements (and consequently OPS 305 shows the vertical sit-
(above the nominal trajectory) or with lack of energy (below uation and SPEC 50 the horizontal situation).
the nominal trajectory), and this is the moment where you Horizontally the idea is to change an approach which can
should decide upon the strategy for the next stage. be from almost any direction into one aligned with the run-
The navigation system will cause a few error messages way such that the final approach can be flown straight to the
when you’re about 300 miles to site, reminding you to in- touchdown point. The principle is to fly straight to an inter-
corporate TACAN and air probe data — till you’re ready to cept point at the tangent with the so-called heading align-
tackle the intricacies of Shuttle navigation, you can safely ment cone (HAC), and then fly around the HAC till the tan-
ignore them and switch perfect navigation on. gent is aligned with the runway (Fig. 6.9).
TERMINAL AREA ENERGY MANAGEMENT 73

overhead

touchdown

straight in

Figure 6.9: Overhead and straight-in TAEM patterns.

As seen from the figure, there are two possibilities to Note: What VERT SIT shows is always relative
do that for every direction of approach — the overhead to a selected horizontal pattern — guidance always
(OVHD) variant first leads past the runway and an angle assumes you will follow the planned groundtrack.
> 180 deg around the HAC, the straight-in (STRT) variant Once you deviate from the TAEM pattern specified
intercepts the HAC before the runway and uses an angle on SPEC 50, VERT SIT does not provide reliable in-
< 180 deg. formation.

The differences in flight path depend on the direction of The basic idea is to steer the Shuttle down the glidepath.
approach, but are easily several miles. Nominally the over- Above the nominal glidepath, we can push the nose down
head pattern is flown, so if the Shuttle reaches TAEM in- and increase airspeed while extending the speedbrake, be-
terface low on energy, the approach can be down-moded to low we can retract speedbrake and fly optimum glide ratio
straight-in which allows to recover part of the energy. Com- to recover some of the energy. The nominal glidepath uti-
bine this with the idea that you might choose the opposite lizes a moderate amount of speedbrake to dissipate energy,
runway direction (which again changes the path by several so again there is some leeway.
miles) and there’s a lot of leeway. Generally airspeed should not exceed 270 KEAS (the
limit line above the nominal trajectory) to avoid exceed-
The graphical display of this is provided in the Shuttle ing dynamical pressure limits, airspeed on the nominal tra-
by HORIZ SIT (SPEC 50) (Fig. 6.10). Its center portion jectory is some 230-240 KEAS and optimum glide path is
shows the touchdown point, the final approach as a straight around 220 KEAS.
line and the HAC as a circle connecting with it, all relative On the far right is an energy estimator scale. If the
to the Shuttle symbol and three predictor circles. The idea triangle is above the line marked as STN (S-turn), then
is to aim the predictors at the HAC tangent, then fly around guidance recommends to not fly straight to the HAC inter-
the HAC till the final approach is reached, zero bank angle cept but rather approach in a wavy line to increase length
and descend to the runway. of the flightpath and deplete more energy. If the triangle
falls below the line marked as MEP, guidance recommends
In addition, SPEC 50 always shows the selected runway
to down-mode the approach to a so-called minimum entry
direction (items 3 and 4) and the selected pattern (item 7)
point.
and can be used to down-mode the pattern when low on en-
ergy. The lower portion is dedicated to TACAN information Following nominal procedures, the head-up display
filtering — the useful readings are azimuth to next guidance (HUD) is first switched on and used to monitor the flight
point and range (though these are duplicated on the PFD). during TAEM. The corresponding symbology is shown in
Fig. 6.11. The central part is the flight director cage with the
Now let’s focus on the vertical part of the pattern which guidance diamond — this is a fly-to indicator which com-
is concerned with energy management (Fig. 6.10). mands pitch and roll maneuvers with the aim to bring the
diamond inside the cage.
The center portion shows a nominal glidepath as altitude Left to it is the Nz indicator measuring forces. A pitch
vs. range (this is different from the entry displays). and roll ladder indicates attitude. A cross marks the bore-
sight, i.e. the direction the nose of the Shuttle is pointing.
74 CHAPTER 6. COMING HOME

Figure 6.10: Horizontal and vertical situation displays during TAEM.

designed for hypersonic aerodynamics are very powerful in

the lower atmosphere. At the AoA characteristic for TAEM,
the stabilizer fin is finally in the airstream, restoring dynam-
ical yaw stability, and the rudder becomes effective.
The aerojet DAP controls rudder to minimize beta
(sideslip), i.e. unlike in many airplanes, there is no reason
to use the rudder pedals during a turn, and in fact doing so
will cause some amount of sideslip.
The nominal TAEM exit point is seven miles to touch-
down at an altitude of 12.000 ft, aligned with the runway.
Use the HSI needle to perfectly align the Shuttle with the
runway during the last stage of the HAC.

Final approach
The final approach phase is the only mission phase where
you must rely on your piloting skills — like in reality, there
Figure 6.11: TAEM mode of the HUD. is no automatic mode available to do it for you, and the
workload during the final thirty seconds is moderately high.
Also, unlike practically any other mission phase, this one
is flown more by looking outside than by instruments. As
The far left is filled by an airspeed tape showing KEAS, and usual, practice makes perfect.
the far right has the altitude tape showing thousands of feet.
Text on the bottom indicates which of the guidance You can directly start into the final approach by us-
phases — S-TURN, ACQ (waypoint acquisition) or HAC ing --aircraft=SpaceShuttle-approach and
(heading alignment cone) — is currently active, below is the setting the landing site airport. For instance,
--airport=KEDW will give the approach into Ed-
control mode (AUTO or CSS). The lower right has a speed-
brake indicator showing actual and commanded speedbrake wards Airforce Base.
position. The final approach is subdivided into four guidance
The autopilot is able to fly the complete TAEM pattern phases — outer glideslope (OGS), pre-flare (FLARE), in-
right to the HAC tangent into the final approach, however ner glideslope (IGS) and final flare (FNLFL) — the desig-
nominal procedure is for the Commander to take control nations in brackets are the HUD designations for the phases.
once the Shuttle goes sub-sonic. While guidance is available for these phases, there is no au-
The flight characteristics during TAEM are very good, topilot taking over controls, i.e. the Shuttle must be flown
though control inputs need to be gentle. The huge elevons manually following the guidance symbology.
FINAL APPROACH 75

will lead to an airspeed increase — it is necessary to use

the speedbrake to maintain airspeed around 300 KEAS. If
speedbrake is set to AUTO, the software will do this au-
tomatically, if the pushbutton is set to manual, the throttle
controls command speedbrake.
As soon as you are on the glideslope and have the real
runway in view, de-clutter the HUD (see Fig. 6.13) by press-
ing Shift + h . The first declutter will make the virtual
runway vanish, the second will remove the pitch ladder ex-
cept for the horizon line and replace the KEAS and alti-
tude tape by alphanumerics — we’ll need those for preci-
sion during flare.
If everything is nominal, at some 5000 ft above ground
radar altitude should become available. This is indicated by
an ’R’ after the HUD altitude string. At 3000 ft, arm gear
( ctrl + g ). There will be an on-screen message to remind
you if the ’help’ message stream is active. Do not forget
this step otherwise the landing gear won’t deploy later, and
nothing is quite as embarrassing as having come all the way
Figure 6.12: Approach mode of the HUD, full symbology.
from space to do a belly-landing!
Between 3000 and 2000 ft, expect a second set of tri-
angles approaching the flare triangles from below as a re-
The main aid for the approach and landing phase is the minder that you’re supposed to pull into pre-flare soon. As
HUD (Fig. 6.12). Its symbology differs from TAEM in sev- soon as the triangles merge (at 2000 ft), pull up such that
eral important points: First, the flight director/guidance di- the velocity vector follows the triangles. This aims to re-
amond combination is changed into a velocity vector sym- duce sinkrate from 11.500 ft/min or 190 fps to about 720
bol. Unlike the flight director, this moves across the HUD ft/min or 12 fps. As soon as you fly shallower than 12 de-
and points to where the Shuttle is currently going (i.e. aim- grees, the Shuttle will lose airspeed. Smart braking logic (if
ing it at some point on the runway will make the Shuttle hit the speedbrake is on AUTO) will decide upon a speedbrake
that point). setting during this phase and keep it till touchdown — this
The velocity vector is accompanied by flare triangles affects the length of the final flare.
which always indicate what descent angle to fly — the idea
is to keep the symbol between the triangles (note that the
velocity vector usually moves very little left and right on
the HUD because the Aerojet DAP automatically maintains
zero sideslip).
At lowest de-clutter level, the HUD displays a virtual run-
way overlaid on the real one (in case clouds block the view
when the approach is begun). Comparing the real with the
virtual runway allows an immediate check if the state vector
is good.

Note: If real and virtual runway do not agree, the

navigation state is bad and guidance symbology may
not be used.

The virtual runway also contains an aim point and a

touchdown point. To start the final approach, pitch down
and aim the velocity vector at the aim point, then fine-tune
the outer glideslope with the help of the flare triangles (they
will steer you on the optimal glideslope). The descent angle
is 17-20 degrees — this is an extremely steep glidepath if
you’re used to the visuals of e.g. airliner approaches which Figure 6.13: Decluttered HUD commanding flare.
typically follow a 3 degree glidepath. The steep descent
76 CHAPTER 6. COMING HOME

If you follow the flare triangles faithfully (they actually

adjust to what you’re doing, if you delay the flare they’ll
command more crisp movement) you should come out of
the final flare at between 300 and 600 ft above ground (the
HUD changes phase to IGS), and this is also the moment to
deploy gear. Since gear is a powerful drag device, expect
additional strong reduction in airspeed. Maximal gear ex-
tension speed is 312 KEAS, at this point you ought to be
well below that.
Typically there should be no more than a couple of sec-
onds on the inner glideslope before you’re low enough for
the final flare.
As soon as the HUD changes to FNLFL and the flare
triangles move up, start the final flare, carefully monitoring
radar altitude to gauge descent rate. Your aim is to reach
zero altitude when the airspeed has dropped to around 200 Figure 6.14: A safe touchdown in Vandenberg.
KEAS. Despite a powerful ground effect, expect that the
nose needs to go 12 degrees above the horizon. This will
mean you have poor view on the runway. Also, handling Training TAEM from the carrier air-
characteristics deteriorate rapidly below 230 KEAS, so it
would not be a bad idea to touch down a little faster on the craft
first few tries.
You can experience the first glide tests done with the Shuttle
Note: If you did not de-clutter the HUD, you will and in that way also train the later part of a TAEM pattern
find it extremely difficult to flare properly as the exact by using the glide scenario
distance to the ground is very hard to gauge by eye --aircraft=SpaceShuttle-glide
in the touchdown attitude.

The nominal touchdown point is, unlike for airplanes,

about 2500 ft behind the threshold. When weight on the
rear wheels is sensed, the speedbrake is automatically de-
ployed to full. Activate the drag chute ( shift + c or arm
and deploy pushbuttons left of the HUD) as soon as you’re
sure you won’t bounce, wait till airspeed has dropped to 180
KEAS to initiate a gentle de-rotation, i.e. bring the nose
down and let the nose gear carefully take weight.
Only when all three gears are on the ground, apply brakes
( b ) and use nose wheel steering to correct any errors of
the rollout path. Remember to jettison the drag chute ( c
or pushbutton right of the HUD) at 60 kt, otherwise it may
damage the engine bells.
It usually takes quite some distance to bring the Shuttle Figure 6.15: Riding the Shuttle Carrier Aircraft.
to a complete stop. This is the reason the nominal landing
sites have rather long runways. Trying to get the Shuttle to
fit into a shorter runway like for an emergency landing can You will find yourself sitting on the back of the Shut-
be a bit of a challenge — we’ll cover the techniques used tle Carrier Aircraft (SCA), a modified Boeing 747 (see
for that later. Fig. 6.15). While riding the SCA, the Shuttle controls won’t
The limit simulation during touchdown is fairly realistic, work, you have to go where the SCA carries you — but us-
i.e. you need to keep vertical speed limits rather tightly ing Carrier Aircraft → Controls you can click on the left
or you will experience burst tires or broken gearstruts and and right arrows to bank the SCA into the direction you
difficulties in controlling the rollout as a result. want to go.
The simulated mission ends with the call ’Wheels stop’ On the VERT SIT and HORIZ SIT displays, you can fol-
— there’s no ground checkout procedure included. Con- low where you are relative to a nominal TAEM and let the
gratulations — you’re back to Earth! carrier deliver you right into the patttern. The ride is per-
MANUALLY FLYING AN ENTRY 77

haps a little bumpier than in reality, which has to do with The simplest way is of course to follow the guidance nee-
the difficulties of dragging the simulated Shuttle along with dles on the PFD, but that doesn’t train you for the situations
external forces while tightly coupling it tightly to the carrier where you need to take over manually — so let’s not do that
position, but it works. and learn how to correctly plan the entry from scratch.

• First, it is easier to learn the procedure by only tak-

ing the roll channel to manual and leaving the pitch
channel under automatic control. In this way, you can
spend all your time on controlling the trajectory —
for the most part, pitch is a pre-programmed Mach-
dependent target value that just needs to be held.

• When q passes 10 psf trajectory management can start.

Roll the Shuttle to a high bank angle (usually 50-70
deg will be necessary) and monitor the vertical accel-
eration scale on the lower right corner of the PFD. Stop
the roll at the bank angle when this reads zero.

• Now look at the Shuttle symbol relative to the nomi-

nal trajectory and what the guidance box is doing. If
Figure 6.16: Separation from the SCA. you’re above the trajectory, you need more drag, i.e.
sink more — so roll to a somewhat higher bank angle.
If you are below the trajectory you need less drag, so
Once the carrier is where you need it to be, you can ini- roll back to a somewhat lower bank angle.
tiate the pitchdown maneuver — either by Shift + i , or
by the button on the control menu. The carrier will begin to • Drag is slow to adjust, so wait for at least 15 seconds
pitch down, and once the attitude is okay for separation, a to gauge the effect of your maneuver. If e.g. you sink
callout on-screen (and spoken if text to speech is activated) faster, drag should be increasing and the guidance box
will inform you. should indicate your progress back to the center. If
The Shuttle is separated the same way from the carrier that doesn’t work, or doesn’t work fast enough, add
bank angle (or reduce if you’re below). Cross-check
and from the ET, i.e. you can either use d to separate,
with the drag scale on the left of ENTRY TRAJ.
or you can switch the ET sep mode on panel C3 to MAN,
remove the switch guard and press the separation button. • When you are reasonably close to the center trajectory,
Following separation (see Fig. 6.16), you can expect a roll again to the bank angle at which the vertical accel-
callout that you’re vertically clear of the SCA and can start eration is zero — apparently your current sinkrate is
to maneuver, while the SCA will continue the descent (if good enough.
you do not separate, you will eventually be flown into the
ground, there is no detailed simulation of the SCA flight • The sinkrate required to follow the nominal entry tra-
pattern provided — likewise you can’t chase the SCA to a jectory changes quite a bit in different regimes, so be
landing). prepared to follow it.

• Monitor the ∆Az periodically — when it grows to-

Manually flying an entry wards about 15 deg or -15 deg, you need to execute a
roll reversal, i.e. roll to the same bank angle in oppo-
Nominally the entry is flown by the autopilot and the crew site direction. So if you’re currently 70 degrees banked
just monitors its action. And at first it might seem that this right, you want to end up 70 degrees to the left. The
is a near-impossible task to do manually because there are roll reversal should be executed reasonably fast, but
so many factors to consider and the whole maneuvering is you don’t need to yank controls all the way.
very different from how an airplane is operated. However, it
isn’t actually that hard to learn and can be a rewarding expe- • During the roll reversal, the Shuttle generates more lift
rience. In particular, if the flightpath prior to entry interface than when banked steeply, so when the maneuver is
is unusually steep or shallow, it might even be necessary to over you’re typically above the desired trajectory — so
fly the first portion by hand to prevent the AP from mis- expect the need to increase the bank angle some more
judging the situation. initially to get back down.
78 CHAPTER 6. COMING HOME

• After a while, you’ll get the hang of how long it takes

for the bank angle to translate into a drag variation and
how quickly you need to maneuver.

• When you’re ready, you can also take the pitch channel
on manual. The time to do pitch corrections is while
you wait after a roll maneuver to show its effect on
drag. Take a quick look at the pitch scale on the left
side of ENTRY TRAJ and correct such that the triangle
(actual AoA) move right on top of the arrow (desired
AoA). The AoA is, especially during the early entry
phase, rather crucial for temperature control, so if you
neglect to correct pitch, chances are high that you burn
up in the atmosphere.

• Do not forget the roll reversals, if you allow ∆Az to

grow too large, it will be impossible to turn back in
time and you won’t be able to reach the landing site.
At hypersonic velocities, turning around is a very slow
process.

• Finally, just prior to TAEM transition at the end of EN-

TRY TRAJ 5 you can pitch the Shuttle down into the
small AoA region and start flying it like a regular air-
plane.
Chapter 7

Troubleshooting

The nominal procedures and checklists are designed to The Shuttle doesn’t hold a clean trajectory and
operate the Shuttle trouble-free, but sooner or later you’ll crashes after liftoff
probably slip in procedures and encounter a problem. You
will find the most common ones summarized below. If you intend to fly autopilot, check that

• the launch guidance dialog has a valid target and reads

Launch ’active’
The Shuttle spawns inside the pad, or lying flat next to • the PITCH an YAW/ROLL pushbuttons on panel F2
the pad are on AUTO

This is a bug caused by a timing issue combined with • throttle pushbutton is on AUTO
the fact that FG does not natively initialize objects on their
tail — the pad needs to be placed into the scenery by the If you’re trying to fly manually, check that
FG model manager, and upon startup, the software requests
this. However, if you are on a slow computer or try to start • the PITCH an YAW/ROLL pushbuttons on panel F2
in an area densely populated with other objects, the pad is are on CSS
inserted into a queue and may not yet be ready by the time
the Shuttle is placed. • throttle pushbutton is on MAN
Try to
• your control device is properly configured and cen-
• start FG paused to allow the model loader queue to tered upon liftoff
empty
• auto-coordination is off
• start in a less densely populated area

Engines do not ignite and a launchpad abort is called For instance, non-centered pedals sending a rudder
signal upon liftoff can kill you.
There’s insufficient pressure in the hydraulics system to
operate the main engine valves or gimbal the engines — in The Shuttle breaks apart a minute into the flight
that condition, launching would be suicidal, so the launch
is aborted. This happens due to aerodynamic overstresses during the
Check that early ascent and can either be a violation of dynamical pres-
sure or wing bending moment. Check that
• all three APUs are running (APU/HYD MEDS page)
• all three engines were throttled to 67% of rated power
• hydraulic main pump pressure switches on panel R2
about 30 seconds into the flight — if not, likely no AP
are in ’high’ position
is engaged or auto-throttle is switched off
• hydraulic pressure in all three systems is larger than
2600 psf (APU/HYD MEDS page) • all pitching maneuvers were initiated and ended gently

• unlimited fuel (called ’fuel freeze’) is not enabled in • no large AoA was forced prior to max. dynamical
the FG options pressure

79
80 CHAPTER 7. TROUBLESHOOTING

The Shuttle breaks apart during late ascent Insertion

This likely happens due to structural overstress by accel- The OMS burn result is different from the target
eration forces. Check that
This means that either the computation is wrong or the
• all three main engines are throttled back when acceler- OMS has lost an engine. Check that
ation reaches close to 3 g during late ascent
• the current weight of the Shuttle is entered correctly
• all maneuvers were initiated and ended gently, espe- on the OMS maneuver display (if you’re flying with
cially when linear acceleration is already large it’s easy realistic navigation)
to violate limits by hard maneuvers
• both engines reach 100% of rated thrust during the
Autopilot steers the Shuttle westward and fails to burn
reach orbit
If an engine is out, check the cause (not armed, no fuel,
When starting up FG, you did not pass a --heading op- damaged,. . . ) and conduct a single engine OMS burn in-
tion to orient the launch stack properly. As a result, the tail stead if needed.
fin of the Shuttle points westward, and pulling back after Also note that small deviations to the burn target (e.g. a
liftoff leads to the wrong launch azimuth. There is usu- 1-2 miles in apsis for a long burn) are normal — null them
ally not enough performance to reach a retrograde orbit, so using the RCS in translational mode once the OMS has
you will end up on a ballistic trajectory back into the atmo- finished.
sphere.
OMS engine does not ignite or does not reach suffi-
• Make sure to start FG with the proper orientation for cient chamber pressure
the launch pad by passing a heading parameter west-
ward — this will point the tail fin eastward. The engine may be damaged or not receiving enough
fuel. Hydrazine freezes at around 35 F, so insufficient heat-
APU overheating occurs, followed by loss of control ing of the fuel lines may be the cause, in this case the
function can be restored. If there is a problem with tank
The APUs need to be cooled by water spray boilers and pressure, a crossfeed may be needed. Otherwise the engine
overheat if that is not done. Once the APUs fail, there is may have genuinely failed and a single engine or RCS burn
no hydraulic pressure left to gimbal or throttle the main en- needs to be planned. Check
gines. Check that
• OMS tanks have both fuel and oxidizer
• all three water spray boiler controllers on panel R2 are • OMS propellant tanks pressurized
switched on
• tank isolation valves open
• the water spray boiler N2 valves are open
• OMS fuel line heaters on
• APU temperatures on APU/HYD MEDS do not rise
into the red region • on DISP 89, OMS fuel line temperatures above 35 F

ET can’t be separated after MECO • if you just switched heaters on, allow for a few minutes
to thaw lines
It may sometimes occur that the AP tried a last-second • OMS engine switches are in ARM/PRESS position
correction before commanding MECO, in this case the
Shuttle will have non-zero attitude rates at MECO. This will • OPS page reads MNVR EXEC
inhibit automatic ET-separation. There are two ways to deal
with this issue:
On orbit
• either transit to MM 104 and wait for the RCS to null
rates, then initiate the normal ET separation d The Shuttle does not respond to control input

• or force ET separation overriding the rate veto by the Most likely it’s an orbital DAP configuration issue.
GPC Ctrl + d Check that
ENTRY 81

• Orbital DAP pushbutton of the control section is not Payload doesn’t attach
AUTO
Payloads in the bay can only be grabbed under the cor-
• you have taken RHC rather than THC (toggle via m ) rect angle, for most the end effector needs to be pitched
down. Simply maneuvering the effector to the right point
• you are in MM 104 or higher and ET has been discon- is not enough. Payloads in space are difficult to grab as the
nected Shuttle is always moving ever so slightly and the tolerance
is small— be persistent, this is difficult to do in reality just
• flight controller power switch is on
as well.
• RCS has propellant
• correct attachment angle when grabbing objects inside
• your joystick is plugged in the bay
• correct attachment point when grabbing objects in
The payload bay door does not move space

This is either an electrical or a mechanical issue. Check Retention system does not operate
that
Check that
• Ku antenna, RMS arm and any payload are properly
stowed • payload bay mechanical power switch is on
• payload retention system logic power is on
• the door actuators are powered
Avionics bays overheat
• there is no strong temperature difference between left
and right side of the Shuttle that causes tensions Either the water or the freon loops are not operational and
therefore no heat is carried from the avionics bays or no fan
RMS arm does not move in the avionics bay is operational (one should be running
at all times). Also, power-hungry equipment like the engine
This may have a couple of reasons. Check controllers need to be off, the radiator is unable to vent their
large heat load. Check
• RMS arm power
• payload bay door open
• RMS arm joint temperatures — joints do not drive in
cold temperatures, operate the heater • radiator loop controllers on

• MPM deployment status — arm needs to be properly • bypass valves not on bypass
deployed and unlatched • at least one water pump on
• no saving or braking option engaged • freon pumps on

• auto sequence switch is in ’proceed’ when driving auto • difference between freon in and out temperature
sequences
• avionics bay fans on
• joint positions outside software or hardware stop re- • power consumption of the electrical system
gion
Switch the flash evaporator system on and put it into hi
Payload is attached but can not be moved load operation if you need to cool down quickly.

Most likely it is still held in place by the retention system

— check Entry
• payload retention system powered Shuttle loses control shortly after q has passed 10 psf.

• payload bay mechanical power on It’s very likely you didn’t trim properly before entry — if
there’s too much fuel in the rear OMS tanks, the control sur-
• all latches of the chosen payload released faces don’t have enough authority to control pitch. Check
82 CHAPTER 7. TROUBLESHOOTING

• OMS fuel largely gone after OMS burn (if not, dump) Note that if you are too far from the runway, you may
not be able to get meaningful guidance information.
• aft RCS fuel sufficient

• CoG in the valid range for entry (query MCC if neces- HUD doesn’t show guidance symbology
sary)
Likely either no landing site is selected, or the selected
No Shuttle symbol on ENTRY TRAJ displays airport is not supported as landing site by guidance. For few
regular landing sites, auto-detection is implemented, i.e, if
This either means there’s no landing site set — in which you start an approach into Kennedy Space Center (KTTS)
case guidance doesn’t know the range to what should be guidance is automatically on and set to KTTS, but this does
computed and can’t display anything, or that your actual not work for other airports. Check that
range at the velocity range of the display is off so much the
symbol isn’t shown. Check • airport and runway are in the list of supported sites (see
appendix B)
• landing site and runway selected on SPEC 50
• entry guidance is set to the right airport/runway (SPEC
• current range to landing site about 2000 miles or less 50 or the GUI)

Shuttle symbol on ENTRY TRAJ displays moves Gear can’t be deployed

backward
Part of the procedure is to arm gear at 3000 ft altitude, if
Normally the Shuttle symbol should move from upper the gear is not armed, it won’t deploy. Check for
right to lower left. If you see movement from left to right,
this is a sign that you’re too far from the landing site. Check • Gear armed talkback or gear armed message on FG-
native HUD.
• range to landing site
Gear was out too early and can’t be retracted.
Shuttle breaks apart with a thermal protection
system failure It really can’t in reality, this is done by the ground crew.
The landing gear is a powerful drag device and nominally
This happens due to on of two possible reasons. Either it should be deployed a few seconds prior to touchdown, at
a properly working TPS was overstressed by flying a bad between 300 and 600 ft altitude. Deploying it too early can
AoA or a too steep entry trajectory — or the TPS was not be a fatal mistake which can not be corrected.
properly working in the first place. Check that
Gear breaks during touchdown
• payload bay door is closed during entry

• ET umbilical door is closed Touchdown really needs to be very gentle, the Shuttle is
fast and heavy and tires and gear struts are not reinforced.
• star tracker doors are closed Practice makes perfect. Check that
• AoA is no more than 3 degrees different from the nom- • vertical speed at touchdown does not exceed 9 ft/s
inal AoA displayed on ENTRY TRAJ (that’s 540 ft/min)
• no agressive drag increase is flown during high heating
It takes some practice to do a good touchdown, don’t be
discouraged if this doesn’t work the first few times.
Approach and touchdown
Drag chute does not deploy
DPS never leaves ENTRY TRAJ 5
The drag chute deploys only if weight is sensed on all
Usually the change to MM 305 should occur automati- rear wheels, as mistakenly deploying it in air would be
cally at TAEM interface, however sometimes this doesn’t fatal. It is quite possible during a landing that only one
happen. This is no particular reason for concern, simply do wheel is on the ground yet. To prevent this, plan a few
the OPS transition manually typing seconds of coast phase during which the Shuttle rolls and
loses airspeed (and lift) before trying to deploy the chute.
• OPS 305 PRO
APPROACH AND TOUCHDOWN 83

Shuttle overshoots the runway

The Shuttle is not decelerating very fast once on the

ground, which is the reason the nominal landing sites have
very long runways. This can be a problem if you try to get
into an airport with a relatively short runway. Check that

• Airspeed is around 200 KEAS at the touchdown point

• Drag chute is deployed after touchdown
• Full brakes are applied after weight on nose gear
84 CHAPTER 7. TROUBLESHOOTING
 Part II

Advanced topics

85
Chapter 8

Nominal procedures

This chapter summarizes various procedures which have Procedure

not been mentioned earlier but which may or may not be-
• Close the receiving side tank isolation valves to dis-
come relevant during some part of a regular mission. They
connect the tank which can not supply propellant from
all have in common that they don’t deal with an acute emer-
the manifold. That is, if you want to feed left RCS
gency and hence do not have to be executed in a time-
from right RCS, close the left RCS tank isolation valve
critical fashion.
both for manifolds 1/2 and the dual valves 3/4/5 A and
3/4/5 B.

• Open the supply-side crossfeed valves. In the above

Crossfeeding example, these are right RCS crossfeed valve 1/2 and
3/4/5.
Background
• Check the receiving side tank isolation valve talkbacks
Nominally, each OMS engine or RCS thruster module feeds to read CL.
from its own fuel and oxidizer reserves. However, it can
happen for a couple of reasons that these are not available. • Check the supply side crossfeed valves to read OP.
For instance, the RCS propellant may deplete due to exces-
• Open the receiving side crossfeed valves — here left
sive maneuvering while there is still OMS propellant left.
RCS crossfeed valve 1/2 and 3/4/5.
Or the helium pressure might have dropped to the point that
the tanks can no longer be pressurized. Or, due to an OMS All fuel is now drawn from the supply side, in the above
problems, a de-orbit burn must be made with the RCS en- example both left and right RCS jets are fed from the right
gines, in which case there is usually not sufficient RCS pro- RCS tanks.
pellant for the burn duration needed. In such cases, a cross
feed procedure may be required. Note: Do not connect tanks directly during cross-
In general, cross-feeding works between left and right feedings, as they may be pressurized to different lev-
OMS, between left and right RCS and from OMS to RCS, els. Rather, follow the correct sequence to make sure
but does not work from RCS to OMS (the propellant only one tank acts as supply for one or more mani-
throughput of the RCS feeding lines is not large enough folds to avoid damage.
for the needs of the OMS engines). There is no capability
to cross-feed the forward RCS, if its propellant flow ceases If you want to end a crossfeed, follow the inverse order —
it is unavailable, but the forward RCS module is also not first close all crossfeed valves, then open the tank isolation
essential for attitude control, and most importantly it is not valves on the former receiving side.
needed during entry.
Cross-feeds are set up by closing the isolation valves
and opening the cross-feed valves or a manifold. The
Mounting the COAS frame
relevant switches can be found on the overhead pan-
els (detailed cockpit only) or in the menu dialog
Background
OMS/RCS Propellant Flow . The crew optical alignment sight is a device similar to a
HUD that uses a combiner to project a recticle pattern into
infinity. Looking through the combiner, it is possible to get
An in-sim checklist for cross-feeding is available. a calibrated view along one of the Shuttle’s body axes, or to

87
88 CHAPTER 8. NOMINAL PROCEDURES

use the pattern to measure an angular distance. For instance, IMU alignment using COAS
measuring the apparent size of a rendezvous target with a
known size allows to estimate the current distance to the Background
target, keeping the target in the center of the pattern allows
to fly a precise alignment of the Shuttle with respect to the The attitude shown by the PFD is derived from a set of gy-
target. In addition, the COAS is used to establish an intertial roscopes which sense any rotation around one of the axes,
reference frame using star sightings if needed. the inertial measurement units (IMUs). However, this pro-
cedure is not perfect, and so the attitude maintained as part
The device itself can be mounted when needed in for- of the state vector starts to drift over time. This would lead
ward position, next to the CDR HUD looking forward, or in to a wrongly indicating PFD and worse, poor guidance dur-
upward position in the rear of the flightdeck. Of course the ing entry.
mounting procedure needs to be done with care such that Star tracking cameras pointing left (-Y) and upward (-Z)
the view through the device is indeed correctly calibrated are used to correct such attitude drift in orbit. They have
to represent the Shuttle’s body axis. In both positions, the a catalogue of bright stars in memory, and whenever the
device is powered from the AC3 bus. COAS is kept stowed camera happens to have such a star in view, the logic com-
during launch and entry. pares actually seen position with the position expected if the
internal attitude model would be correct and automatically
applies the required corrections.
It may however happen by chance that the star trackers
do not have any bright star in view for an extended period
of time, or that one of the tracking cameras is off, or not
operating because it is pointed at Earth or the Sun. In this
case, attitude errors accumulate, and they can grow to the
point that the star tracker can no longer be sure whether a
star seen really is the star expected or not — in this case,
the trackers report FALSE TRK as their status.
Once that happens, the crew needs to correct for the at-
titude error manually. The principle is the same as for the
star trackers — the Shuttle is pointed at a bright star, once
the body axis is aligned with the star a mark is taken, and
from two such marks on identified stars, the error is derived.
This is known as the crew optical alignment sight (COAS)
procedure.
Figure 8.1: The foreward COAS device plugged in. A COAS procedure can only be done during the night
portion of the orbit when the stars are out and require that
the crew can recognize constellations correctly.

There is a failure scenario to simulate attitude errors

to the point that star trackers no longer function. Sim-
Procedure ulating attitude errors is only done if realistic rather
than perfect navigation is chosen.
The device is mounted in the chosen position using the sim-
ulation options menu. From there, it needs to be powered
on using the switch next to the plug (make sure the AC3 bus Procedure
is powered). If everything is set up correctly, there should Sights can be taken along the -X axis by pointing the bore-
be a recticle pattern visible when looking through the com- sight marker of the Shuttle HUD at a star, by using the for-
biner glass. The view manager ( Ctrl + v ) has suitable ward COAS recticle or along the -Z axis using the aft COAS
views defined for both sighting positions. device.
The large wheel can be used to adjust pattern brightness
to the outside lighting conditions. To point e.g. the Shut- • Wait for the night portion of the orbit when stars come
tle’s -Z axis towards a specific star, mount the device in aft out
position, look through the glass and maneuver the Shuttle • Study the constellations and look for a major star you
till the star is centered in the recticle. The forward COAS recognize — use a star atlas if needed
and the HUD boresight both are aligned with the same body
axis, so either device can be used. • Open GNC SPEC 22 via SPEC 22 PRO
IMU ALIGNMENT 89

• The display will now show ∆BIAS values which tell

you how much the state vector would be corrected if
the marks you have taken are used. You can expect to
get an accuracy of perhaps 1-2 degrees by pointing this
way, so if the values show much more, it likely means
you’ve been taken the attitude reference for the wrong
star

• Once you have sensible ∆BIAS values, you can update

the state vector via item 28. This should bring the star
tracker back into operating accuracy.

Note: Updating the state vector with any mark, how-

ever bad, will zero the errors. The reason is that the
update forces the state vector to agree with the new
Figure 8.2: COAS functionality on SPEC 22. input, so by definition the errors get zero. Don’t think
of the shown error as the difference between reality
and instruments — this is never shown. The error is
• Enter the ID of the first star you want to use as refer- merely the difference between what the state vector
ence as item 21. The list of supported stars is given is before the measurement and what the measurement
below says — none of which has to be true.

ID star
The star ID is also a valid target ID to be used in OPS
11 Shedir
201 as item 8 (tracking target) — using star tracking,
12 Mirphak
you can automatically point the Shuttle towards the
13 Mizar
approximate direction of a reference star prior to a
14 Arcturus
COAS — of course you will be off by the attitude
15 Betelgeuse
errors.
16 Procyon
17 Spica
18 Antares IMU alignment
19 Canopus
20 Sirius Background
As mentioned above, usually the star trackers are used to
• Select the forward COAS reference (item 26) or aft
provide attitude fixes. However, they do not automatically
COAS (item 27), dependent on what you want to use
fix attitude problems, but rather the three inertial measure-
• Maneuver the Shuttle such that the bore sight of the ment units need to be re-calibrated periodically (usually ev-
HUD or the recticle center points at the chosen star ery 1-2 days) based on star tracker information. This is done
on the SPEC 21 page (see Fig. 8.3).
• Press Ctrl + a or the ATT REF pushbutton to mark The IMUs provide attitude by virtue of being an inertially
the star when it is centered — the GPC stores the non-rotating gyroscopically stabilized platform. Thus, by
sensed gyro positions at this point measuring the angles on the (rotating) IMU frames, it is
possible to deduce the orientation of the Shuttle in space via
• You need a second star to compute an attitude, so if you a transformation matrix. However, the gyroscopes are not
think the mark is good, accept it by item 23, otherwise perfectly stabilized platforms, they drift over time, which is
you can mark again using Ctrl + a and overwrite the what the star trackers aim to correct.
previous mark Using star pointing angles from the tracker thus provides
information on what the true vehicle attitude is and (via an
• Once the first mark is accepted, enter a second star via inverse transformation matrix) what the IMU frame angles
item 21, ideally far in angle from the first one (this should be, compared with what they actually are.
will give better accuracy). Take a second mark with There are now two ways to make use of the star pointing
this star, then accept that as well using item 23 once vectors: Either, one can physically rotate the IMU angles
you’re satisfied such that the actual frame angles are aligned with what they
90 CHAPTER 8. NOMINAL PROCEDURES

• If the star tracker is operating properly, there should be

three stars in the star table on the upper right. Select a
pair of them (with a good separation angle) using items
17 to 19 — each of the selected stars will be marked.
• If the star table is not filled, check star tracker oper-
ations. You can start by commanding a self-test via
items 1 and 2 for each of the cameras. Also, check
whether the system is powered and doors are open.
Now the pointing vectors are available to the GPC, so we
can proceed with the actual IMU alignment.
• Bring up the IMU ALIGN page SPEC 21 PRO
• Select the IMUs you want to align via items 10 to 12,
Figure 8.3: The SPEC 21 page. usually you would do all three
• Make sure the alignment method is set to 13 STAR
ALIGN
are — or one re-computes a transformation matrix such that
• On the lower left there should now be ∆ angles shown
the actual angles do not change but the correct Shuttle atti-
— these are the angles by which the IMUs will be
tude is still derived from them.
moved. Usually they should be similar in magnitude,
Rotating the gyroscopes around isn’t easy though - if just one of the set is dramatically different to the oth-
there’s a fast but inaccurate method called slewing which ers, consider IMU to IMU align (see below)
can achieve about 1.2 deg/s of rotation and a slow but accu-
rate method called pulse-torquing that achieves 0.007 deg/s. • Once you are satisfied, do ITEM 16 EXEC — you
In contrast, re-computing a matrix works essentially instan- should now see the ∆ angles slowly reduce.
taneously.
• The procedure will terminate on its own and wipe the
However, it is not unproblematic. Any gyroscope can
star table on SPEC 22 as a precaution.
not properly sense attitude at the Euler singularity — imag-
ine an aircraft pitched up to 90 degrees, and one can no In case you need to do a matrix alignment, rather than se-
longer tell a difference between roll and yaw. This is called lecting item 13, use item 15 before doing ITEM 16 EXEC
a gimbal lock. To avoid gimbal locking, the three IMUs are — you will see the ∆ angles vanish immediately.
all oriented in slightly different attitudes, so normally they The IMU to IMU alignment procedure is slightly differ-
can’t all reach gimbal lock a the same time. ent. For this, you do not need to select any stars from the
When errors are corrected in the transformation matrix, star table.
the singular points (at 90 degrees pitch up and down) ex-
pend into singular rings at somewhat lower pitch — and • Select the IMU you want to align via items 10-12
eventually, a situation may arise in which all three IMUs go
• Select the IMU you want to use as reference via ITEM
into gimbal lock at the same time, at which point the Shuttle
14 + X EXEC where X is the reference unit
has no longer valid attitude information.
For that reason, the preferred method is physically align- • Now do ITEM 16 EXEC — you can now observe the
ing the units using pulse-torquing rather than re-computing ∆ angle move fast while the unit is slewing, and then
the matrix. Only in situations where a quick result needs slowing down when pulse-torquing is used as the angle
to be achieved should slewing or matrix re-computation be decreases.
used.
• The procedure will also terminate on its own.

Procedure In case you need to end the alignment procedure prema-

turely, use ITEM 17 EXEC. As mentioned before, the mis-
IMUs can be aligned to another IMU (this is the only case alignment can grow so large that the star tracker no longer
where slewing is used) or to star tracker data. Since nor- provides pointing vectors — in this case, a COAS procedure
mally we want to use star tracker data, we need to make has to be used. Executing the COAS update implicitly uses
sure we have it first. a matrix alignment, so a COAS should always be followed
by a proper IMU alignment once the star tracker provides
• Open SPEC 22 via SPEC 22 PRO pointing vectors again.
FUEL CELL PURGE 91

Fuel cell purge

Background
Electrical power for the Space Shuttle orbiter is created by
three fuel cells which generate electricity by converting the
reactants hydrogen and oxygen into water (which is then
utilized to satisfy the Shuttle’s water needs).
Normally the fuel cells operate as a closed loop, i.e. the
reactants are cycled through the stack, the water vapour is
removed and the gases are fed back with nothing venting
overboard. Over time this leads to a deterioration of the fuel
cell power output as inert gases and other contaminants ac-
cumulate in the stack and prevent the desired reaction from
occurring.
For this reason, fuel cells need to be purged periodically.
During a purge, the cell changes into an open flow where Figure 8.4: Fuel cell status on SM DISP 69.
spent hydrogen and oxygen are vented overboard along
with water vapour to carry the contaminants with them. To
prevent the exhaust vent of the purge lines from freezing • Leave the purge valve open for two minutes, moni-
over, it needs to be heated prior to a purge. tor that reactant flow does not reduce again during the
The need to purge a fuel cell can be recognized from purge (that is a sign of line freezing).
monitoring the main bus voltages. As the power consump-
tion of the equipment aboard the Shuttle remains fairly con- • Close the purge valve again and monitor that reactant
stant, dropping fuel cell efficiency will mean they get sat- flow drops, indicating that the valve is indeed closed.
isfied by higher current and lower voltage, till the voltage
eventually drops below the equipment operating voltage • Purge other cells as required.
and devices cease to function. A CWS warning light will
illuminate before actual failures occur. • Leave the line heater on for another 30 minutes to pre-
vent leftover water from freezing in the line.
Procedure
Fuel cells can (and should) be purged with electric load on, Note: Line freezing is actually simulated, so take the
so there’s no need to disconnect the cell that is to be purged procedure including the heating period seriously.
from the main bus, however no more than 10 kW should
be demanded from the cell during a purge (nominally the
whole orbiter uses 14 kW which are supplied by three cells, Cabin air EVA preparation
so this is an exceptionally high load).
• Open SM DISP 69 via SPEC 69 PRO on an MDU that Background
shows systems management.
During a spacewalk, astronauts breathe a low pressure pure
• Switch on the purge line heater ( Electric system GUI oxygen atmosphere. This is done to pressurize the space
dialog) . Monitor the line temperatures (PURGE LN suit as little as possible, because a suit pressurized to a stan-
O2 T and PURGE LN H2 T1/T2 and wait till they dard atmosphere would be far too inflated and hence inflex-
reache 69 F (h2) and 79 F (oxygen T1) and 40 F (oxy- ible to be of actual use.
gen T2) and above. This may take a while if the purge However, reducing the ambient pressure creates the dan-
line is cold. ger that nitrogen dissolved in the blood forms bubbles and
causes decompression sickness. For that reason, nitrogen
• Open the purge valve of the fuel cell you want to purge.
needs to be flushed out as much as possible before a space-
Monitor FLOW O2 and FLOW H2 on the left hand
walk. This can be done by pre-breathing pure oxygen in-
side of the display — it should increase for the fuel cell
side the suit for some four hours. However, this time can
that is currently being purged as reactants are vented
be significantly reduced by lowering the cabin air pressure
overboard.
while keeping oxygen partial pressure. For instance, after
• Monitor the fuel cell voltage — if the purge takes ef- 24 hours at 10.2 psi rather than the usual 14.7 psi pressure,
fect, it should increase. the pre-breating time is reduced to 40 minutes.
92 CHAPTER 8. NOMINAL PROCEDURES

Procedure
Surprisingly enough, the Shuttle is not equipped with regu-
lators to automatically lower cabin atmosphere pressure in
preparation of EVAs — this has to be done manually and
moreover will trigger CWS warning lights.
Note that the procedure described below differs from the
original one because some elements (mid deck controls and
airlock valves) are not modeled.

• Open the environment monitoring DPS page via SPEC

66 PRO

Figure 8.6: N2 and O2 supply valves on panel L2.

• Open the oxygen supply valve again to bring oxygen

partial pressure to around 3.2 psi

While the cabin pressure is reduced, you will have to

manage the atmosphere by yourself by opening the supply
valves for oxygen and nitrogen — add nitrogen when you
have a high oxygen partial pressure (it should not go above
3.6 psi) to keep cabin pressure at 10.1 psi, add oxygen when
you have a low oxygen partial pressure (it should not drop
below 2.7 psi).
Figure 8.5: Cabin atmosphere parameters on SM DISP 66. When the EVA preparation is over, simply open supply
valves to let the automatic regulators restore cabin pressure.

• Look at cabin pressure (CABIN PRESS) and oxygen Note: Always make sure the vent valves are closed
partial pressure (PPO2, three readings) in the upper left after using them!
section. PPO2 should remain above 2.7 at all times.

• Close all oxygen and nitrogen supply valves on panel Extravehicular activity
L2 (SYS 1 and SYS 2) — this will prevent the auto-
matic regulation from trying to bring pressure up to Background
14.7 psi again.
If you dare, you can actually get out and take a spacewalk
• Next locate the vent isolation and vent valve on the (the simulation currently does not check whether you’ve
upper part of display L2 and open both — this will let been pre-breathing the right atmosphere). This only works
the cabin atmosphere escape into space. Anticipate the in orbit and simulates the use of a simple manned ma-
cabin atmosphere CWS light to come on. neuvering unit (MMU) which allows you to apply small
amounts of thrust to move around. In reality, the MMU
• Monitor cabin and oxygen partial pressure carefully. has been used only a handful of times from the Shuttle and
At around 12 psi cabin pressure, close the vent valves was quickly deemed too dangerous. Once you’re out some
again distance from the Shuttle, you might start to realize why!
• Open the oxygen supply valve and monitor oxygen
flow (O2 FLOW) climb Procedure
• Re-supply oxygen till oxygen partial pressure reaches There are no outside visuals provided for you spacewalk-
about 3.3 psi ing, it’s a first person view experience and you can’t pan the
view (think of how a space suit restricts the field of view).
• Vent cabin air till pressure reaches 10.2 psi During the spacewalk, the Shuttle needs to be in an inertial
DEPLOYING THE AIR DATA PROBES 93

attitude hold DAP (as you can’t control it since the con- This is not an option during entry, as the sheath of plasma
trol input will be used for the MMU). Also, once you are created during the early entry phase leads to a communica-
spacewalking, you need to get back to the airlock, you can’t tions blackout, and by the time communication could be
simply leave the view. re-established, the Shuttle may simply not be in range of a
tracking station till very close to the landing site.
• Using v , go to the EVA view looking out into the The upshot of this is that coming from the hypersonic into
payload bay from the airlock. the supersonic regime (say around Mach 5), the orbiter can
be expected to have sizable state vector errors. GNC tries
• When you’re ready to step out, press Shift + e to to acquire TACAN stations along the entry track as soon
start the spacewalk — you’re now adrift. as possible to correct the error (the TACAN at the landing
site can be expected to be received with some 400 miles to
• m toggles from rotational to translational control,
site), but this does not provide altitude information. Thus,
and on-screen message will tell you which is currently
for a good altitude estimate during the last part of the flight,
active. Using the input device or keyboard, you can
barometric altitude is obtained.
apply bursts of thrust or continuous thrust to maneuver
There are two air data probes stowed behind thermal pro-
around.
tection tiles at the forward lower fuselage. The probes con-
• There is no rate-controlled DAP or attitude hold DAP tain pressure and temperature sensors which allow to derive
available, you need to null all rates by carefully ob- angle of attack, Mach number, equivalent airspeed (KEAS),
serving your surroundings. Nulling a 3-axis rotation true airspeed (KTAS), dynamic pressure (q) and barometric
may become a challenge quickly, and judging what is altitude. Some of the raw readings from the data probes (i.e.
happening with the combined effect of rotations and before merged into state vector information) can be found
translations is tricky. on SPEC 51.

• You don’t have to worry about consumables or reaction

mass, the simulation provides an unlimited reserve in Procedure
the MMU.
The air data probes are deployed around Mach 5. The
• If you want to end the spacewalk, you need to get switches for the left and right probe are found on panel C3
back to the airlock (where you started). Then press (Fig. 8.7). Do not deploy the probes much before that as
Shift + e again — this will change the view from they need heat shielding during the hot parts of the entry,
free-floating to anchored and allow you to change into and moreover doing so will open gaps in the thermal pro-
other views again. tection.

Deploying the air data probes

Background
As discussed before, much of the Shuttle’s navigation is
based on maintaining a state vector. If GNC knows the
current position, velocity and attitude, and if rate gyro as-
semblies (RGAs) accurately sense attitude change rates and
inertial measurement units (IMUs) sense accelerations, then
the equations of motions can be used to compute state vec-
tor updates based on the sensed accelerations and attitude
changes. Thus, in theory this is all that is needed for navi- Figure 8.7: The air data deploy switches.
gation.
In practice, the procedure is error-prone. No measure-
ment is perfect, any error in attitude will mean that sensed Simply put both switches into deploy position to activate
accelerations get attributed to the wrong direction, and any the probes. The deploy/heat position in addition heats the
error in the magnitude of sensed acceleration will lead to er- probes, but this is not necessary and operationally not used.
rors in updated velocity and ever-growing errors in position. You can check on the ADTA section of SPEC 51 whether
For that reason, during orbital operations the Shuttle is pe- the probes are measuring data. If you are using realistic
riodically tracked by ground radar and state vector updates navigation, you in addition need to authorize the use of air
are uplinked. data by the selection filter on SPEC 51.
94 CHAPTER 8. NOMINAL PROCEDURES

Air data is a crucial ingredient of the area navigation, pro- starts they are already in the approximately correct position,
viding the altitude that radio navigation can not measure. hence no transients affect the accuracy of the burn.
In addition, air data is used e.g. to derive gains for airfoil
movement. If the probes do not deploy properly, this will Procedure
generate a PROBES software CWS warning message, and
if no air data is available to the navigation system by Mach To do a single engine burn, set item 2 or 3. To point the
2.5, a NAV EDIT ALT warning message will be created as engines through the CoG enter via items 6-8 the values P
a reminder. +0.4, RY -5.2 and LY + 5.2 (if you use the right engine, LY
If the air data probes ever need to be stowed, that has to won’t matter and vice versa). Then program and execute
be enabled by the pair of stow enable/inhibit switches also the burn normally.
on C3. Moving the deploy switches back into stow position The system will store the last trim values used by the
without enabling stow before will not issue a command. TVC DAP at the end of the burn, so if you stay with sin-
gle engine burns, you don’t need to enter trim values again.
However, if you want to go back to a two engine burn, en-
Single engine OMS and RCS burns ter instead P + 0.4, RY + 5.7 and LY -5.7 (note the sign
reversal!).
Background
Note: Expect that the simulation gravitates to
Apart from the insertion and the de-orbit burns, there are slightly different values, dependent on the lateral dis-
sometimes burns required to correct the orbit slightly. These tribution of payload etc. — this is no reason for con-
generally have relatively low ∆v requirements. However, cern.
the OMS engines don’t have infinitely precise cutoff timers,
any burn duration specified is only approximate. This is no As mentioned earlier, the trim values entered will only
problem for long burns, but for burns which last just a few affect the first few seconds of the burn, after that the TVC
seconds, the relative error can be comparatively large. DAP will alter the values as needed — so a failure to enter
The situation can be improved by utilizing just a single proper trim values for a long burn is just wasteful, but not
engine for the burn — in which case the acceleration is likely to cause an actual problem.
cut in half, the burn duration doubles, and the relative er- If you need to do an RCS burn, expect a long duration
ror shrinks accordingly. There are other situations which (since the RCS is significantly less powerful than the OMS).
may force the use of a single engine, for instance problems Most likely such burn will be an emergency de-orbit burn,
with the other OMS engine or the propellant supply. In case in which case the RCS propellant reserves won’t be suffi-
of problems with both OMS engines, there is also the pos- cient, which means that you need to set up and OMS to
sibility to do a burn using the RCS translational engines. RCS crossfeed first.
Items 1-4 on the MNVR DPS page control which engines After setting item 4, the burn can be programmed regu-
are used — the options are both engines, left engine, right larly (the avionics will take the reduced thrust into account).
enging or the RCS. Trim angles for OMS engines don’t matter in this case since
There is, however, a complication when using a single the engines are not used at all. Unlike the OMS, the RCS
engine: The net thrust has to be vectored through the Shut- is not ignited automatically (though the correct DAP is cho-
tle’s CoG to prevent unwanted rotations. In a two engine sen) — when the burn timer reaches TIG, you need to fire
burn, since the thrust is left-right symmetrical, only the en- the RCS by hand using the THC (and also end the burn
gine pitch has to be vectored. This is no longer the case for when the TGO counter reaches zero).
a single engine burn — here the engine also has to be vec-
tored in the yaw axis to push the Shuttle a little ’sideways’.
The TVC DAP automatically takes control of such trim
The OMS kit
issues by nulling rotation rates — but the engine gimbal is
Background
not fast, so it takes a few seconds to go through the whole
gimbal range. In particular, this means when the engines are As it is, the Space Shuttle is limited to rather low orbits <
gimbaled for a two engine burn when a single engine burn 600 km altitude. The main reason for that is the rather lim-
starts, the engine has to gimbal in the yaw axis for a few ited availability of OMS propellant (which would mainly be
seconds to reach the correct position — which will make a used to raise and lower orbit after MECO).
burn which lasts only a few seconds in the first place go off If higher orbits are desired, the solution seems rather
quite a bit. straightforward - carry extra propellant as part of the pay-
The solution to this is to manually enter engine trim an- load (and decrease the payload mass that can be carried ac-
gles — a few seconds prior to the burn, the engines will cordingly). This is what the OMS kit modules were de-
be gimbaled to the specified angles, and when the burn signed to do - provide an extra up to 1500 fps velocity
LAUNCH PERFORMANCE 95

change worth of OMS propellant via modules in the pay- corresponding Helium valve switches are also found on
load bay. panel O8.
The idea got as far as having all the switches and internal
wiring in place — but OMS kit modules were never flown in As of now, the crew does not have insight into the
actual reality. In the simulation, they can be selected in the OMS kit status via the DPS systems summary - pre-
simulation options menu or specified as part of the mission sumably there would have been an extra page to ac-
file. cess this information, but it is unclear how it would
have looked.
Procedure
Note: The maximal OMS kit (1500 extra fps) that
can be carried by the Shuttle is rather heavy and lo-
cated aft in the payload bay. As such, it poses a sig-
nificant danger during launch aborts (especially con-
tingency aborts) because of the CoG shift towards the
rear and the long dump time of the extra propellant,
which in essence means the Shuttle might not be in a
correct trim when it hits the atmosphere in an emer-
gency. Contingency procedures were devised long
after the idea of the OMS kit appeared in the Shuttle
program, so this is an issue that simply has no good
solution (and never needed to be addressed in reality
as the kit was not actually flown).

Launch performance
Figure 8.8: The propellant dialog with an OMS kit installed.
Background
The default launch profile assumes a rather standard trajec-
Once the OMS kit is installed, it can be connected to ei- tory and the AP generally tries to steer towards that trajec-
ther OMS or RCS via the crossfeed manifold. The switches tory. For launching an empty Shuttle into a low inclination
for the kit valves are located on panel O8, right next to the mission, there are sufficient propellant reserves to cope with
left OMS valves. The sequence of operation is as for a nor- the fact that this trajectory is not optimized. However, for
mal crossfeed: more marginal missions like heavy payloads into high or-
bits at high inclination, the default trajectory may simply be
• Close the receiving side tank isolation valves to dis- too wasteful.
connect the tank which can not supply propellant from One example is lofting the trajectory. If the Shuttle
the manifold. reaches orbital speed (and MECO) at a higher altitude, the
• Open the OMS kit crossfeed valves. OMS engines need to do less work to increase orbital al-
titude later. However, lofting is generically risky as it in-
• Check the receiving side tank isolation valve talkbacks creases the stresses in case of a contingency launch abort
to read CL. when the Shuttle falls back into the atmosphere. Con-
versely, depressing the trajectory increases surivial proba-
• Check the OMS kit crossfeed valves to read OP. bility in case of an abort, but requires longer OMS burns to
• Open the receiving side crossfeed valves. climb to a certain orbit.
Another example is the throttling back of the engines dur-
To check the level of propellant still available in the OMS ing the highest dynamical pressure phase when passing the
kit, the overhead analog gauges are used. When the rotary lower atmosphere. By default the throttle commands are set
switch is turned to either OMS fuel or oxidizer position, the for an empty Shuttle which accelerates rapidly off the pad
center gauge labeled FWD/KIT shows the remaining frac- — for a Shuttle lugging a heavy payload, the time the en-
tion of OMS propellant in the kit tanks. Likewise, when gines are throttled back can be reduced, or even the value
an OMS kit is installed, the propellant dialog shows the kit to which they are throttled can safely be increased — all of
tanks along with their details (see Fig. 8.8). which improves ascent performance.
Like left and right OMS, the kit tanks need to be pres- Finally, to target a very high apoapsis, it is useful to make
sured by Helium to be able to supply any propellant. The part of the OMS-1 burn already during ascent. The reason
96 CHAPTER 8. NOMINAL PROCEDURES

is that if OMS propellant is burned at SRB separatuin, the A similar effect can be achieved during second stage bal-
Shuttle gets lighter for the remainder of the stage 2 ascent listic climb when the Shuttle pitches down further. By vec-
— which means more acceleration for the same propellant toring yet more thrust to the horizontal, the arctop altitude is
consumption. This is known as the OMS assist burn. depressed and orbital speed is built faster, otherwise the op-
There are no easy expressions what parameter settings posite is true. One degree in bias angle corresponds to about
to use for a particular mission — you have to use external 7000 ft — by adjusting the bias angles in sync with the de-
simulation software or simply experiment a bit. sired MECO altitude, a smooth trajectory can be achieved.
Note that the ASCENT TRAJ display is scaled with the
provided loft value — if the loft value is 30.000 ft, then the
Procedure
Shuttle will be shown on the reference trajectory when it is
All changes to the default launch profile have to be i-loaded actually 30.000 ft higher than for a default launch with zero
into the launch section of the mission file. loft specified.
The first triplet of parameters determines how the engine
throttline during passage through the atmosphere is done.
throttle-down-time-s sets the time after liftoff when Circuit breakers
the engines are throttled down, throttle-up-time-s is
the time when the engines are throttled back to full power, Background
and throttle-down-to-percent is the value down to The Shuttle, with its many systems, also has a large number
which the engines are throttled (this can’t be set smaller of circuit breakers installed. A subset of them is imple-
than 67%). mented and working in the simulation (most importantly on
An OMS assist burn (to be started after SRB separation) the overhead panels O14 to O16).
can be programmed by setting oms-assist-burn to true, Circuit breakers serve two purposes: First, they protect
and oms-assist-duration-s determines how long the the electric system against power surges when a circuit be-
OMS engines are fired. gins to fail — they trip and cut the power to the affected
Finally, to loft the trajectory, trajectory-loft-ft system. In this role, they are currently not availble in the
can be used, which if positive increases the MECO altitude, simulation.
if negative decreases it. However, the parameter only sets However, they also serve another purpose: Moving
the final altitude that is targeted for MECO (see Fig. 8.9) — around in zero gravity, it is fairly easy to accidentially bump
if no adjustments are done to the previous phases, the AP into a switch and activate some system that should be off
will be able to steer there only relatively late. — think for instance switching MLS on while in orbit —
the system serves no purpose but still consumes valuable
power. It is thus operational procedure to save some sys-
tems by pulling their circuit breakers when the system is
not needed — circuit breaker and switch then form a logi-
cal AND, power can only flow if both are on.

Procedure
Saving systems via their circuit breakers is no operational
necessity unless you’re in the habit of randomly clicking
Figure 8.9: Phases during ascent.
spots in the cockpit, but if you’re working with the origi-
nal NASA flight data files and following the detailed proce-
Rather, when the trajectory is lofted or depressed, the dures outlined there, you may find it useful to operate the
whole arctop altitude should be shifted as well. This can be circuit breakers. You can toggle their state by simply click-
accomplished using srb-climbout-ang-bias-deg and ing on them.
ballistic-climb-ang-bias-deg.
The first parameter determines what is done after the
max. qbar phase is over and engines are throttled back
up. At this time, the Shuttle (still with boosters attached)
pitches down to 45 deg and starts to build orbital speed. By
providing a negative bias angle, the climb can be made to
be more shallow, which means faster buildup of speed and
reduced arctop altitude. Conversely, by increasing the angle
a few degrees, the arctop altitude can be increased.
Chapter 9

Failures

Types of failures These have a short associated text describing what will hap-
pen and create a randomized problem for the user to deal
In reality, space flight has to deal with the possibility of with.
component and systems failures. While a simulation can A third way to trigger failures is to insert a failure point
be made to always work perfectly, the ability to simulate into the mission file. This is a time point at which the
failure modes has driven lots of the design of the Flightgear probability for a certain component to fail is evaluated —
Space Shuttle simulation. this mode is most useful to specifically train certain failure
The Shuttle is designed around a philosophy of triple re- modes, for instance abort procedures for an engine failure
dundancy — the failure of any one critical system should in a given time interval.
allow to continue the planned mission, the failure of a sec- Since both a selected failure scenario and a failure
ond system should allow a save return to Earth, and only the point create a failure that is known to the user, the
third consecutive failure of the same system leads to loss of Failure training menu item allows to create a random fail-
vehicle and crew. For that reason, critical systems are al- ure at a selectable difficulty level. In this way, it is possible
ways installed threefold — there’s there main engines, three to train to diagnose and deal with problems without prior
APUs and hydraulic systems, three main electrical buses, knowledge.
three fuel cells,. . . The last path is to trigger the failure manually using the
Therefore many failures won’t initially appear very property browser (or remote access to an FG instance from
prominently to the crew, because the redundancy allows to an instructor station via a tool like Phi) - all components
more or less just continue operations. For instance, even which can be failed are located under
with two hydraulics systems down, when hydraulic pres- /fdm/jsbsim/systems/failures/
sure is severely reduced, the handling of the Shuttle in the and normally the number between 0 and 1 describes the
atmosphere is not dramatically different for normal maneu- percentage at which a component is working (a thruster may
vers (though the effect will be felt when maneuvering agres- e.g. fail gradually by not providing full thrust). In case a
sively). component fails binary, any value that is not 1 is usually
Simulated failures won’t occur on their own (i.e. com- interpreted as failure by the simulation.
ponents will not fail for no good reason). There are five
different ways to trigger failures in the simulation. The first
is violating limits. For instance, touching down with a high The caution and warning system
vertical speed exceeding structural limits has a certain prob-
To catch the most critical failures, the Shuttle has a hard-
ability to blow tires or even break the gear struts. Overheat-
ware caution and warning system (CWS) that works inde-
ing an APU by neglecting to cool it will cause failure of the
APU, and subsequently of the associated hydraulic system. pendently of the flight software and announces detected off-
Letting RCS fuel lines cool below operating temperatures nominal conditions by warning lights and — for especially
by neglecting thermal management will cause the lines to critical conditions — alarm tones and klaxons.
freeze and render the jets non-functional. The CWS lights are found in the center of panel F7 (and
should usually not be illuminated). There are both red
Damage due to limit violations can be disabled in the and yellow lights, separating dangerous conditions requir-
simulation options by choosing the ’soft’ simulation ing immediate attention from serious problems which nev-
mode. ertheless leave some time to deal with them.
If a critical failure occurs, the light indicating the failure
Another way to experience failures is to select a failure will illuminate and at the same time the Master Alarm push-
scenario inside the simulation from the dropdown menu. button indicators (on the eyebrow panels F2 and F4 as well

97
98 CHAPTER 9. FAILURES

CABIN ATM
If the air pressure in the cabin exceeds 15.53 psi or is lower
than 13.74 psi, or if the oxygen partial pressure exceeds
3.6 psi or is lower than 2.7 psi, this warning light illumi-
nates. Usually this is a sign that something is wrong with
the environment system, and the condition can get danger-
ous quickly because air may run out. The light will however
also illuminate if, in preparation of a planned EVA, cabin air
pressure is intentionally lowered to about 10 psi.
Figure 9.1: The CWS lights.

AV BAY/CABIN AIR
as on the aft panel A7) will illuminate and an alarm tone
This warning light monitors the action of the various fans.
goes off. The alarm tone can be silenced by depressing any
It illuminates if the temperature in one of the avionics bays
of the Master Alarm pushbuttons, but the CW light is active
exceeds 130 F or if the cabin air goes stale when it is no
as long as the dangerous condition persists.
longer circulated by the air scrubbers.
There are two special alarm tones for especially danger-
ous conditions — a klaxon for rapid cabin depressurization
and a siren in case a fire is detected. These conditions also H2O LOOP
illuminate the Master Alarm pushbuttons, but they do not
have CWS warning lights on the matrix. A fire alarm will The H2O loop is responsible for cooling the avionics. If an
however illuminate the red smoke detection light on panel off-nominal condition with one of the pumps is sensed, this
L1. Also for these alarms, the tone can be switched off by light illuminates, indicating that the avionics bays may soon
depressing the Master alarm button. overheat if the backup pump is not switched on. This may
It is possible to test the proper functioning of the CWS at worst result in the loss of all avionics.
matrix lights via the Caution/Warning Mode switch on
panel C3. In the ACK position, no light will be illuminated,
but depressing any of the forward Master Alarm pushbut- FREON LOOP
tons will illuminate the whole matrix (the Ascent position
works the same as the regular position, except that the Com- If the freon temperature is sensed too warm, then this light
mander’s Master Alarm button is not illuminated when an illuminates, indicating that the proper functioning of the
issue is detected — it remains NASA’s secret why this is thermal management systems need to be investigated. Usu-
implemented that way). ally this would point to a problem with the radiator loop
The various lights denote the following failure condi- or a failure to activate the flash evaporators when the ra-
tions: diator loop is down. The light also indicates if the freon
is too cold, usually indicating that less cooling action than
currently available is needed.
MAIN BUS UNDERVOLT
Voltage on one of the main electrical buses is less than 26.4
V. This may be caused by disconnecting the bus from the FUEL CELL STACK TEMP
fuel cells, by a fuel cell failure or by a shorted circuit on the This light is an indication that one of the fuel cells is run-
bus. It can be expected that there is lots of associated equip- ning too hot. This is a dangerous condition because the cell
ment failures (a third of the MDUs goes black for instance) might explode and cause lots of damage and hence usually
when a bus experiences a power loss. it means the fuel cell need to be switched off by closing the
reactant valve.
AC VOLTAGE
This light illuminates when one of the AC buses registers FUEL CELL PUMP
less than 115 V. if the light illuminates without a MAIN
BUS UNDERVOLT, this means that the AC bus is either Problems with the fuel cell coolant pumps are indicated
not powered by the alternator or is shorted. A failed AC by this status light. It is triggered by an off-nominal pres-
bus without the main bus usually implies that equipment sure differential before and after the pump, indicating that
with rotating components such as pumps or fans can’t be coolant is no longer circulated through the fuel cell and the
operated, but avionics remains functional. stack might soon overheat.
THE CAUTION AND WARNING SYSTEM 99

OMS TVC APU TEMP

A thrust vector control failure of the OMS means that one or If the temperature in one APU reaches more than 290 F, this
both engines can no longer be gimbaled, i.e. accurate atti- light illuminates, indicating that the APU is not properly
tude control during an OMS burn is no longer possible. The cooled by the water spray boiler. This may indicate that the
light illuminates if the logic detects a difference between boiler is failing, has no water left or was not switched on.
commanded and actual OMS gimbal position on one of the Operating an overheating APU will lead to turbine failure
engines. and subsequent loss of the hydraulic system rather sooner
than later.

LEFT (RIGHT) OMS

AIR DATA
The OMS warning lights illuminate when there’s either a
problem with tank pressure (too large or too small), indi- The air data warning light signals a dilemma in the air data
cating that fuel flow to the engine might soon cease due to systems, i.e. the set of air data probes shows measurements
lack of pressure, or if the OMS engine fails to reach more which can’t be reconciled by internal consistency checks.
than 80% of nominal chamber pressure when commanded Since the redundancy management software can’t decide
to ignite. Both are indications that the OMS engine might which set to trust, no air data is utilized as long as the condi-
not be available any more. tion persists, and the crew needs to manually de-select part
of the data set on SPEC 51 to solve the dilemma.

FORWARD (LEFT, RIGHT) RCS

BACKUP CW
Like the OMS lights, the RCS warning lights illuminate if
tank pressure is too large (greater than 312 psi) or too small The backup warning light illuminates for a host of condi-
(smaller than 200 psi), and thus if the propellant flow to the tions which need to be announced even if one of the warn-
jets might cease. ing lights fail. Usually they also trigger a software CWS
alarm which explains the details.

HYD PRESS
Engine status lights
If one of the hydraulics systems registers a pressure of less
than 2400 psi, this warning light illuminates. Usually the In addition to the CWS lights, there are three status lights
warning would occur during on-orbit operations of one of for the main engines in the lower part of panel F7. They
the hydraulics circulation pumps ceases to operate and the have an upper red and a lower yellow portion (see Fig. 9.2).
pressure in the system gradually drops as a result — if all
pumps are on GPC control, they’re periodically cycled on
to maintain pressure at about 2600 psi. Note that the fact
that this light is off does not mean the hydraulics system is
ready for operation — for that, it must be pressurized by the
APUs and read about 3000 psi operating pressure.

APU OVERSPEED
Figure 9.2: The engine status lights.
The light illuminates if one of the APUs reaches 92880 rpm
or more, indicating that the turbine is no longer operating
safely. If enabled, this condition normally causes and auto-
The red portion of a status light illuminates whenever the
shutdown to prevent damage.
engine provides no thrust (either because it has failed, or
because it has intentionally been shut down). The yellow
APU UNDERSPEED part illuminates whenever the engine can no longer be throt-
tled but continues to provide thrust at current throttle setting
The underspeed light is illuminated if one of the APUs fails (this is known as a lockup condition).
to reach more than 57600 rpm, indicating that the turbine The engine status lights illuminate fully post-MECO and
is not sufficiently powerful to provide operating pressure to remain on till ET separation, this is perfectly normal and no
the hydraulic system. reason for concern.
100 CHAPTER 9. FAILURES

Software CWS messages software running any systems-management functionality,

which means that SM failures are only detected by the BFS.
The Shuttle’s GPCs continuously analyze the incoming data
stream and detect off-nominal sensor readings. Some of If you see an SM ALERT but no flashing message,
them are just inserted into the telemetry, other conditions it’s most likely a BFS-specific message which you
more likely to require quick attention are handled by the can only view by switching one IDP to the BFS GPC.
software caution and warning (CWS) system.
Software CWS messages appear on the fault line at the A summary of all fault messages produced by the simu-
bottom of a DPS screen, flashing in red, with the latest mes- lation is given below.
sage on top. The correct response is to depress the ACK
key on the keyboard to stop the flashing, and continue to
G23 X RCS Y JET
press ACK to bring subsequent messages on-screen. The
fault line can be cleared with the MSG RESET key and the This message announces a problem with one or more of the
FAULT SUMM key brings up DISP 99 with a list of the last RCS jets. The jet is denoted by pod and direction, so F RCS
15 failure messages. L JET denotes a left-firing jet of the forward RCS module.
Failures which are simulated and detected are fail off, fail
Critical failures are not annouced by the software on and leak.
CWS system but have warning lights on the hardware A jet failed off has received a command to fire, but the
CWS system. sensors do not detect an actual jet firing, i.e. the jet doesn’t
work. A jet failed on has received a command to shut down,
The messages appear in a standard format, for instance but the sensors still register a firing jet. Finally, a jet with a
S88 EVAP OUT T 1 * 4 19:20:10 leak shows fuel flow despite not having received an ignition
The first triplet is the ID of the DPS screen which can command and no firing sensed.
show more information, for instance this message points Details of the failures are found on GNC SPEC 23 where
towards systems management (S) DISP 88 (that’s environ- also the capability to close fuel flow to a jet failed on and the
ment thermal properties). The next part is the actual fault capability to de-select failed jets (and their opposing coun-
message, mentioning here evaporator out temperature. The terparts) from the jet table exists.
next fields are rather technical and display detailed info on
what particular sensor is responsible (usually not simulated) X OMS GMBL
and what class the message falls in, the last numeral is the
GPC announcing the fault and the trailing entry is the time Problems with the gimbal of the OMS during a burn or a
the fault was detected. gimbal test register with this message where X stands for
Until an I/O reset is done, detected faults are reported the left or right OMS, this L OMS GMBL denotes a prob-
once, and acknowledging and resetting the message will lem with the left OMS gimbal.
make the fault warning go away, regardless of whether the Generally this means that the OMS thrust vectoring will
problem is fixed or not. An I/O reset for a given GPC can be have problems controlling the Shuttle (though limited capa-
done via the corresponding key on one of the keyboards. Be bility exists even if one engine is still capable of gimbaling).
prepared that this might cause a number of already reported GNC OPS 202 (in orbit) or the corresponding MNVR page
problems to re-create fault messages if they were not actu- can be used to start gimbal tests and to select the backup
ally fixed. The real point of I/O resets is that some equip- gimbal system for the OMS engines, as well as switch to
ment might be off at times, and this is of course is regis- single engine OMS or RCS burns in case the function can
tered by the CWS system as ’non-communicating, possibly not be restored.
faulty’ — so the software announces this and stops trying
to communicate with the system. Thus, when the system is
X OMS QTY
switched on, it requires an I/O reset to get it to communicate
with the GPCs. If the OMS propellant is running in one of the pods, this is
If the condition is sufficiently important, the software announced with this message. For instance, the right OMS
CWS messagte is backed up by a short alarm tone and the pod announces as R OMS QTY. Whether this is problem-
SM ALERT light on panel F7 is momentarily illuminated to atic or not depends on whether it happens in the last stage
get the crew’s attention. Both the tone and the illumination of the planned de-orbit burn at the nominal end of a mission
will stop without crew intervention. or is the result of a leaking tank.
Most failure messages are the same for PASS and BFS, In the latter case, single engine OMS burns or a cross-
but there are a few that are specific to either software. feed may be required to end the mission safely via an un-
In particular, during ascent and entry, BFS is the only planned de-orbit burn.
SOFTWARE CWS MESSAGES 101

X OMS PC MPS ELEC X

This message signals a problem with an OMS engine during This message indicates an electric lockup on one of the
operation — the measured chamber pressure did not reach main engine where X identifies the engine as left, right or
above 80% of rated power despite the engine commanded center. The message is a software backup alarm, the main
full on. As usual, X denotes the engine, i.e. L OMS PC indicator in the cockpit is a yellow engine status light from
refers to the left engine. the hardware CWS. The crew needs to fly procedures for
Usually this means the engine is no longer safe to oper- engine lockup in response.
ate and should be shut down — there may however be sit-
uations in which the current OMS burn is critical, in which
case it may need to stay on. MPS HYD X
A hydraulic lockup (i.e. a failure in one of the three hy-
X OMS TK P draulics systems during powered flight) is indicated by this
message where X identifies the affected engine. Like in
The OMS tanks are pressurized with helium and the valves
the case of the electric lockup, the software message is a
operated with nitrogen, and this message indicates that one
backup, the main indication of the failure is a yellow en-
of the pressure readings are too low. In the case of helium,
gine status light. The crew needs to fly procedures for en-
that may mean that only a limited amount of the fuel left in
gine lockup in response.
the tank is accessible before pressure runs out, in the case
of nitrogen, it may signal that the number of future burns
with the affected engine is very limited. SSME FAIL X
The message may signal the need to set up a crossfeed or
to not use the affected engine any more. The complete loss of a main engine is announced by this
message where again X identifies the affected engine. Like
the other engine status messages, this is a software backup,
X RCS TK P the main indication of an SSME failure is a red engine status
As above, the message refers to an abnormal pressure read- light. Likely this message means that a launch abort needs
ing in one of the RCS tanks where X identifies the affected to be initiated.
system (so F RCS TK P would indicate a problem in the
forward RCS module).
SSME REPOSFAIL
X RCS LEAK Prior to entering the atmosphere, the main engine nozzles
are repositioned upward to be out of the airstream. If this
Since propellant and oxidizer are used in given pre- maneuver could not be performed before the MM 304 tran-
determined proportions, the relative filling fractions of oxi- sition because the TVC isolation valves were closed or be-
dizer and propellant must always be the same. If they’re not, cause not at least two hydraulic systems were fully pressur-
one of the tanks is leaking, and a deviation in the announced ized, this message is triggered.
RCS tank contents triggers this message.
Dependent on where the leak is located, closing off in-
dividual manifolds (and hence groups of thrusters) may or S67 AC VOLTS
may not be able to deal with the problem. Otherwise, the
RCS jets can be isolated from their own propellant supply Voltage on one of the three AC buses is detected to be crit-
and utilize OMS propellant via a crossfeed. ically low — below 115 V. This likely indicates either an
issue with the fuel cells or a shorted circuit, or something as
trivial as the AC system not connected to the main bus. De-
ET SEP INH pendent on what the problem is, it may require to tie buses
For external tank separation, the Shuttle should have near to draw power from the remaining good fuel cells or to iso-
zero rotation rates in order to avoid smashing into the tank late the AC bus and use the backup systems. Details are
after the separation has been initiated. If, for whatever rea- found on SM DISP 67 where the electrical system status is
son, this is not the case when the GPC commands sepa- displayed.
ration, this message is produced to alert the crew that the
separation has to be done manually. SEL AUTO
The correct response is to null the rates, arm the manual
separation sequence and depress the separation pushbutton This message is a mostly harmless reminder that in prepara-
on panel C3. tion to an OMS burn, the Shuttle is trying to maneuver into
102 CHAPTER 9. FAILURES

burn attitude but the orbital DAP selection is still in a man- G50 RM DLMA TAC
ual mode, so the crew should either cancel the automatic
As above, the message signals that the TACAN system is
maneuver or select the AUTO DAP pushbutton.
in a dilemma, i.e. is unable to decide which of two or more
conflicting data streams is valid. As a result, the data can not
PROBES be utilized by navigation till the crew resolves the dilemma
manually.
If Mach number is below 2.5 and the switches are set to
deploy the air data probes but the software has not regis-
tered the signal that both probes have actually reached their HIGH G
deployed positions after the usual deploy time, this error When the Shuttle during atmospheric flight maneuvers too
is created. It signals an issue with the mechanical parts of hard, i.e. with accelerations larger than 2.2 g, this message
probe deployment and usually means that one of the probes is triggered to remind the crew that structural limits will be
is stuck or damaged. approached soon.

G50 NAV EDIT ALT OTT ST IN

This message appears if by Mach 2.5, air data has not been During TAEM, the energy estimate for the Shuttle can es-
incorporated into the navigation state by the crew and hence tablish that the current approach is not guaranteed to reach
the altitude information provided by navigation might be the touchdown point safely. This message suggests to
bad. change from an overhead to a straight-in approach.

G51 RM FAIL ADTA OTT ST IN

If the redundancy management software of the air data sys- Like the above message, this low energy warning suggests
tem has been able to identify a failed transducer unit and to change the TAEM pattern from a nominal to a minimal
automatically de-select it, this alert is generated to notify entry point and thus shorten the path that the Shuttle has to
the crew. There is no further action required as redundancy fly before touchdown.
management has taken care of the issue and the air data
stream is considered good. GPC CONF
When the GPCs are instructed to load an invalid memory
G51 RM DLMA ADTA configuration, this message appears (note that this is only
possible if realistic DPS memory management is selected).
If on the other hand the redundancy management software
It indicates that the GPC and bus assignment table needs to
of the air data system has not been able to identify the faulty
be revised, as it is currently invalid.
unit but just registers incompatible data, no unit is dese-
lected and no air data can be used by GNC until the dilemma
The following messages are BFS-specific failure mes-
is resolved. Crew de-selection of one or more transducers is
sages which can appear during ascent and entry:
required. This message is also accompanied by a hardware
CWS light.
SM2 AVBAY TEMP
G50 NAV EDIT TAC The temperature in one of the avionics bays is too high,
indicating a problem with cooling or, at worst, a fire.
This alert signals that by an altitude of 130.000 ft, no
TACAN data has been incorporated into the navigation and
hence the state vector positional information might be bad. SM2 AV BAY FAN
The cooling fan in one of the avionics bays is not operating
G50 RM FAIL TAC at sufficient performance, leading to heat buildup.

Similar to the air data system, if TACAN redundancy man-

SM1 CABIN PRES
agement has been able to identify one receiver as faulty and
has automatically de-selected it, this alert is generated to Pressure inside the cockpit is too low, indicating that there
notify the crew. No further action is required, the remain- might be a leak somewhere, or the life support system isn’t
ing TACAN data is considered good. operated properly.
MEDS FAULT MESSAGES 103

SM1 CABN DP/DT POLL FAIL

Cabin pressure is dropping faster than a safety threshold, The message indicates that the reporting IDP has no con-
indicating a leak somewhere. tact to a GPC and is hence unable to display meaningful
DPS pages. The reporting IDP is appended to the message.
SM1 CABIN PPO2 This can happen because the IDP was intentionally isolated,
or because the GPC went bad, or because there is no GPC
Oxygen partial pressure in the atmosphere is too low so that running in the requested major mode.
the atmosphere becomes dangerous to breathe.
PORT CHANGE
SM1 CABIN FAN
If an MDU can not establish contact to its designated IDP, it
The cabin ventilation fan isn’t operating, which means that may be configured to automatically try the secondary port to
no air is circulated through the air scrubbers, which could contact a different IDP. If this happens, it continues to dis-
lead to a dangerous CO2 concentration. play normally, but a PORT CHANGE message is produced
to alert the crew that there has been a problem with one IDP.
SM2 FREON FLOW One cause that can produce the message is switching an IDP
off. The reporting MDU is appended to the message.
There’s no flow detected in the freon loop, indicating that
Note that the MDU can also be configured to not auto-
the thermal management does not work properly and heat
matically change ports, in which case it announces MDU
buildup will soon occur.
IS AUTONOMOUS and manual port change may be re-
quired.
MEDS fault messages
To make matters a bit more confusing, the MEDS system Limits
has its own set of fault messages created by the IDPs. These
refer to problems inside the display system (IDP, MDU, For a craft like the Space Shuttle, determining the actual
keyboard) and appear in white on the MEDS fault line (see failure limits under stress is something that hasn’t been
Fig. 9.3). done. NASA documents for contingency aborts for instance
describe maneuvers which might possibly work, indicating
that the Shuttle might hold up or not. While the Shuttle
is certified for operational limits, some of these are almost
certainly violated during a normal contingency abort pat-
tern. If the survival of the crew is at stake, certified limits
do not matter, the actual structural breakage points do.

Figure 9.3: A software CWS (red) and a MEDS fault mes- Aerodynamic forces
sage (white) showing on their respective fault lines.
The first major type of stresses are aerodynamic forces,
measured by the dynamical pressure q. During ascent,
MEDS fault messages are acknowledged with the forces regularly reach some 650 psf and higher — the forces
edgekey MSG ACK menu option, reset with the MSG RST are in fact much stronger than during a nominal entry. The
option and a summary can be accessed via the MEDS menu reason is that during entry the tolerable forces are in fact
structure on the fault summary page where the message list much lower due to a condition known as actuator stall —
can also be cleared. the hydraulics system is beyond some q simply unable to
MEDS fault messages announced by the simulation are move the airfoils against the pressure, which leads to loss
given below. of control. During ascent, no airfoils are needed to provide
control since thrust vectoring fills that role, hence the toler-
able q is much higher.
I/O ERROR
For the simulation it has been assumed that actuator stall
This is followed by the MDU ID (for instance IO ERROR gradually becomes relevant beyond 375 psf (the number to
CDR1 refers to the left-most screen in the cockpit) and which the controls are certified is quite a bit lower) and that
indicates that the announcing IDP could not reach the de- the structural limit of the ascent stack is at 820 psf.
vice. This message is for instance generated if an MDU is Actuator stall will not necessarily destroy the Shuttle, it
switched off. will just render the control surfaces ineffective, and depen-
104 CHAPTER 9. FAILURES

dent on the current state of the vehicle, this will evolve into properly detatch from the heat shield and a quickly unman-
a recoverable or non-recoverable condition (for instance, if ageable heat flux.
the Shuttle has a strong upward velocity when actuator stall The simulation requires that the total thermal stress re-
occurs, it moves into thinner air, q is likely to drop, causing mains below some limit. The stress is defined as the in-
airfoils to become responsive again). tegrated heat flux caused by temperatures at the TPS ex-
The most likely failure point of the structure during as- ceeding 1500 K. This means that a short-duration high tem-
cent seems the connection between orbiter and the external perature spike (as characteristic for e.g. a marginal single-
tank, and if the attachment struts fail the orbiter would be engine TAL abort) does not usually cause a TPS failure, but
likely to smash into the tank structure with engines ignited, a prolonged exposure to high temperatures does. If the TPS
causing an explosion. This scenario is what the simulation is damaged, the limit will be correspondingly lower (up to
assumes. the point where surviving an entry is no longer possible).
A different kind of limit is imposed by the wing bending Note that within this model, the failure can occur after
moment. This is both a function of q and AoA and requires the maximal heating phase as long as sufficient excess tem-
to adhere to a tight AoA schedule when passing through perature is collected during the phase.
the region of maximal dynamical pressure (between 30 and For a thermally safe trajectory, it is thus necessary that
50 seconds into the flight). To minimize the wing bending first all gaps (umbilical doors, payload bay doors, star
moment, AoA needs to be slightly negative which drives the tracker doors) are closed. Then, the Shuttle needs to be
need for inverted flight. flown within the nominal AoA schedule (at peak temper-
During the final approach, the landing gear has to be low- atures, 40 ±3 deg AoA) — any deviation from the nomi-
ered, and the limit beyond which it is damaged in the pro- nal AoA will cause additional heat load and hence increase
cess is 312 KEAS. Also, once on the runway, the drag chute thermal stress. Finally, the trajectory should be managed to
has a fail pin which cuts it loose if deployed beyond 230 be as cool as possible — for instance agressively increas-
KEAS. ing drag to get onto the nominal entry trajectory quickly
will cause more heat load than remaining above the nomi-
nal profile while temperature is high and adding drag later
Acceleration limits when the thermal stress is lower.
Both during ascent and entry, g-forces are strong, reaching To aid in flying a good entry profile, a limit callout is
nominally up to 3 g. The direction is however different, available which warns about unusually high temperatures
during ascent forces act along the body X-axis, during entry on the heat shield. This does not necessarily mean that the
the main force is lift acting along the body Z axis (hence the trajectory is bad, but it is always a sign that drag should be
accelerometer changes to display NZ in OPS 3). decreased as soon as possible (usually be decreasing bank
During ascent, an acceleration of 3.9 g is assumed to angle). As the entry autopilot is aware of thermal limits and
cause a fatal failure of the ET attachment struts, again re- tries to stay clear of them, this is mainly a concern when
sulting in an explosion. The NASA contingency abort train- piloting manually.
ing material speculates that the first failure point for high The scenario assumed for a TPS failure is a fatal breakup
NZ is the attachment of the OMS pods at some 4.2 g. Con- of the Shuttle into a debris shower.
tingency flight software will generally try to keep g-forces
below 3.9 however. The simulation of NZ limit violations
assume that a wing can be ripped off by sufficiently large
Touchdown limits
accelerations exceeding this value. Significant force acts on the wheels when the orbiter
touches down (touchdown velocity is generically high,
Thermal limits above 200 KEAS). To limit stress on the main gear, ver-
tical speed at touchdown should not exceed 9 ft/s. After
A nominal entry at Mach 27 causes significant heat load main gear contact, the Shuttle coasts for some time before
which would be enough to destroy the Shuttle (abort en- the de-rotation starts and the nose gear is brought down. To
tries typically happen at smaller velocities not exceeding limit forces acting on the nose gear, the angular speed at de-
Mach 10 and have much reduced thermal stresses, though rotation should not exceed 2 deg/s. Dependent on how se-
the aerodynamic and g-forces tend to be higher). vere the limit violations are, the simulation will either blow
Surviving this heat load requires correct functioning of tires or break gear struts, leading to changed friction coeffi-
the thermal protection system (TPS) — all gaps need to cients and loss of controll during rollout.
be closed and equipment stowed (ET umbilical door, star A different kind of limit is imposed by the fact that the
tracker doors and payload bay door closed) — and in addi- body flap touches the ground if the nose points higher than
tion the safe AoA schedule need to be strictly flown. Devi- 14 degrees at touchdown, a condition known as tailscrape.
ations from the AoA schedule cause the shockwave to not The need to avoid tailscrape limits touchdown airspeed
LIMITS 105

from below, as (dependent on weight) the nose needs to go

too high to limit vertical speed to avoid tailscrape for low
airspeed.

Minor limits
In addition to the general structural limits described above,
many components are modeled with their own operating
limits (and associated failures). For instance fuel lines will
freeze if the temperature drops below hydrazine freezing
point, APUs will fail if they’re insufficiently cooled, so will
the avionics bay, the cabin atmosphere will go stale if not
vented etc. These limits are described in this manual where
their systems are introduced.
106 CHAPTER 9. FAILURES
Chapter 10

Trivia

Here’s a collection of small details you may or may not crew input, regardless of whether the Shuttle is under man-
enjoy about the simulation. ual or automatic control.

Exhaust plumes RCS firing sounds

If you look carefully at the SSME exhaust plumes during It might take you to de-activate the nose RCS pod in SPEC
liftoff, they’re very sharp jets with something like a string 20 before you notice, but the aft RCS pod firings do not
of brighter spots (’shock diamonds’) showing inside. Yet, cause audible sounds in the cabin while the forward RCS
as Atlantis climbs higher, the plumes look very different — firings are heard very prominently.
they widen, and the shock diamonds disappear. This is as described by the astronauts, and the reason
is that sound does not propagate well enough through the
The reason is the interaction of the rocket exhaust with
structure of the orbiter fuselage (there’s of course no air to
the surrounding air. Initially air pressure keeps the hot gases
carry it).
close together and they can’t expand, and the air pushing
Similarly, if you let all air escape from the cabin, you’ll
back at the exhaust jet causes standing pressure waves to
no longer hear any in-cockpit sounds (the simulation won’t
form which generate the shock diamonds. Higher up air
kill you, but it’s not exactly a recommended course of action
pressure drops and finally disappears which makes first the
either).
standing waves go away and finally allows the plume to ex-
pand naturally.
Incidentially, this is also the reason rocket engine thrust
is lower at sea level than in vacuum.
Visibility of stars
Fairly often, the question whether the simulation should
show the stars also during the day portion of the orbit is
Airfoil motion during launch raised, arguing that since the sky is black due to the lack of
atmosphere, one should be able to see them.
Observing a launch carefully in outside view, you will no- However, the behaviour of the simulation is very likely
tice that the elevons are deflected after liftoff till about Mach correct ’as is’. The chief reason we don’t see stars during
2.5. Why does this happen? the day is not that the atmosphere absorbs their light (the
The reason is not that thrust vectoring would not have atmosphere is equally well there during the night where we
enough control authority — during the first stage with SRBs do see them), it’s that our eyes have the ability to adapt to
ignited, thrust vector control is more than sufficent to ma- a wide range of lighting conditions. With fully darkness
neuver the Shuttle. Rather, the maneuver is done to reduce adapted eyes (which takes about 30 minutes stepping out
the wing bending moment (which is a major structural con- from a brightly lit area), we can see very faint objects in
cern during the first part of the flight). By unlucky chance, a the sky which we would never see with brightness adapted
shockwave caused by an instrumentation box on the SRBs eyes.
happens to cause an unexpectedly high bending moment, So what matters is not so much the atmosphere (or lack
which needs to be carefully tackled. By flying negative thereof) but the brightness adaption state of the eyes. And
AoA and deflecting elevons, the moment can be reduced during the day portion of an orbit, there are plenty of bright
to safe values. surfaces to see, starting from the Shuttle itself over the mag-
In the simulation (as in reality), the elevon deflection hap- nificant display of Earth underneath (and of course the Sun
pens automatically during launch without the need of any itself). So the eyes during the day portion of an orbit are

107
108 CHAPTER 10. TRIVIA

normally daylight adapted — hence one does not see stars. Educational control modes
This is in fact confirmed by the testimony of Apollo astro-
nauts who did not see stars standing on the lunar surface. The real Shuttle utilizes a heavy layer of flight controllers
(Theoretically, since there’s no atmospheric light scatter- to map control inputs via the DAPs to actual engine gimbal
ing, one could darkness-adapt eyes during the day portion angles, jet firings or airfoil movements. This is by and large
of an orbit by looking away from all lit surfaces — then one also what the simulation does.
would see stars. This however is not simulated.) Flying a rate-holding DAP in the atmosphere is really
comfortable, the Shuttle handles almost like on rails — for
instance during entry one can simply put it into an extreme
bank angle/ high AoA attitude — and it’s going to simply
Power consumption stay there without any need for trim or control input. In
fact, the scheme never needs rudder or aerodynamic trim
because the DAP implicitly auto-trims the vehicle such that
It’s really hard to spot even with the electrical systems dis- controls centered never command any motion.
play open, but the MDUs consume power dependent on
For educational purposes, it is possible to take some of
their brightness setting in the simulation — really. In fact,
quite some effort has been made to identify plausible power this out. The key binding is Ctrl + m . On the launch
consumption of all sorts of components and simulate an pad, this will switch to a DAP in which the control input
electrical system in which the voltage on each bus gradu- directly sets the engine gimbal angles. Launching in this
ally decreases when more power is demanded. As a result, mode, you will have to cancel out atmospheric forces your-
you can actually watch fans, pumps or actuators come on self, and later, outside the atmosphere, struggle hard to null
by looking at the power consumption. all rates manually.
A similar mode is available during entry — here the DAP
allows you to directly control airfoils like for a plane, and
only a minimal amount of yaw stabilization is done auto-
matically. Using this DAP will help you to appreciate the
Computation speed enormous difference in response of the vehicle from hyper-
sonic to subsonic speeds — while very large airfoil motion
The actual Space Shuttle DPS ran on 1970 hardware. This is needed above Mach 20, the gains in the subsonic regions
means that on a modern computer, even emulating the func- are much reduced.
tionality in a simulation is much faster. For instance, during Successfully flying a launch by direct gimbal control or
a TAL abort, an OPS transition from 1 to 3 has to be made, surviving an entry when directly controlling airfoils is quite
and the description of the procedure instructs the crew to an achievement and can be a source of pride!
wait for almost a minute for the software to load before
switching to BFS.
In contrast, the FG simulated flight software has of course
all OPS sequences in memory at the same time and just The head-up display
doesn’t make it accessible unless the simulation reaches the
proper sequence — which makes OPS transitions happen at A real HUD is not just projected on a glass in front of the
lightning speed, essentially without any delay. cockpit window but really symbology projected at infinity
There are very few tasks for which the software takes such that the eyes never have to change focus when looking
notable time, most importantly numerical fast-forwarding outside and reading the HUD. That is to say, all HUD sym-
of the trajectory computation for orbital maneuver planning bology appears as if it would be at a great distance. Tech-
(even then, usually it’s just a few seconds delay). The sim- nically, that is achieved by a collimator which makes the
ulated trajectory prediction by the way does not utilize sim- outgoing light rays parallel.
ple 2-body pointmass orbital computation but follows the One consequence is that a symbol projected over some
equations of motion in non-spherical gravity. This is likely distant feature outside really stays there when moving the
something that’s way beyond the capabilities of the original head around, i.e. the HUD symbology does not show par-
software, however an actual mission would have had the ca- allax movement (while the HUD frame, being nearby, of
pability on the ground, and in reality burns did not need to course does). The HUD image appears to be fixed when
be computed by the crew, the computations were done by moving the head and just vanishes when moving the head
mission control and sent to the Shuttle. Since no mission outside the projected cylinder of light.
control is simulated, it seems reasonable to provide this ca- This behavior is fully implemented in the simulation
pability in a different way. (which is not readily apparent unless you move the view).
 109

Main engine ignition Rain splashes

The Shuttle is not supposed to be flown through rain — bad
Watching real Shuttle footage of engine ignition, you can
weather in fact is a reason to divert to a different landing site
see that a few seconds prior spark generators start to fire,
— but this is a simulation. Should you ever wish to simu-
which then ignite the gas mixture stream that appears as
late a Shuttle landing in rain, this is actually supported, you
soon as the valves open. At first the flame is faint red as
can expect rain splashes on the windshield and the spray of
combustion is inefficient, then it brightens quickly to an
water emerging from the wheels as you touch down. In fact,
orange color before it finally stabilizes and settles to the
landing in poor weather is quite a challenge since you have
near-invisible flame that will remain during ascent. During
to completely rely on the HUD overlay to give you accu-
the first moments, the flame is far more turbulent than later
rate information on where the runway is right till the final
on, and also the distribution of shock diamonds has not yet
seconds.
reached their final pattern.
You can also try to land the Shuttle during a thunderstorm
This whole behavior is actually modeled in the simula- — but then you will discover rather quickly why any storm
tion, so it’s worth watching the engine ignition sequence in the vicinity is a reason to divert to a different landing site!
from an outside view once in a while.

Faulty sensor readings

In general it is fairly easy to make a simulated sensor show
a wrong value. It is much harder to make that value plau-
sible and characteristic for a certain fault. Consider for in-
stance the air data probes. Both altitude and Mach number
are derived from a combination of ram pressure and static
pressure readings. If the static pressure measurement fails
because the tube is partially blocked and the sensed value
is too low, the formulae to deduce altitude and Mach num-
ber will show characteristic deviations — both altitude and
Mach number will be over-estimated, but due to the expo-
nential pressure drop in the atmosphere altitude usually will
be dramatically off.
Whenever possible, such deviations characteristic for a
particular failure of a piece of equipment have been entered
into the simulation (although admittedly in many cases this
is not easily possible).

Re-entry visuals
How exactly does the Shuttle look like when it is exposed
to the fierce heat of atmospheric entry? It seems completely
impossible to take a photograph of this as reference mate-
rial, but exactly that has been done and underlies the visual
pattern of the glow. During an entry, the airborne Spitzer
Infrared Telescope has been pointed at the Shuttle from be-
low and taken a shot of the temperature distribution across
the TPS at Mach 12 — this is not quite at the peak heating
phase, but nevertheless provides rather detailed information
on where the heat has been fiercest. This picture has been
used to extract a temperature map, which has been applied
to create the distribution for all entry phases.
110 CHAPTER 10. TRIVIA
 Part III

Orbital maneuvering and Rendezvous

111
Chapter 11

Orbital maneuvering basics

In this chapter, we will go over the basic ideas behind defined as perpendicular to the orbital plane (and as such, it
orbital maneuvering, i.e. how to go from one stable orbit is perpendicular to the prograde vector). Finally, the radial
by a series of maneuvering burns into any other stable orbit. direction is perpendicular to both prograde and normal and
We’ll do this in an idealized setup and ignore a couple of completes an orthonormal coordinate system. By virtue of
real-world complications — most importantly perturbations being orthogonal to the normal vector, it also must lie in the
to the orbit due to the fact that Earth is not a perfect sphere. orbital plane.
The math behind computing the required burns typically For a spacecraft in a circular orbit, the radial direction
involves computing orbital elements from state vectors and is always outward along the line from the center of Earth to
vice versa. It is not not hard if you know how to do vector the spacecraft. For an eccentric orbit, this is not exactly true,
algebra, but it is somewhat beyond the scope of this man- but for all orbits the Shuttle is able to reach, the difference
ual — if you want to compute some examples yourself, all is too small to matter.
expressions are on Wikipedia and there are various online
tools for the task found in the internet.
The prograde burn
The coordinate system Each maneuvering burn can be characterized by the velocity
change ∆v by which it alters the state vector (a burn does
As briefly mentioned before, to describe maneuvering not directly alter the position part of the state vector). How-
burns, special coordinate system of prograde, normal and ever, to see what a burn does, it is most useful to express
radial directions is most useful. the velocity change along the prograde, normal and radial
Let’s recapitulate the essentials: The spacecraft in a direction.
closed orbit moves in an ellipse (in inertial coordinates) If a burn alters only the prograde velocity component,
which always remains in the same 2-dim plane, the orbital it accelerates (or decelerates) the spacecraft along its cur-
plane. Earth’s center of mass is always in one focal point rent trajectory. Such a burn alters the kinetic energy of the
of that ellipse. Two orbital elements (inclination and longi- spacecraft, but because it is in the orbital plane, it can not
tude of the ascending node) describe the orientation of the alter the orientation of the orbital plane. In fact, a prograde
orbital plane. Two more (semimajor axis and eccentricity) burn alters the shape of the orbit ellipse.
describe the shape of the ellipse, one describes the orien- If the orbit is eccentric, it alters in general both semi-
tation of the ellipse (argument of the periapsis) and the fi- major axis and eccentricity — or, if one prefers, apoapsis
nal element (true anomaly) describes where on the ellipse and periapsis. How precisely depends on where the burn
the spacecraft currently is. For a situation without pertur- is done. If it happens exactly at the apoapsis, a prograde
bations, all elements except the last one are unchanged in burn alters only the value of the periapsis. If done exactly at
time. the periapsis, it alters only the value of the apoapsis. If done
Equally well the orbit is uniquely described by the state anywhere between apoapsis and periapsis, it alters both val-
vector at a given time. This consists of three position and ues, but the closer it is to the periapsis, the more of the
three velocity vector components. However, all components change affects the apoapsis and vice versa.
of the state vector change in time. A prograde burn does not directly alter the argument of
The prograde direction is defined to be a vector of unit the periapsis, but since it can turn an apoapsis into a new
length along the current velocity vector (and a burn into the periapsis (and vice versa), it can flip the parameter by 180
opposite direction can be called retrograde). The prograde degrees.
vector is always in the orbital plane. The normal direction is Maneuvers where a prograde (or retrograde) burn is exe-

113
114 CHAPTER 11. ORBITAL MANEUVERING BASICS

A few useful expressions for orbits:

If the radius at the periapsis is RP , the radius at the apoapsis RA and Earth’s radius RE then the orbital altitudes for the
periapsis and apoapsis above Earth are

rP = RP − RE and rA = RA − RE

The eccentricity e is then

RA − RP
e=
RA + RP
and the semi-major axis a is
RA + RP
a=
2
Conversely, given a and e, the radii are

RA = (1 + e)a and RP = (1 − e)a

two orbits intersect over the equator. Assume the spacecraft

is at the ascending node — any northward normal burn will
increase inclination, any southward velocity change will de-
crease inclination. At the descending node, the pattern is
reversed.
If the original orbit is not circular, the argument remains
valid in principle, but then a simple normal burn is not suf-
ficient to change one orbit into the other. In fact, let’s con-
Figure 11.1: Effect of a prograde burn (at black square) on sider two observations: First, a transfer from one circular
a circular orbit. to another orbit through a normal burn requires to change
the velocity component at the intersection point — if that
change is large, say 90 degrees, in essence the whole veloc-
ity of the spacecraft in one direction needs to be killed and
cuted at an apsis are known as raising or lowering the op- then re-established into another direction — large inclina-
posite apsis. For instance, if the Shuttle after MECO is in tion changes are propellant-costly.
an eccentric orbit with an apoapsis of 400 km and a periap- But second, the magnitude of the orbital speed is higher
sis of 120 km altitude, a burn at the apoapsis can raise the closer to the planet and decreases far away. Thus, any or-
periapsis to 400 km and circularize the orbit. bital plane change on an elliptic orbit is best done far from
the planet to save propellant. Another way to understand
this is to realize that the lever arm to to tilt the angular mo-
The normal burn mentum vector is larger far away from the planet, hence the
task is easier.
When a burn is directed out of the orbital plane, kinetic en- In actual fact, normal burns play very little role in Shut-
ergy of the spacecraft is not changed. As a result, the shape tle maneuvering, because realistically the OMS propellant
of the ellipse that describes the orbit is unchanged. What is reserves are insufficient to change inclination by more than
however altered is the direction of the angular momentum a few tenths of a degree - inclination has to be established
vector — in other words, a normal burn changes the orbital during launch and remains more or less fixed for the rest of
plane. the mission.
Dependent on when it is executed, it may affect inclina-
tion, longitude of the ascending node or both parameters.
The effect of a normal burn can be pictured as follows: The radial burn
Any two circular orbits intersect at exactly two points. A
burn at any of these points that changes the velocity of the A radial burn again is in the orbital plane and hence inca-
first orbit into the velocity of the second orbit transfers a pable of changing inclination or longitude of the ascending
spacecraft from the first into the second orbit. node. What it does is to rotate the apses around the nor-
For instance, if only an inclination change is desired, the mal axis, i.e. the location of apoapsis and periapsis shift
IN-PLANE ORBITAL MANEUVERING 115

and in addition the eccentricity may change. Since the ki- However, prograde (and retrograde) burns may equally
netic energy of the spacecraft does not change (the burn is well be used to move the argument of the periapsis. To
perpendicular to the velocity vector), the orbital period re- go from one elliptic orbit to another one with the same ec-
mains unaltered. centricity and semi-major axis but different argument of the
For instance, if you do a radial burn inward from a cir- periapsis, the required sequence of burns is
cular orbit, you will end up distorting the orbit. First the
spacecraft will move inward, but 90 degrees later it will • wait till you reach the apoapsis
reach the new periapsis, then go outward again, reach the
former orbital radius 180 degrees away from the burn and • at the apoapsis circularize the orbit by a prograde burn
then a new apoapsis 270 degrees away — in other words,
the eccentricity of the orbit is changed. • wait till you are 180 degrees away from the desired
new periapsis location

• burn retrograde to lower the periapsis to reach the de-

sired eccentricity

In fact, insertion and de-orbit are only special cases of

these general sequences. After MECO, the Shuttle is usu-
ally on an elliptic orbit moving towards the apoapsis, and
the insertion burn circularizes that elliptic orbit. In contrast,
the de-orbit burn lowers the periapsis such that the orbit in-
Figure 11.2: Effect of a radial inward burn (at black square) tersects the atmosphere a few thousand miles ahead of the
on a circular orbit. landing site.
It should be apparent at this point that orbital maneuver-
ing is very different from controlling an airplane or even
On the other hand, assume the spacecraft is right after the
a rocket during launch — at its core, it is a waiting game.
apoapsis and starts to descend — by burning outward, the
Burns are just nudges given to the trajectory at just the right
descend can be slowed or halted, and so the periapis will
location, and all orbital transfers take time — there’s no
be reached later (it will be reached eventually because the
such thing as an immediate transfer into a different orbit.
burn does not alter orbital energy) — the location of the
apses rotates away from the spacecraft.
Generally in actual maneuvering radial burns are most
interesting to compensate for ignition time issues - assum-
Changing the orbital plane
ing for instance you’d like to circularize an orbit, but for The location of the orbital plane is characterized by the in-
some reason can’t execute a burn right at the apsis. Adding clination and the longitude of the ascending node. Suppose
a small radial component to it will give you the same out- two orbits are circular with the same radius but in a dif-
come if you burn a bit later. ferent plane — then it’s easy to see that regardless of what
their parameters are, they intersect in exactly two points, the
In-plane orbital maneuvering so-called nodes.

When staying in the orbital plane, the transition from any

orbit to any other orbit can be accomplished by a sequence
of properly timed prograde and retrograde burns only, no
radial burn is required.
One of the simplest examples is the Hohmann transfer.
This is a transfer between two circular orbits, either to a
higher or to a lower orbit. Assuming that we’d like to trans-
fer to a higher orbit, the sequence of burns is Figure 11.3: Effect of a normal burn (at black square) to
change the orbital plane (here inclination).
• burn prograde till the new apoapsis is has higher as the
desired radius of the target orbit
• wait till you reach the apoapsis If the orbits differ in inclination only, the nodes are lo-
cated right at the equator. If they differ just a bit in the
• at the apoapsis circularize the orbit by a new prograde longitude of the ascending node, the nodes are very close to
burn the point where the orbits reach the largest latitude.
116 CHAPTER 11. ORBITAL MANEUVERING BASICS

Rules of thumb for the Space Shuttle:

1% of propellant ∼
= 6 fps velocity change
1% of propellant ∼
= 6.6 km (3.5 nm) apsis change
1% of propellant ∼
= 0.013 degree inclination change
100 fps (or 16% propellant) ∼
= 110 km (60 nm) apsis change
100 fps (or 16% propellant) ∼
= 0.22 degree inclination change
Note that transfer between circular orbits requires two apsis changes!

The general idea of a plane-changing burn is to make it In particular, the Shuttle is completely incapable of
alter the velocity vector from the orginal orbit to the tar- changing the orbital plane by more than a fraction of a de-
get orbit right at the node (where the position vector is gree using the OMS engines, thus the orbital plane is to
the same). Thus, conceptually any plane change can be a very good approximation established at launch and all or-
achieved with just a single burn. For instance, to lower in- bital maneuvering done during a mission is in-plane. In fact,
clination, a spacecraft needs to execute a normal burn ori- changing the plane is so hard that when OMS fuel needs to
ented into the southward direction when crossing the equa- be dumped prior to entry, this is done by adding a normal
tor from the southern to the northern hemisphere — or al- component to the de-orbit burn on the grounds that it does
ternatively a burn oriented into the northern direction when not significantly alter the trajectory anyway — the effect
crossing from northern to southern hemisphere. of the atmosphere on the trajectory due to the high bank
To achieve a plane change, the original and the target angle flown during entry is higher by far. Only during a
orbit don’t have to actually intersect or be circular — the rendezvous mission, small corrections to match planes with
nodes are still well-defined concepts if one considers the the rendezvous target are done.
projection of the two orbits to a standard radius — in this Also, higher orbits generally have less margin for prob-
case, transferring from the original to the target orbit in- lems (such as propellant leakage, or excessive RCS usage,
volves in-plane maneuvering as well. to be compensated via an OMS to RCS cross-feed) due to
In-plane and plane-changing burns can be interchanged the higher requirements for insertion and de-orbit.
rather freely, but since changing the orbital plane is equiv-
alent to changing the angular momentum of the spacecraft, Note: If the propellant is depleted too much for a
due to the larger lever-arm at large radii it is less-propellant- de-orbit burn to intersects the atmosphere, there is no
costly when far away from the planet. Thus, a maneuvering way for the Shuttle to return to Earth.
plan to transfer from a high circular orbit to a low circular
orbit at different inclination would be to The consequence is that the launch (both in terms of tim-
ing and in terms of azimuth) needs to be carefully planned
• wait till the equator is crossed to bring the Shuttle into the desired plane — it can’t be ma-
neuvered there later.
• burn to change inclination
• burn to lower periapsis
Maneuvering using PEG-7 targeting
• wait till periapsis is reached
As we’ve seen above, orbital maneuvering requires to alter
• burn to lower apoapsis
the velocity components of the state vector at certain times.
whereas if the task would be the opposite transfer, the Thus, a maneuvering plan for the Shuttle can be cast into the
in-plane maneuvering would be done first. form of a list of ignition times and the three components of
∆v (in the prograde, radial and normal coordinate system).
Since these velocity components are not computed by the
Finite propellant Shuttle avionics, this is known as external targeting or the
Powered Explicit Guidance 7 (PEG-7) scheme.
In practice, the Shuttle only has OMS propellant for about During orbital operations, the Shuttle avionics is directly
600 fps of velocity change. Given that the propellant is set-up to be programmed with PEG-7 burn targets — the
needed to do the insertion into orbit (the OMS-2 burn) and relevant DPS page is OPS 202 (see Fig. 11.4).
to do the de-orbit burn, this places rather tight limits on what Burns from a maneuvering plan have to be entered one
the Shuttle can actually do. by one, only the next to be executed can be programmed
MANEUVERING USING PEG-7 TARGETING 117

The following expressions can be used to determine the prograde (or retrograde) ∆v requirements for a Hohmann transfer:
Taking the standard gravitational parameter GM as 3.986004418 · 101 4 m3 s− 2 and R1 as the radius of the inital circular
orbit and R2 the radius of the desired circular orbit, the requirement to enter the proper elliptical transfer orbit is
r r !
GM 2R2
∆v1 = −1
R1 R1 + R2

and the velocity change to leave the transfer ellipse and circularize at R2 is
r r !
GM 2R1
∆v2 = 1−
R2 R1 + R2
If R1 and R2 are not very different (as is the case for the orbits the Shuttle can reach), the two values also won’t be
radically different, explaining why the rules of thumb can be formulated for ’apsis change’ in general.

MET and add 15 minutes — the time so obtained we pro-

gram as ignition time TIG in items 10-13.
To get the 50 fps retrograde burn, we set item 19 to -50
and load the plan using item 22. The reward ought to be that
TGT HP on the left side is lowered by about 30 miles. Using
automatic maneuvering (item 27) and the timer to maneu-
ver (item 23) and the EXEC key when the timer reaches 10
seconds (just like for the insertion burn) the first maneuver
can be executed.
At this point, TTA in the center of the display will indi-
cate when the new periapsis is reached. Again, by looking
at current MET and adding that time, we can determine TIG
for the second circularization burn (it can be programmed
right after the insertion burn, but especially if attitude con-
trol thrusters are used, the result will be more exact if we do
Figure 11.4: The MNVR page used for all orbital maneu- it again just 15 minutes ahead of time).
vering of the Shuttle. With the new TIG and again a retrograde burn of 50 fps,
we should then approximately get into the desired orbit.
Of course the maneuvers done that way are not very pre-
cise — you can expect to miss the apsis by as much as a
ahead of time. All ignition times are specified in mission few miles. To get better than that requires something far
elapsed time (MET) and entered as items 10 to 13. The beyond rules of thumb or analytical expressions — a real
PEG-7 targets are then entered as items 19 (prograde), 20 offline orbital solver.
(normal) and 21 (radial) and loaded via item 22, at which
point the prediction for the apses on the center-right part of
the display is updated.
Since during OPS 2, only external targeting is available,
we need to use some external tool to determine them —
either the rules of thumb shown before, or analytical ex-
pressions for e.g. a Hohmann transfer, the Shuttle avionics
won’t help you.
To do a Hohmann transfer from a circular orbit into an-
other circular orbit 30 miles lower, we could for instance
use the rules of thumb to select retrograde burns of 50 fps.
The first burn of the maneuver could be at any time, so to
give the Shuttle time to maneuver into burn attitude and
us the opportunity to check everything, we look at current
118 CHAPTER 11. ORBITAL MANEUVERING BASICS
 Part IV

Reference material

119
Chapter 12

DPS pages

GNC functionality missing. The items shown on the display are all meaningful
— item 2 shows the current contingency abort region color
Guidance, navigation and control software is available in all code and is used to arm the abort, item 4 executes a con-
OPS, usually run on a redundant set of three or four GPCs tingency abort. Item 5 toggles whether yaw steering during
running PASS. a contingency abort should be used or not (this is only se-
lectable before the abort). Item 6 activates single engine roll
control (SERC) — normally this should get active automat-
ASCENT TRAJ (OPS 101, 102, 103) ically as soon as the avionics detects a two engine failure,
The trajectory displays (the display title is LAUNCH TRAJ but if for some reason this does not happen, SERC can also
1 prior to liftoff in OPS 101 and ASCENT TRAJ 1 after be selected manually.
liftoff in OPS 102) are the main DPS pages for the flight
into orbit. In the graphical portion of the display, they show MNVR (OPS 104, 105, 106, 202, 301, 302, 303)
the desired trajectory as a function of speed (ground-relative
on ASCENT TRAJ 1, inertial on ASCENT TRAJ 2) on the One of the most often used DPS pages, the maneuver dis-
horizontal and altitude on the vertical axis. Numbers on AS- play is used in all ops to monitor the current orbit and pro-
CENT TRAJ 1 indicate the approximate pitch values which gram, execute and monitor OMS burns — be it for orbital
should be flown at the indicated points. The Shuttle is repre- insertion, de-orbit or just small corrections. Items 1-4 on
sented as a triangle with 30 second predictor circles shown. the upper left allow to select which engines are used for a
Numerical readouts display current throttle setting and (on burn, the selection ranges from double and single OMS en-
TRAJ 2) remaining propellant. gine burns to an RCS translational burn.
tem 5 allows to set the roll attitude during the burn —
this defaults to 180 degrees, i.e. head-down, but can be set
to any value desired.
Item 9 is used to specify the current weight of the or-
biter such that the requested velocity changes can be used
to compute burn durations. Items 10-13 allow to specify the
planned ignition time.
The lower part of the left column is used to enter burn
targets, both as PEG-4 and PEG-7 (items 19-21). Executing
item 22 loads the burn targets into guidance, computes burn
duration and a prediction for the orbital changes. Item 23
sets the CRT timer to count to the specified ignition time.
The current and future orbit can be seen in the center of
the right column where TGT HA and HP show the target
height of apoapsis and periapsis whereas CUR shows the
current values. Left to it is another important number, TTA
Figure 12.1: The OPS 101 page. (or TTP) shows the time to apoapsis (or periapsis) which is
useful to plan the timing of burns. Also, for de-orbit burns,
REI just above shows the predicted range from entry inter-
The displays in-sim differ from the original in that la- face to landing site (only in OPS 3) and the timer shows the
bels for the target pitch at certain parts of the trajectory are time to the interface as TFF.

121
122 CHAPTER 12. DPS PAGES

Once a burn is entered and verified, item 27 can be used inertial vector or a chosen target. It also displays current
to automatically maneuver the Shuttle into attitude if the inertial attitude, attitude target, attitude change rates and at-
orbital DAP is set to AUTO. Once in attitude, a flashing titude errors.
EXEC will appear in the upper right corner as soon as the
specified ignition time is reached and needs to be confirmed
by the EXEC key to initiate the burn.
When a burn is in progress, DVTOT and TGO show the
integrated velocity change and elapsed time of the burn. Be-
low are the body coordinate components of the sensed ve-
locity changes, the sensed values subtracted from the calcu-
lated ones (since the OMS engines are not aligned with the
body axis, always expect more than one component to build
velocity). At the end of the burn, these should be close to
zero, and any residuals can be nulled using the RCS as long
as the display is of type MNVR EXEC.
The lower part of the center column is devoted to gimbal
management — item 34 can be used to execute a gimbal
check, and items 28-33 are used to select which gimbal sys-
tem is used for which engine. Above are the current gimbal
angles of the engines in pitch (P) and yaw (Y). Figure 12.3: The OPS 201 page.
The lower part of the right column is relevant for off-
nominal situations. Using item 35 burn targets for an ATO
or AOA can automatically be computed. Items 36 to 38 The layout of the display in-sim is identical to the orig-
finally allow to execute a forward RCS propellant dump. inal, but not all functions are available. Items 1-4 allow
to enter the time of a future maneuver. Items 5-7 specify
inertial attitude pointing angles. Item 8 allows to select a
tracking target, valid IDs are 2: center of Earth 3: Earth-
relative target specified by items 11-13, 4: center of the Sun
or the number of any star in the star catalog (see appendix
B).
Items 11-13 allow to enter latitude, longitude and altitude
of an Earth relative target (used for ID 3). Item 14 allows
to choose the body vector to be aligned, valid selections are
1: +X, 2: -X, 3: -Z axis. Item 15-17 are for entering offsets
to the chosen body axis, but only item 17 (the roll offset
omicron) is implemented.
Item 18 starts a current or future pointing maneuver, item
19 tracking and item 20 a rotation around the current axis.
Item 22 allows to display all angles below with respect to
the -X axis when piloting the Shuttle from the mission spe-
Figure 12.2: The OPS 104 page. cialist station. No other functions are implemented.

Almost all functions on the display are supported, includ- ENTRY TRAJ (OPS 304)
ing PEG-4 targeting in OPS 1 and 3 (the functionality is not The entry trajectory displays (a series of five) is used for
provided in reality in OPS 2). Items 24-26 can be used to about the last 2000 miles to landing site to manage the
override the automatic burn attitude computation in reality, hypersonic part of the atmospheric entry. In the center, it
this is however not yet possible in the simulation. shows the nominal trajectory on a plot of range to site vs.
velocity. This is surrounded by other tranjectories. The out-
UNIV PTG (OPS 201) ermost lines represent thermal and structural as well as low
energy limits. A Shuttle symbol displays the current state
The universal pointing page supplies automatic attitude of the orbiter and is followed by a series of tracers to gauge
control routines in orbit, such as pointing and tracking the trend of the trajectory. A square symbol called the guid-
where a Shuttle body axis is aligned with either a specified ance box shows into which direction the Shuttle is going
GNC FUNCTIONALITY 123

to evolve. For converged guidance the box is on top of the

Shuttle symbol. Point tracers follow the guidance box to
gauge their trend. The dotted lined represent the specific
drag value indicated in their annotation.
The scale to the far left shows the nominal AoA for the
current Mach number (arrow) and the actual AoA (trian-
gle). Directly right of it is the drag scale, showing desired
drag (arrow) and actual drag (triangle). The scale on top,
known as the phugoid scale shows the roll commands. The
alphanumerics show on the upper left azimuth difference to
the landing site and on the lower right side force, trims and
drag bias. Item 4 allows to (de-)select special low energy
flight rules for the autopilot.

Figure 12.5: The OPS 305 page.

The lines on the upper left are the alpha vs. Mach sched-
ule to be flown during a contingency abort and are not used
in nominal situations.
The display as implemented is a somewhat simplified
version of the real counterpart. The energy indicator is more
trustworthy than the limiting trajectories and should be used
when flying manually to gauge what to do.

RTLS TRAJ (OPS 601)

Figure 12.4: The OPS 304 page. During the powered phase of a return to launch site (RTLS)
abort, the OPS 601 trajectory display is shown for orien-
tation. The central graphical portion of the display shows
Since internally the guidance does not work quite on the two reference trajectories — a lower one for an early RTLS
same principles as the original Shuttle PASS software, al- and two parts of a late RTLS. Both trajectories display site-
phanumerics are only partially meaningful, but the low en- relative velocity on the x-axis vs. altitude on the y-axis. The
ergy entry mode is correctly implemented and toggling item zero site-relative velocity line is shown in the middle.
4 will instruct the AP accordingly. The actual trajectory flown during an abort is typically
in the space between the two sets. Note that none of the
VERT SIT (OPS 305) trajectories is meant as hard limit, i.e. the actual trajectory
may be temporarily outside the set, and this is no reason for
The two vertical situation displays are used during TAEM concern. As in the ascent trajectory display, the Shuttle is
and final approach to manage altitude during the remaining shown as a triangle with two predictors for estimating the
range to touchdown. The center part of the display shows trend.
the nominal trajectory as a plot of altitude over remaining The options are identical to ascent — item 2 arms a
range, framed by an upper and lower limiting trajectory. contingency abort, item 4 executes the contingency abort,
The upper represents the 270 KEAS limit that should not item 5 toggles yaw steering enabling, item 6 activates sin-
be exceeded during TAEM, the lower is an approximate op- gle engine roll control in case this doesn’t activate auto-
timal lift/drag curve. A Shuttle symbol slides down on it, matically. The only unique parameter is GUID in the up-
with the Shuttle’s angle indicating the current descent rate. per right which shows how long guidance estimates for the
To the far right are indicators for nose position and an pitch-around maneuver (change from flying away from the
energy estimator. The energy estimator has an upper line site to flying back towards the site) — pitcharound happens
above which S-turns should be flown to deplete energy, and when the counter reaches zero.
a lower line below which the approach pattern should be The display is a somewhat simplified version of the
down-moded to a closer aim point. Alphanumerics show real one, but all options (including the color-coded RTLS-
commanded and actual state of the speed brake. specific contingency aborts) are fully implemented. When
124 CHAPTER 12. DPS PAGES

manifold and the status of the manifold valves, the aerosur-

face section shows the surface positions. All others show
symbols only if the device is not responding.
In the simulation, RCS jet indicators, aerosurface po-
sitions, TACAN, air data probes (ADTA) and microwave
landing system (MLS) are properly shown, failures for the
other devices are not simulated and hence the fields will al-
ways be blank.

SYS SUMM 2 (DISP 19)

The second GNC systems summary page shows the detailed
status of the RCS and OMS thrusters. The upper part dis-
plays the OMS, with fuel (FU) and oxidizer (OXID) quanti-
ties in the upper left, and the various tank pressures ranging
Figure 12.6: The OPS 601 page. from the He-system and the N2 system to the regulators and
valves.
The RCS is divided into three sections showing forward,
left and right with helium pressures, tank pressures and the
flying the RTLS manually, the aim is to steer the Shuttle to- individual manifold valves and the status of the isolation
wards the left-most point of the trajectory with less then 5% valves. Jet failure announcement per manifold is also avail-
propellant left in the ET. able here.
Note that the correct RTLS site needs to be specified in
the mission file for the display to show meaningful guidance
information, as FG can not know your launch site from the
start parameters.

SYS SUMM 1 (DISP 18)

The first GNC systems summary page shows the status
summary of RCS jets, aerosurfaces, the various hand con-
trollers — rotational (RHC), translational (THC) and speed-
brake/thrust (SBTC), of the GPCs, the flight control sys-
tem and the various navigational sensors (inertial units, ac-
celerometers,. . . ).

Figure 12.8: The DISP 19 page.

This display is fully functional and the parameters re-

spond to problems with the propellant distribution, pressure
losses or valve opening and closing in a plausible way.

DAP CONFIG (SPEC 20)

The DAP configuration page is used to set the parameters
used by the various orbital DAPs, in particular gains, dead-
bands and thrusters to be used. The first column is used to
configure DAP A (with the primary, alternative and Vernier
Figure 12.7: The DISP 18 page. jet selections), the second column does the same for DAP
B. The current selection is shown in overbright. The third
column is used to edit a currently not used configuration
The RCS jet summary shows failures sumarized for each that can later be assigned to a DAP.
GNC FUNCTIONALITY 125

items 10 to 12 and the alignment method can be chosen

using items 13-16.
The center right block allows to set BITE masks for the
IMUs to exclude certain failure modes from the self-testing,
and the lower right allows to set sensitivity thresholds for
the navigation model and to access the memory units.

Figure 12.9: The SPEC 20 page.

In FG, only the first two columns are functional, and

while the alternative selection can be edited, it has no effect
on the flight dynamics. The various items set (for DAP B
and Vernier analogous): 10: target rotation rate 11: attitude Figure 12.10: The SPEC 21 page.
deadband 12: rate deadband 13: gain in rotational speed for
a single pulse in pulsed mode 14: whether thrusters should
be fired to compensate the rotation caused by translational In FG, all readouts and mode options are meaningful, and
thruster firings 15: whether forward, aft or all RCS pods are the selection filtering is fully simulated. Accelerations and
used for pitch 15: whether forward, aft or all RCS pods are attitude sensed will subtly differ based on the IMU mis-
used for yaw 17: velocity change for a single translational calibration model, and all three alignment methods on the
pulse in pulsed mode. page are fully supported.
The options described for DAP A and B are actually Currently no BITE masks can be set, and no mass mem-
functional and there are no sanity checks done on the val- ory access is possible, likewise the threshold cannot be al-
ues entered, thus be aware that it is quite possible to enter tered.
values which do not lead to converging automatic solutions
(too tight deadbands) or are extremely wasteful with pro- STRK/COAS (SPEC 22)
pellant.
The function of the star trackers and the crew optical align-
ment system (COAS) is controlled from SPEC 22. There
IMU ALIGN (SPEC 21)
are four different blocks on the page. The upper left allows
The IMU ALIGN page is used to monitor the function of to select the operating mode for the two star tracker sys-
the on-board inertial measurement units (IMUs) and to re- tems, the lower left shows the status output for the trackers,
calibrate them when needed. On the upper left is a status the upper right shows a table of the last stars that have been
display which shows for each of the three units its own sta- tracked, and the lower right is used to operate COAS.
tus description and the temperature string (which indicates The status of the star trackers can be NO TARGET (there
whether the cooling fans are working). Items 21-23 allow to is no star in view of the current camera), HI RATE (the
down-mode each unit to standby, items 4-6 allow to mode Shuttle’s angular motion is too high to do tracking), FALSE
them back to operational and items 7-9 allow to de-select TRK (the orientation error of the state vector is too high to
each unit from the filter, which prevents their data from in- do star tracking, a COAS procedure needs to be performed)
fluencing the state vector. and OUT FOV (a rendezvous target to be tracked is not in
On the center left is a readout of the raw accelerations view of the camera). The shutter status is usually OP (open)
sensed by all three units for comparison. On the lower left unless the Sun is in camera view when it changes to CL
are the currently sensed attitudes (X, Y, Z) and the expected (closed).
attitude changes during an alignment process (∆X,. . . ). While the star tracker is not simulated in all details, the
The alignment process itself is controlled on the upper simulation includes the essentials of the operation and al-
right, where the unit to be aligned can be selected using lows to do COAS procedures. Items 1 to 10 set the star
126 CHAPTER 12. DPS PAGES

Figure 12.11: The SPEC 22 page. Figure 12.12: The SPEC 23 page.

trackers mode to self-test, star track (the normal function (items 9, 11, . . . 39) are used to deselect a jet (i.e. cease to
for attitude control), target track (usable during rendezvous issue any firing commands to it) — this is useful to not use
operations), instruct to break the current tracking and into thrusters which have failed ON in the past. Items 40-44
idle. Item 13 and 14 set a magnitude threshold on what stars allow to toggle the manifold valves — these are software
are tracked while items 15 and 16 can be used to manually overrides for the valve switches on the overhead panels.
control the shutter (which protects the tracker cameras from Doing so cuts the fuel supply of a thruster failed on and de-
the glare of direct sunlight) manually using items 15 and 16. activates it. Using item 51, this can be done automatically
Items 17 to 19 (de-)select a star in the above table for if the system detects a fail on thruster.
reference, item 20 clears the star table. SPEC 23 should be used to diagnose any weird behavior
COAS operations require to point the Shuttle towards a of the RCS and to fix it by de-selecting failed thrusters (and
star and are described in detail in chapter 8. Item 21 allows their symmetrical counterparts) to retain a balanced RCS
to enter a reference star ID, item 22 selects which Shuttle action.
axis is to be oriented, item 23 is used to accept a COAS
mark for the chosen star, items 26 and 27 select where the RM ORBIT (SPEC 25)
COAS mask is mounted and items 28 and 29 instruct GNC
to use the currently processed COAS mark to update the
state vector.

RCS (SPEC 23)

The RCS utility page is used to monitor and control the var-
ious RCS thrusters (aka, the jet table). Next to RCS fuel
pressure readings on the right side, it shows the status of
every individual thruster on the left side in terms of failure
mode (FAIL), automatic deselection inhibition (DES INH),
deselection (JET DES) and priority (PTY). The top part of
the page allows to set various general configuration options.
The failure modes displayed can be OFF (thruster are re-
ceived a firing command but is not on), ON (thruster has
received a shutdown command but remains on) and LEAK
(thruster is off but losing fuel).
In FG, most of the items are functional, though the au- Figure 12.13: The SPEC 25 page.
tomatic deselection inhibition and the crossfeed quantity
gauging options are not. Items 1-3 are used to select which
RCS system is to be shown (forward, left or right) and the The orbital redundancy management utility has the func-
jet column labels will change accordingly. The deselections tion of testing and managing the flight controllers. If Com-
GNC FUNCTIONALITY 127

mander or Pilot rotational hand controller (RHC) or trans-

lational hand controller (THC) are deflected, the display
shows the signal sensed from the device for diagnostic pur-
poses.
In the simulation, this is a fairly involved way to test
whether your joystick is properly connected. Dependent on
the chosen view and the control mode, the system will sense
your control input as coming from Commander or Pilot side
and from RHC or THC respectively. It’s unlikely you’ll ever
use this display much, but it’s there just in case.

REL NAV (SPEC 33)

The relative navigation specialist function page is used for
rendezvous navigation. It is a conceptually fairly compli-
cated page which can’t be explained in a few lines. Very Figure 12.15: The SPEC 34 page.
simply put, it allows to specify a device (Ku band antenna
as radar ranger (RR), star tracker (S TRK) or crew optical
alignment system (COAS) to act as sensor for obtaining the
rent position to an intercept with the target for specified ini-
relative state vector between Shuttle and rendezvous target,
tial and final times. In addition, the Shuttle’s current state
and then allows to configure how exactly the raw sensor
vector is accessible in the lower right of the display.
information should be incorporated into the relative state
The page is functional in a simplified way. Items 2-5 are
vector.
used to enter the ignition time (TIG-1) for the first maneu-
ver of the two burn sequence, items 13-16 do the same for
the second burn of the sequence. Item 17 can be used to
enter the desired transfer time if TIG-1 is already set (this
automatically fills 13-16). Items 18 to 20 are used to enter
offsets (in kft) to the target in proximity coordinates. The
computation can be started using item 28, the parameters
of the first burn will then be displayed in the upper part of
the page when the computation finishes. Finally, item 29
is used to access the parameters of the second burn. Both
items 28 and 29 will automatically enter PEG-7 parameters
into the MNVR page.
For correct function, the page needs an orbital target i-
loaded in the mission file and all transfers will be computed
to that target.

Figure 12.14: The SPEC 33 page. HORIZ SIT (SPEC 50)

The horizontal situation special function page serves a dual
purpose. In orbit, it allows to enter the landing site as item
The implementation of relative navigation does not in- 41 (see valid indices in Appendix B — in OPS 1, also the
clude all the details of raw sensor information and Kalman TAL site can be chosen as item 40 which then automatically
filtering, but is done to give the basic flavour of the pro- becomes landing site if an abort is called).
cedure. The relative state vector displayed will have errors, During approach, items 3 and 4 allow to change the run-
and these errors can be reduced by utilizing one of the avail- way. Items 7 and 8 toggle the TAEM pattern selection from
able sensors. overhead to straight-in and from nominal entry point to min-
imal entry point.
ORBIT TGT (SPEC 34) When TACAN receivers come into range, the lower por-
tion of the display shows range and azimuth or ∆ azimuth
The orbital targeting page provides the interface to the Shut- (items 34 and 35 select between the options). The lower
tle’s Lambert solver. In short, such a solver can compute a left portion manages incorporation of other data (GPS, drag
two-burn solution for a transfer orbit from the chaser’s cur- altitude or air data altitude) into the state vector.
128 CHAPTER 12. DPS PAGES

Figure 12.16: The SPEC 50 page. Figure 12.17: The SPEC 51 page.

The central portion of the display shows a graphical rep- The upper right part selects flight control options and at-
resentation of the TAEM pattern, with the touchdown point mosphere model for the entry part of a mission. Below
a dot, connected with a straight line (the final approach) tothe status of the IMUs is displayed and options to de-select
the heading alignment cone. The Shuttle is shown fixed in failed IMUs are provided. Further below the air data probes
the lower portion of the display, with a series of predictor (ADTA) display their sensed parameters (if deployed) —
dots indicating the trajectory trend. altitude, AoA and Mach number and items 34 to 37 allow
Landing site and TAEM pattern selection and the incor- to de-select failed probes to prevent them from being incor-
poration of the various data sources are fully functional. porated into the state vector.
Also a manual state vector update via items 10 to 16 is fully Finally, the lower right part provides overrides for the ET
supported. separation switch (items 38 and 39), the ET umbilical door
Note: The display will only show guidance information switch (item 40) and the manual vent door control switches
if the correct landing site is chosen and guidance is active (item 43 and 44).
— this is not the case by default if you select e.g. a TAEM In the simulation, the left part of the display is fully im-
scenario directly. Also, the landing site index needs to point
plemented. From the entry FCS, only the repositioning of
to a valid site where guidance is implemented. the SSMEs when hydraulics powers up again can be en-
abled and disabled (item 19) and the atmosphere model can
OVERRIDE (SPEC 51) be selected using items 22 to 24 but filters can not be cho-
sen. The air data probe status will be correctly shown when
The override utility page provides a mixed bag of tricks, the probes are deployed, and the display can be fully uti-
having to do both with launch abort configuration, deselec- lized to do redundancy management for the air data system
tion of failed navigation components and software overrides or IMUs in case of failures. All software switches on the
for failed switches. lower right work as described.
The left column provides a range of abort selections.
Items 1 to 3 are software overrides for the abort rotary
switch and allow to arm TAL or ATO with items 1 and 2 Systems management functionality
respectively and to call the armed abort with item 3 (note
that RTLS can be called via OPS 601 PRO, so there’s no Systems management is only available in SM OPS 2 (which
need to provide a software switch here). is nominally only loaded in orbit when one of the GPCs run-
Items 4, 50 and 51 change the throttle mapping from ning PASS is taken out of the redundant set running GNC.
rated power to 109% of rated power, i.e. in an emergency If so desired, a GPC can be commanded to load SM during
the engines can be throttled extra this way. Item 5 enables any mission phase on SPEC 0.
and disables the OMS to RCS interconnect for a propellant
dump, item 6 arms the dump, item 7 and 8 start and end it. ANTENNA (OPS 201)
Item 9 allows to manually enter the percentage of OMS fuel
that should be dumped. Dumps for forward and aft RCS can The center part of the antenna management display shows
be controlled with items 13 to 14 and 15 to 16 respectively. a graphical representation of the orbiter fuselage with the
SYSTEMS MANAGEMENT FUNCTIONALITY 129

Ku-band antenna pointing angle as a square. The idea is to

show line of sight — if the square is above the line, then
the antenna is free, otherwise the signal is blocked by the
fuselage.
The left column shows details of the Ku antenna, for in-
stance commanded and actual elevation and azimuth and
TDRS targeting selections. The lower portion shows the
list of all TDRS satellites with their longitudes, the informa-
tion which are currently in view and to which the antenna is
pointing. The right column shows the status of the S-band
antennas and identifies the ground station (if in reach) with
a short ID.

Figure 12.19: The SM OPS 202 page.

software switches in case the hardware switch fails, item 14

activates the software bypass and items 15 to 17 correspond
to the three switch positions.
This display is fully functional in the simulation, both
automatic and manual mode to access the payload bay door
are implemented and all latch gangs can be driven individ-
ually.

ENVIRONMENT (DISP 66)

Figure 12.18: The SM OPS 201 page.
The environment system display summarizes the status of
the cabin atmosphere flow and composition, the various
fans and the supply and waste water system. On the upper
The page is largely functional in-sim and represents what left is the cabin air pressure) (PRESS) and pressure change
the Ku-antenna is doing, whether it’s used for radar rang- rate (dP/DT), and the O2 and CO2 partial pressures (PPO2
ing or searching for a TDRS uplink. Also the S-band an- and PPCO2). Below is the status of the nitrogen and oxygen
tenna shows correct line-of-sight info to a ground station system in terms of flows into the cabin and various pres-
or whether a TDRS connection is made. The usefulness of sures. On the lower left is the status of IMU fans and the
both is somewhat limited in simulated context as there is humidity separator. The upper right shows the avionics bay
no sophisticated ground station to talk to, so whether the fan status and temperatures. Below are quantities, pressures
antenna works or not is unfortunately not overly relevant. and line temperatures of the supply water system, followed
by the waste water system. The lower right shows the status
PL BAY DOORS (OPS 202) of the regenerable CO2 removal system, however this is not
installed on Atlantis.
Opening and closing of the payload bay doors is usually All atmosphere parameters with the exception of airlock
done via hardware switches in the cockpit, but the DPS con- pressure are meaningful in the context of the simulation. Of
tains a software backup page for all the tasks since the func- the water systems, only the quantity readings are modeled.
tion is crucial. Fan indicators and temperature indicators on the other hand
On OPS 202, all switches and talkbacks are available, interface fully with the thermal model.
and in addition the microswitch status can be directly seen.
In particular, item 1 and 2 switch AC power to the door
ELECTRIC (DISP 67)
motors on and off and item 3 toggles the automatic opening
and closing program. Items 4 to 13 allow to individually The electric system display page shows voltages, currents
drive the motors corresponding to the various latch gangs and power usage for various subsystems in the Shuttle. The
in manual mode where the talkbacks show whether a latch left part shows the DC system, the right part AC voltages.
gang or door is open or closed. Items 14 to 17 finally are From top to bottom, the DC systems are: Fuel cell (FC),
130 CHAPTER 12. DPS PAGES

Figure 12.20: The DISP 66 page. Figure 12.22: The DISP 68 page.

CRYO SYSTEM (DISP 68)

main buses (MN), forward and aft power control devices
The cryogenic storage display shows the status of the oxy-
(PCD), and essential buses (ESS).
gen and hydrogen tanks used to power fuel cells and to gen-
erate water for the environment system in the process. The
displayed readings are (top to bottom) pressures (P, TK P),
quantity (QTY), temperature (T), heater temperature (HTR
T), heater status (HTR), manifold pressure (MANF P) and
valve status (VLV).
Since the cryogenic storage system is not explicitly mod-
eled in FG, the display largely shows generic values charac-
teristic of a properly functioning system and quantity read-
ing acording to mission elapsed time.

FUEL CELL (DISP 69)

The fuel cell display contains more detailed diagnostics for
the internal state of the fuel cells. In addition to the volt-
age and current readings also available on DISP 67, there is
also the flow of oxygen and hydrogen to the fuel cells, the
reactant valve status and temperatures and pressures on the
Figure 12.21: The DISP 67 page. left column.The right column shows temperature readings
on relief lines and purge lines above and pH values of the
water below.
Except for the pH readings, all values shown on the dis-
The display is fully functional in FG and reflects any
play are meaningful in the context of the simulation, in par-
loads currently hanging on the corresponding bus, bus ties
ticular temperature readings of the fuel cells or abnormal re-
and the connection status between DC and AC system faith-
actant flows may indicate problems with the stack and low
fully. During nominal operations, the total power consump-
line temperatures are an indication that the line heaters have
tion of the orbiter (without payload) should be about 14 kW.
to be activated, otherwise the lines may freeze.
Only a handful of circuit breakers aboard the Shuttle are
simulated, so in case of a malfunction of equipment, the dis- COMMUNICATIONS (DISP 76)
play will show possibly unplausibly high currents which in
reality would be limited by a breaker. However, the net re- The communications display shows a quick summary of
sult is the same, voltage on a bus will collapse when there’s the function and configuration of the various antenna sys-
a serious electrical problem. tems aboard the Shuttle. The upper left quadrant covers the
SYSTEMS MANAGEMENT FUNCTIONALITY 131

Figure 12.23: The DISP 69 page. Figure 12.25: The DISP 78 page.

S-band antenna, the upper right the Ku-band antenna, the the cabin. The lower left shows a quick summary of the fuel
lower right the space-to-space radio and the lower left the cell system along with total current drawn and total power
network signal processors (NSPs) responsible for signal en- consumption. The right column shows the electric system
coding and decoding. summary with fuel cell and bus voltages, voltages of the
AC buses and at the bottom fuel cell stack temperatures and
pressures.
All systems are simulated and the parameters driven ac-
cording to the status of the respective subsystem.

SYS SUMM 2 (DISP 79)

Figure 12.24: The DISP 76 page.

Apart from the technical details of the NSPs and the vari-
ous self-test indicators, the display is largely functional and
responds to the various ways of configuring the antenna sys-
tem.
Figure 12.26: The DISP 79 page.
SYS SUMM 1 (DISP 78)
The system summary page show a host of information. The The second systems summary shows pressures and tem-
first one is divided into four different areas. The upper left peratures of the cryogenic storages of hydrogen and oxygen
shows the status of the cabin smoke detectors. Below is in the upper part of the display. On the lower left is the sta-
a summary of the cabin air pressure, pressure change and tus of the APUs — exhaust gas temperature (EGT) as well
composition in terms of O2 and CO2 partial pressures, as as lube oil in and out temperatures, speed and fuel quantity.
well as flow readings from nitrogen and oxygen system into At the bottom left is the thermal status of the avionics bays
132 CHAPTER 12. DPS PAGES

and the fans. The right column shows the status of the hy- HYD THERMAL (DISP 87)
draulics system — pressures and temperatures. Below are
The thermal status of the hydraulics system can be ac-
the water supplies for the spray boilers and the valve status,
cessed on DISP 87. On the right column, this shows for
and at the bottom a summary of the H2O and freon loop.
all three hydraulics systems the temperature readings for a
Cryogenic storages are not explicitly simulated, so the large number of sensors distributed across the Shuttle. The
parameters show generic values for a properly functioning left part of the display summarized general characteristics
system. Also, the quantity of hydraulic fluid is not explicitly of the hydraulics system in the upper part, shows the cur-
tracked. All other parameters are meaningful and reflect the rent selection of hydraulics system to device to be operated
simulated state of APUs and hydraulics. and in the lower part shows the currently sensed gear tire
pressures. During orbital operations, the hydraulics system
is not driven by the APUs but fluid is circulated using the
circulation pumps to provide an even temperature distribu-
APU HYD (DISP 86) tion and prevent fluid from freezing up and blocking the
line.
A detailed summary of the APUs and the hydraulics system
can be accessed on DISP 86. The left column shows several
important parameters of the APUs, among them exhaust gas
temperature (EGT), speed, fuel quantity and temperatures
for the gear box (GBX BRG T) and the gas generator bed
(GG BED) which allow to gauge proper function of the tur-
bines. The upper right part summarizes the status of the hy-
draulics system in terms of reservoir temperature (RSVR T)
and pressure (RSVR P) and accumulator pressure (ACCUM
P) which is the desired operating pressure. The lower right
part shows parameters of the water spray boiler cooling sys-
tem in terms of controller status (CNTLR), water reserves
(H2O QTY) and vent temperature (VENT T).

Figure 12.28: The DISP 87 page.

In the simulation, all temperatures and pressures shown

on the display are meaningful and reflect a combination of
the action of the circulation pumps, the heat exchanger from
hydraulic fluid to the hot part of the freon loop and the am-
bient temperature in various parts of the Shuttle. Also the
tire pressures reflect the state of the landing gear and will
indicate flat tires properly.

APU/ENVIRON THERMAL (DISP 88)

The thermal status of the APU and the environment system
Figure 12.27: The DISP 86 page.
is displayed on DISP 88. The display is divided into four
sections. The upper left shows the status of the two freon
loops, with current freon flow and temperature readings at
Nearly all parameters are meaningful in the context of the the radiator inlet, the radiator outlet and the evaporator out-
simulation and reflect proper heat buildup by operating the let to gauge current cooling efficiency.
turbines as well as the interaction of the hydraulics system The water loop is shown in the upper right, with pump
with both the APUs and the hydraulic circulation pumps. pressure, temperature and pressure differential as well as
Since it is only accessible in SM OPS 2 while the APUs are heat exchanger flow. The lower right section shows the
typically used in OPS 1 and 3, it is not quite as useful as it APU fuel supply thermal conditioning, whereas the lower
could be. left shows the evaporator nozzles and feedlines. All these
SYSTEMS MANAGEMENT FUNCTIONALITY 133

Figure 12.29: The DISP 88 page. Figure 12.30: The DISP 89 page.

PDRS CONTROL (SPEC 94)

areas have to be thermally controlled by dedicated heating
elements to avoid freezing of water or hydrazine, and the The payload deployment and retrieval system (PDRS) has
temperature readings reflect the action of the heating ele- several functions that can be accessed via the special func-
ments. tion page 94. The upper part is used to select general con-
Most of the displayed values are meaningful in the sim- figurations, in particular items 7 and 8 enable and disable
ulation and the various feedline temperatures respond to the software stop while items 9 and 10 enable and disable
the action of the heating elements. The freon temperature software auto brake. The section below is used to assign
readings are taken from the overall thermal simulation of automatic RMS movement programs to one of four slots
the Shuttle and based on a fairly sophisticated calculation. (which can then be chosen using the rotary switch) using
The simulation does not include thermal expansion of flu- items 13 - 16, whereas item 17 can be used to select the
ids, hence the values for the freon and water accumulator starting point of a selected auto sequence.
quantities are not based on any detailed simulation. Below are the controls for operator controlled automatic
mode, items 18-20 are used to enter a target position, items
21-23 to enter a target attitude, item 25 to make a reach
limit check given the commands. Finally, on the lower left
software readouts for the RMS arm ready to latch indicators
PRPLT THERMAL (DISP 89) are available.

Thermal properties of the propellant feedlines can be mon-

itored on DISP 89. The display is organized into six ar-
eas, showing the left and right OMS pods to the left, OMS
and RCS crossfeed lines to the upper right, forward RCS in
the center-right and aft RCS on the lower right. Tempera-
tures for multiple sensors located at different places are all
shown. For proper operation, the feedlines should be kept
in the range of 55 to 75 F.
In FG, all temperatures are meaningful (though no de-
tailed simulation of the difference between sensors is done)
and reflect the ambient temperature at the location as well
as the action of the OMS and RCS heating elements. An
exception is the OMS fuel injector temperature (FU INJ).
Since fuel is routed around the engine nozzle for cooling
purposes, the temperature at the injector during engine fir-
ing reflects the efficiency of cooling, and the reading should Figure 12.31: The SPEC 94 page.
be about 220 F for a properly operating engine under thrust.
134 CHAPTER 12. DPS PAGES

Both auto-sequences and operator-controlled automatic

mode are functional. However, RMS inaccuracies such as
requiring correction options like POHS are not simulated,
and the available choice of reference points is rather lim-
ited, so only the options mentioned above are available in
the simulation. The reach limit check works, however is
rather coarse, and there are no automatic sequences pre-
programmed by default which could be chosen.

PDRS OVERRIDE (SPEC 95)

Figure 12.33: The DISP 97 page.

Software utility functionality

The systems software utility pages are used to manage the
software itself. A GPC that has not loaded any application
software starts in OPS 0 which is the software management
function, also accessible as SPEC 0.

Figure 12.32: The SPEC 95 page. GPC MEMORY (SPEC 0)

On this central software management display, the configu-
In case of a switch hardware failure, the PDRS over- ration (which GPC runs PASS GNC, which PASS SM and
ride functionality allows to by-pass every PDRS knob and which BFS) can be edited for each OPS as well as the as-
switch in the cockpit and run the RMS arm purely via soft- signment of data buses to the GPCs (the nominal bus as-
ware commands. Each switch has its own section. The drive signment table (NBAT)). This is done on the left part of the
mode switch is represented by the left column, the joint se- display. The upper right part allows to directly view and
lection switch by the middle column, and the right column edit the software code loaded while the central left part is to
represents the rate switch, the auto proceed switch and the initiate a memory dump. The center right part can be used
input axis. The standard procedure is to always select the to directly change the memory loadout of a GPC without
override option before the items for the switch positions be- having to change OPS first.
come active. In addition, for the drive mode switch, the Item 1 allows to select which memory config is to be
enter option has to be selected. edited (for instance G3 is guidance and navigation OPS 3),
Nearly the whole page is implemented in the FG simula- items 2-6 allow to assign a GPC to the config selected via
tion, and selecting the override option for a switch disables item 1. Items 7 to 19 allow to assign data buses to the vari-
the simulated in-cockpit switches and dials. ous GPCs. Changes made in this area take effect when the
next OPS transition is done.
Item 46 allows to select a GPC to directly change mem-
PL RETENTION (DISP 97) ory config, item 45 allows to select a target configuration,
The status of the payload retention system (i.e. whether a and item 47 loads the target configuration into the selected
payload is latched, released or ready to latch) can be moni- GPC. This allows for instance to directly assign one GPC to
tored on DISP 97. For each of the three payload slots, there run SM OPS 2 during launch if this is needed for whatever
is a row of indicators for all five latching systems. reason.
The simulation only includes the option to carry a single Since the simulation does not use the original flight soft-
payload, hence only the first row of the display will be ac- ware, only the NBAT table, the memory configuration table
tive and show the status of the payload retention system for and the direct change of a memory configuration are simu-
the selected payload. lated.
SOFTWARE UTILITY FUNCTIONALITY 135

the timers to include the chosen offsets while item 33 resets

all offsets to zero.
The lower part of the display deals with time synchro-
nization between GPCs and master timing units (MTU).
The simulation includes the upper parts of the display, i.e.
allows to set and run timers and alarms and provide offsets
to the chosen times, but does not include a simulation of the
MTUs or any hardware errors they might experience.

GPC/BUS STATUS (SPEC 6)

Figure 12.34: The SPEC 0 page.

TIME (SPEC 2)
The time utility page is only available in OPS 2. The upper
part allows to set the CRT timer (running in the upper right
corner of every DPS display) and select what time the CRT
time string displays. Item 1 is used to display GMT, item 2
to display MET (note that times for planned maneuvers in
the software are usually MET). Items 9-11 are used to set
the CRT timer as hours, minutes and seconds, items 12 and
13 start and stop the timer. Item 14 arranges the timer to
Figure 12.36: The DISP 6 page.
start at the specified time and count to zero, item 17 starts
at zero and counts to the specified time,
The GPC status display complements the memory man-
agement utility SPEC 0 in that it shows a summary of the
current GPC memory configuration and the current bus as-
signment table. For each of the five GPCs, it shows the
current mode and loaded OPS software and below which
data buses are assigned to which GPC and which IDP (for
historical reasons called CRT on this display) is talking to
which GPC.
In the simulation, the assignment shown reflects what-
ever table is entered via SPEC 0 and whatever mode the
GPCs are in. However, since failures of the data buses are
not modeled, changing the bus assignments to GPCs won’t
affect the simulation in any meaningful way.

FAULT (DISP 99)

Figure 12.35: The SPEC 2 page.
The fault summary page displays a list of the last 15 soft-
ware CWS fault messages in chronological order with the
The center part of the display allows to add offsets to latest message on top. The message formating follows the
GMT and MET (and so in principle to make the timer dis- standard conventions also used in the DPS fault line. Usu-
play every possible time zone). items 24 to 27 are used to ally the page is accessed via the fault summary button on
enter GMT offsets, items 28-31 to enter MET offsets (each the keyboard, it can alternatively be accessed via SPEC 99
in days, hours, minutes and seconds), and item 32 updates PRO, however this removes all messages from the list.
136 CHAPTER 12. DPS PAGES

Figure 12.37: The DISP 99 page.

All software CWS messages detected by the simulation

(for the complete list see chapter 9) are automatically in-
serted into DISP 99 for future reference.
Chapter 13

Mission files

Mission files are the way to load certain pre-sets for the RMS arm or just a weight parameter that affects launch
simulation to avoid having to configure everything in-sim. and entry performance. Valid options are true or false.
In reality, when the astronauts enter the orbiter, the avionics
is already programmed with launch guidance information, payload: If the payload is explicit, the string identifying
abort targets, Shuttle liftoff weight and similar things — the it (if the previous parameter was false, then the string
crew does not have to do this. can be anything). Valid options are TDRS demo and
In a similar way, mission files allow to load the simula- SPARTAN-201.
tion ready for a particular flight. They include both hard-
ware configuration (what kind of tank is flown, what pay- payload-weight-lbs: If the payload was declared as not ex-
load is aboard, . . . ) and software configuration (orbital DAP plicit, its weight is specified by this parameters. Ex-
gains and deadbands, automatic abort target selections,. . . ) plicitly simulated payloads have their weight declared
By default, the simulation utilizes the file along with the rest of their parameters. Valid entries
SpaceShuttle/Mission/mission.xml, however are positive numbers, however note that there is no
by using the commandline you can point FG to a different sanity check done and values may be entered here that
mission file as are too heavy to be lifted by the Shuttle.
--prop:/mission/filename=myfile.xml
Technically the mission file is a standard xml property oms-kit: An OMS kit module carrying extra propellant can
list (see Fig. 13.1). be flown in addition to a payload. The OMS kit is mod-
The list is divided into several sections, each of which ular and allows to carry extra propellant for up to 1500
can be defined or not. If a section is defined, all of its pa- fps velocity change. Valid entries are none, OMS kit
rameters need to be defined, if a section is not defined, the 500 fps, OMS kit 1000 fps and OMS kit 1500 fps.
content does not matter.
The meaning of the various parameters is given below.

Launch site
Configuration
Some missions require a specific launch site to be mean-
The configuration section details hardware and payload. ingful, in particular the site may be different from KSC. If
that is the case, the site can be declared here and a warning
section-defined: Whether the simulation default configu- message will alert the user who starts at the wrong site.
ration should be altered at all. Valid options are true
or false. section-defined: Whether a particular launch site is re-
quired.
external-tank: What kind of external tank is to be used.
Historically, the tank weight has changed quite a bit
launch-site-iata: The IATA code for the required site,
due to weight reduction programs and usage of differ-
telling any user at the wrong site where he should start
ent materials, so this will affect launch performance.
instead.
Valid options are standard weight, lightweight and
super lightweight.
launch-site-lat: The latitude of the required launch site.
payload-explicit: Whether the payload is an actual 3d
model with a position that can be picked up by the launch-site-lon: The longitude of the required launch site.

137
138 CHAPTER 13. MISSION FILES

<?xml version="1.0" ?>

<PropertyList>

<configuration>
<section-defined type="bool">false</section-defined>
<external-tank type="string">standard weight</external-tank>
<payload-explicit type="bool">false</payload-explicit>
<payload type="string">TDRS demo</payload>
<payload-weight-lbs>5000.0</payload-weight-lbs>
</configuration>

<launch>
<section-defined type="bool">true</section-defined>
<target-inclination type="double">32.5</target-inclination>
<target-apoapsis-miles type="double">90.0</target-apoapsis-miles>
<select-north type="bool">true</select-north>
<tal-site-index type="int">7</tal-site-index>
<rtls-site-index type="int">1</rtls-site-index>
<roll-to-heads-up type="bool">false</roll-to-heads-up>
</launch>

(...)

<orbital-targets>
<section-defined type="bool">true</section-defined>

<object-label type="string">ISS</object-label>
<inclination-deg type="double">42.76</inclination-deg>
<alt-km type="double">280.0</alt-km>
<node-lon-deg type="double">5.07</node-lon-deg>
<anomaly-deg type="double">0.9</anomaly-deg>
</orbital-targets>

</PropertyList>

Figure 13.1: Structure of the mission file.

Launch reached. The periapsis is always given by the MECO

altitude which is fixed in the guidance system, thus
The launch section contains information for the guidance there will always be an OMS burn for orbital insertion
system, such as target orbit or default abort sites. required. As with the inclination, the targeted apoapsis
may not be reachable for all payloads and site latitudes,
section-defined: Whether the simulation default configu-
and there may be an OMS-1 burn required after MECO
ration should be altered at all. Valid options are true
to raise apoapsis in addition to the insertion burn. The
or false.
valid range is hence a function of launch performance,
target-inclination: The targeted orbital inclination in de- targeted inclination and launch site latitude.
grees. Note that a launch may only reach inclinations
select-north: A given inclination can be reached launch-
higher than the launch site latitude and that for very
ing both northward or southward (this leads to an 180
high inclinations and heavy payloads the Shuttle may
degree difference in the longitude of the ascending
not be able to reach orbit at all. The valid range thus
node). Typically safety considerations that launches
depends on launch performance and launch site lati-
should not go over land apply, i.e. from Vandenberg
tude.
one would only ever launch on the southern branch,
target-apoapsis-miles: The targeted apoapsis of the or- from KSC on the northern. Valid options are true or
bit in miles — MECO will occur when this value is false.
COUNTDOWN 139

tal-site-index: In case of a transatlantic abort landing shift the trajectory upwards, shallower angles will
(TAL) is performed, guidance automatically targets a build inertial speed faster.
new landing site. This option sets the index corre-
sponding to the choice (see Appendix B for a table). trajectory-loft-ft: An offset for late second stage guid-
ance for the altitude targeted above or below a nom-
rtls-site-index: In case of a return to launch site (RTLS) inal launch at which the Shuttle accelerates to orbital
abort, guidance steers the Shuttle back to the runway speed.
at the launch site. However, in FG guidance does not
know the launch site (you can launch from any geo- gamma-meco: The flight path angle to be targeted at
graphical position in the simulation), hence you need MECO.
to tell the index of the site to the avionics, or RTLS
guidance won’t work. Valid options are any site index, inclination-targeting: Debug option, leave at a value of 1.
but the option makes only sense if you launch from a
site for which guidance is implemented.
Countdown
roll-to-heads-up: While during early launches the Shuttle
has been kept in heads-down attitude all the way to The countdown section allows to program an automated ig-
orbit, in later missions this was changed. This flag de- nition at a certain time. This makes sense for instance to
termines whether the launch AP will roll the Shuttle precisely target a launch window for an orbital target (de-
shortly after passing Mach 14 (since the thrust vector fined below). The parameters are:
is not aligned with the roll axis, this is a moderately
awkward maneuver and might affect accuracy of or- section-defined: Whether an automatic countdown is used
bital insertion). Valid options are true or false. at all.

throttle-down-time-s: The time after main engine ignition ignition-time-s: The time in seconds after simulation start
at which the engines are throttled down avoid aerody- when the main engine ignition is scheduled. Basically
namic overstress of the vehicle. the Shuttle will rest on the pad for the specified time.
throttle-up-time-s: The time after main engine ignition at automatic-ignition: Whether the engines are automati-
which the engines are throttled back up when the dense cally fired, by-passing the Shift + i command,
atmosphere is passed. or whether just countdown messages will appear but
throttle-down-to-percent: The throttle setting down to no automatic ignition is performed and the simulation
which the main engines are throttled during passage waits for the key command. Valid options are true or
through the densest atmosphere. Usually this is set to false.
67%, but for heavy payloads this may be a larger value.

throttle-nominal-percent: The nominal full throttle set- Ascent trajectory

ting — the engines can be throttled higher than 100%
for launch for better performance, especially when Two section called ascent-traj-stage1 and ascent-traj-
heavy payloads are flown. The valid range is from 95 stage2 allow to define the curve that this visually shown on
to 106 %. the DPS ASCENT TRAJ page. This is needed since several
parameters in the Launch section of the mission file (the
oms-assist-burn: Whether an OMS assist burn during the target inclination, the angular biases for SRB and ballistic
second stage is done. Valid options are true or false. climb or the trajectory loft) change the trajectory from the
default canned profile — which can give the (wrong) im-
oms-assist-duration-s: The duration of the OMS assist
pression that the Shuttle is off-course. Note that the param-
burn in seconds.
eters in these two sections only modify the visuals displayed
srb-climbout-ang-bias-deg: An offset for first stage guid- on the DPS, they have no effect on the autopilot.
ance to the angle targeted for climb-out after passage The following parameters are supported:
through the dense atmosphere. Steeper climbing an-
gles will lead to a higher ballistic arctop, shallower an- section-defined: Whether the trajectory is customized or
gles to a higher inertial speed at SRB separation. not (in this case default values are used). Valid options
are true or false.
ballistic-climb-ang-bias-deg: An offset for second stage
guidance to the angle targeted for climb-out after SRB num-points: The number of points used to specify the cus-
separation up to ballistic arctop. Steeper angles will tom trajectory.
140 CHAPTER 13. MISSION FILES

Each point has to have an index and two coordinates Post-MECO

(see Fig. 13.2 for an example of the basic xml structure).
The number of points declared has to match the parameter A real Shuttle mission is often done with a high degree of
above. automation. For instance, and OMS-1 burn is usually car-
ried out automatically, just as the MPS fuel dump. In this
x: During first stage the Earth-relative velocity in ft/s, for section, parameters for an automated post-MECO program
second stage the inertial velocity in ft/s. can be defined.

y: The altitude in ft at the point when the velocity has section-defined: Whether the simulation default configu-
reached the value specified in the x-coordinate. ration should be altered at all. Valid options are true
or false.

Abort automatic-fuel-dump: Whether the MPS fuel dump

should be initiated automatically. Valid options are
The abort section contains information relevant for the abort true or false.
guidance, insofar as this is not the landing site. Please note
that a meaningful selection requires both a lot of knowl- auto-oms1-burn: Whether the OMS-1 maneuver should
edge about the technicalities of launch aborts and a pre- be initiated automatically. If set to true, the burn pa-
computation of the launch performance during the mission rameters will be entered, the Shuttle will engage au-
using some kind of simulation tool — it’s almost impossible tomatic orbital DAP post-MECO and maneuver into
to ’just guess’ reasonable values. burn attitude, the crew has to depress EXEC to actu-
ally initiate the burn. Valid options are true and false.
section-defined: Whether the simulation default configu-
ration should be altered at all. Valid options are true auto-oms1-burn-peg4: Whether the OMS 1 burn is pro-
or false. grammed in PEG-4 parameters or not (PEG-7 is cho-
sen otherwise). Valid options are true and false.
enable-yaw-steering: During a two engine out contin-
geny abort, the remaining engine can be used to steer oms1-dvx: If the OMS-1 burn is set to true, the prograde
the flight path closer to an emergency landing site. velocity change to be imparted by the burn. This value
This parameter decides whether contingency guidance has to be pre-computed and matched to the apoapsis
should attempt to do that or not. Valid options are true target of the launch guidance.
or false.
oms1-dvy: If the OMS-1 burn is set to true, the normal
yaw-steering-target: If yaw steering is enabled, guidance velocity change to be imparted by the burn. Normally
needs to know into which direction to steer. This is the this would be zero.
angle with the flightpath which should be used (this de-
pends on launch azimuth and where emergency sites oms1-dvz: If the OMS-1 burn is set to true, the radial ve-
are available). Any course is technically valid, but locity change to be imparted by the burn. Normally
nominally only 45 or -45 degrees are chosen. this would be zero.

ato-v-mssn-cntn: This defines the inertial speed (in ft/s) oms1-theta-T: The PEG-4 transfer angle in degrees (only
beyond which an abort to orbit (ATO) will try to steer used if a PEG-4 burn is requested).
into the planned inclination rather than to give up the
mission and steer the most efficient course. What oms1-H: The PEG-4 altitude target in miles (only used if a
is useful depends on payload, target inclination and PEG-4 burn is requested).
launch latitude. Any number will be accepted regard-
less of whether it makes sense or not. oms1-tig-s: The ignition time for the OMS 1 burn in sec-
onds after liftoff.
ato-v-lin: During ATO, an OMS fuel dump is executed.
This parameter defines the inertial speed beyond which orbiter-weight: For non-perfect navigation, the OMS burn
not all of the OMS fuel is dumped but only a fraction. guidance needs the weight of the orbiter after MECO.
This value should be pre-computed and entered here.
ato-v-zero: Complementing the above parameter, this de- Note that using a wrong weight (for instance because
fines the inertial speed beyond which no fuel dump payload was defined but not considered in the compu-
will be considered necessary during an ATO. tation) will lead to the burn missing the burn targets.
DAP 141

DAP dap-A-VRN-rot-pls: Pulse strength for DAP-A Vernier

thrusters — note that the Vernier thrusters can not
All parameters of the orbital DAPs can be adjusted via achieve large rates.
SPEC 20, but they can also be i-loaded to custom values
prior to the mission. This is done in the DAP section of the dap-B-VRN-rot-pls: As above but for DAP-B.
mission file.
dap-A-PRI-comp: Placeholder, currently not used.
section-defined: Whether the simulation default configu-
dap-B-PRI-comp: Placeholder, currently not used.
ration should be altered at all. Valid options are true
or false. dap-A-VRN-comp: Placeholder, currently not used.
dap-A-PRI-rot-rate: Target rotation rate for DAP-A with dap-B-VRN-comp: Placeholder, currently not used.
primary thrusters in degrees per second. This rate is
commanded for full RHC deflection. Any positive dap-A-PRI-p-opt: What RCS pods to be used for the pitch
value will be accepted, but the DAP will be very waste- channel — 0: both 1: aft 2: forward
ful above 5 degrees per second.
dap-B-PRI-p-opt: As above but for DAP-B.
dap-B-PRI-rot-rate: As above, but for DAP-B.
dap-A-PRI-y-opt: What RCS pods to be used for the yaw
dap-A-VRN-rot-rate: As above, but for DAP-A Vernier channel — 0: both 1: aft 2: forward
thrusters. Note that the Vernier thrusters are not very
powerful, so the value needs to be a fraction of a de- dap-B-PRI-y-opt: As above but for DAP-B.
gree, otherwise the time it takes for the Verniers to dap-A-PRI-tran-pls: Velocity change in ft/s for a single
reach the commanded rate will make the DAP useless. translational pulse when firing DAP-A in pulsed mode.
dap-B-VRN-rot-rate: As above, but for DAP-B Vernier dap-A-PRI-tran-pls: As above but for DAP-B.
thrusters.
dap-A-ALT-n-jets: Placeholder, currently not used.
dap-A-PRI-att-db: Attitude deadband (i.e. the max. devi-
ation from desired attitude for which no thruster fire is dap-B-ALT-n-jets: Placeholder, currently not used.
commanded) in degrees for primary thrusters of DAP-
A.
Entry
dap-B-PRI-att-db: As above but for DAP-B.
The entry section is very simple and just contains the
dap-A-VRN-att-db: Attitude deadband for DAP-A
planned landing site for the mission.
Vernier thrusters.
dap-B-VRN-att-db: As above but for DAP-B. section-defined: Whether the simulation default configu-
ration should be altered at all. Valid options are true
dap-A-PRI-rate-db: Rate deadband (i.e. the max. devi- or false.
ation from desired rotation rate for which no thruster
fire is commanded) in degrees per second for primary landing-site-index: The planned landing site (can be
thrusters of DAP-A. changed on SPEC 50) Valid options are any site index
(see Appendix B), also emergency landing sites will
dap-B-PRI-rate-db: As above but for DAP-B. be considered.
dap-A-VRN-rate-db: Rate deadband for DAP-A Vernier
thrusters. RMS auto sequences
dap-B-VRN-rate-db: As above but for DAP-B.
The RMS arm of the Shuttle can be pre-programmed with
dap-A-ALT-rate-db: Placeholder, currently not used. automatic motion sequences which can then be recalled in
space. The mission file allows to i-load six of these se-
dap-B-ALT-rate-db: Placeholder, currently not used. quences. Each sequence consists of a number of points
dap-A-PRI-rot-pls: Rotation rate to be gained by a single which each specify position and attitude of the end effec-
thruster firing in pulsed mode in degrees per second for tor in addition to a delay that is waited when the arm has
DAP-A primary thrusters. reached the designated point. The basic structure is illus-
trated by Fig. 13.2.
dap-B-PRI-rot-pls: As above but for DAP-B. The detailed meaning of the parameters is:
142 CHAPTER 13. MISSION FILES

<rms-auto-sequences>
<section-defined type="bool">true</section-defined>
<num-sequences type="int">1</num-sequences>
<sequence n="0">
<num-points type="int">4</num-points>
<point n="0">
<x type="double">12.0</x>
<y type="double">2.0</y>
<z type="double">0.0</z>
<pitch type="double">0.0</pitch>
<yaw type="double">0.0</yaw>
<roll type="double">0.0</roll>
<delay type="double">0.0</delay>
</point>
<point n="1">
(...)
</point>
</sequence>
<sequence n="1">
(...)
</sequence>
</num-sequences
</rms-auto-sequences>

Figure 13.2: Basic structure of RMS auto sequence definitions

section-defined: Whether sequence are i-loaded at all. Note from the example how both sequences and points
Valid options are true or false. need to be indexed - if this is not properly done, later entries
will simply over-write earlier ones.
num-sequences: The number of sequences declared in the
following. Valid options are 1-6, but the number of
sequence must match what is declared later. Failures
num-points: The number of points declared in the follow- The failure section is somewhat special in that it neither de-
ing. Any positive integer is valid, but the number of fines hardware nor i-loads software functions. Instead, it is
point must match what is declared later. designed to set up a predictable failure scenario for train-
x: The x-axis translational target of the end effector (must ing purposes. For instance, suppose you want to simulate
lie within reach limit). a TAL abort scenario. This requires an engine out during a
particular time interval. Of course you can always switch
y: The y-axis translational target of the end effector (must the engine off with the cutoff switch, but the failure mode
lie within reach limit). definition allows to implement this automatically. The basic
structure of a failure mode definition is shown in Fig. 13.3.
z: The y-axis translational target of the end effector (must Each failure mode has its own xml block which needs to
lie within reach limit). be indexed. Inside each block, the following parameters can
be defined:
pitch: The pitch target of the end effector in degrees (be-
tween -90 and 90)
node: This has to be the property path to a valid
yaw: The yaw target of the end effector in degrees (be- component condition in the simulation. Nearly
tween 0 and 360) all component failure definition is tracked in
/fdm/jsbsim/systems/failures and all com-
roll: The roll target of the end effector in degrees (between ponents have a value between 0 (failed) and 1 (fully
-180 and 180) functional) associated where other values are used to
declare partial failures.
delay: The delay in seconds the end effector waits after
reaching the designated position before proceeding to occurs-met-s: The mission elapsed time (MET) in second
the next point in the sequence. at which the failure occurs. When the MET counter
ORBITAL TARGETS 143

<failures>
<section-defined type="bool">true</section-defined>

<mode n="0">
<node type="string">/fdm/jsbsim/systems/failures/mps/ssme1-condition</node>
<occurs-met-s type="double">135.0</occurs-met-s>
<probability type="double">0.9</probability>
<value type="double">0.0</value>
</mode>
</failures>

</PropertyList>

Figure 13.3: Basic structure of a failure mode definition

has reached the specified value, a check will be done node-lon-deg: Longitude of the ascending node of the or-
whether the failure occurs. bital target in degrees.

probability: A value between 0 and 1 representing the anomaly-deg: True anomaly of the orbiting object upon
probability that the failure mode happens. initialization.

value: The value to which the property node specified is Note that the other orbital elements are not relevant for a
set if the probability check succeeds. circular orbit. The orbital target is initialized at the specified
true anomaly and moves from there according to an analyt-
In the above example, a complete failure of SSME 1 will ical approximation to the J3 orbital dynamics expected for
occur with a probability of 90% 135 seconds into the flight. the nearly circular orbit.
Note that the system neither checks whether a component
has already failed before for other reasons, nor whether the
value specified for the node makes any sense. Notepad
Note that scenarios are an alternative way to specify fail-
ures. The notepad section allows to write text on the in-sim
notepad that can tell about the mission. Up to eight lines
are possible.
Orbital targets section-defined: Whether any text should be written at all.
The orbital target section instructs Flightgear to simulate line1: The text for the first line.
other orbiting objects (stable circular orbits only). These
objects currently exist as coordinate abstractions only, i.e. line1: The text for the second line.
Shuttle guidance can compute distances or rendezvous pa-
rameters, but they can not actually reached (there won’t be a
3d model at the position). The orbital target can be observed
on the groundtrack map.
The parameters specifying an orbital target are:

section-defined: Whether the simulation default configu-

ration should be altered at all. Valid options are true
or false.

object-label: The label that will be displayed on the

groundtrack map.

inclination-deg: Orbital inclination in degrees of the tar-

get.

alt-km: Altitude of the orbit in kilometers above mean sea

level.
144 CHAPTER 13. MISSION FILES
Appendix A

Glossary of acronyms

ADI: attitude director indicator PDRS: payload deployment and retrieval system
ADTA: air data transducer assembly PEG: powered explicit guidance
AoA: angle of attack PFD: primary flight display
AOA: abort once around RCS: reaction control system
AP: autopilot RHC: rotational hand controller
APU: auxiliary power unit RMS: remote manipulator system
ATO: abort to orbit RR: radar ranging
BDA: Bermuda (abort landing) RTLS: return to launch site
BFS: backup flight system SCA: Shuttle carrier aircraft
CAM: computer annunciation matrix SERC: single engine roll control
COAS: crew optical alignment sight SM: systems management
CoG: center of gravity SRB: solid rocket booster
CRT: cathode ray tube TACAN: tactical air navigation
DAP: digital autopilot TAEM: terminal area energy management
DPS: data processing system TAL: transatlantic abort landing
ECAL: East coast abort landing THC: translational hand controller
EI: entry interface TPS: thermal protection system
ET: external tank TVC: thrust vector control
FES: flash evaporator system
GNC: guidance, navigation and control
GPC: general-purpose computer
GPS: global positioning system
HSI: horizontal situation indicator
HST: Hubble Space Telescope
IDP: integrated display processor
IMU: inertial measurement unit
ISS: International Space Station
KSC: Kennedy Space Center
LH2: Liquid hydrogen
LO2: Liquid oxygen
MCC: Mission Control Center
MDU: multi-function display unit
MECO: main engine cutoff
MEDS: multi-function electronic display system
MET: mission elapsed time
MLS: microwave landing system
MPS: main propulsion system
NBAT: nominal bus assignment table
OMS: orbital maneuvering system
OPS: operational sequence
PASS: primary avionics software system

145
146 APPENDIX A. GLOSSARY OF ACRONYMS
Appendix B

Index values of DPS functions

Landing sites and their indices

The sites are selectable on SPEC 50 as item 40 (TAL site, only available during launch) and item 41 (landing site).

Index site status

1 Kennedy Space Center regular
2 Vandenberg Air Force Base regular
3 Edwards Air Force Base regular
4 White Sands Space Harbor regular, lakebed landing
5 Zaragoza Airbase TAL
6 RAF Fairford TAL
7 Banjul International Airport TAL
8 Moron Airbase TAL
9 Le Tube TAL
10 (Ben Guerir) (TAL, does not exist in FG scenery)
11 Bermuda Intl. Airport Bermuda abort site
12 Halifax east coast abort site
13 Wilmington east coast abort site
14 Atlantic City east coast abort site
15 Myrtle Beach east coast abort site
16 Gander east coast abort site
17 Pease east coast abort site
30 Easter Island TAL for Vandenberg launches, never used in reality
32 Diego Garcia emergency landing site
33 Honolulu Intl. Airport emergency landing site
34 Keflavik Airbase emergency landing site
35 Anderson Airbase Guam emergency landing site
36 Amilcar Cabral Intl. Airport emergency landing site
37 Ascension Island emergency landing site
38 Wake Island emergency landing site
39 Lajes Air Base emergency landing site

The following runways are equiped with MLS:

147
148 APPENDIX B. INDEX VALUES OF DPS FUNCTIONS

channel 6 channel 8
Kennedy Space Center KSC33 KSC15
Vandenberg AFB VBG30 VBG12
White Sands NOR32
Edwards AFB EDW06 EDW24
Zaragoza ZZA30
Moron Airbase MRN20
Le Tube FMI22

Stars for COAS

The star ID is entered in SPEC 22 as item 21 to define a reference for a COAS mark or can be used as valid tracking
targets on the OPS 201 page. The following stars are implemented for COAS:

ID star
11 Shedir
12 Mirphak
13 Mizar
14 Arcturus
15 Betelgeuse
16 Procyon
17 Spica
18 Antares
19 Canopus
20 Sirius

GPC memory configurations

The configuration ID is entered on SPEC 0 to determine what software a computer should run. The valid selections are

ID abbreviation description
1 G1 GNC OPS 1 and 6
2 G2 GNC OPS 2
3 G3 GNC OPS 3
4 S2 SM OPS 2
Appendix C

Key bindings

Most functions of the Shuttle can be executed by 3d cockpit switches, GUI options or DPS software switches but for a
few (especially time-critical ones) the following key bindings are used:

Ctrl + q cutoff switch center engine

Ctrl + w cutoff switch left engine
Ctrl + e cutoff switch right engine
Shift + c arm and deploy drag chute
c jettison drag chute
d separate external tank
Ctrl + d force ET separation in unsafe conditions
f trim body flap up
Shift + f trim body flap down
Ctrl + b engage BFS
Ctrl + g arm landing gear
Shift + g deploy landing gear
Shift + h declutter HUD
Shift + i main engine ignition / initiate SCA pitchdown
m toggle controls from RHC to THC
Ctrl + m force the DAP into a different major mode
Ctrl + a mark a COAS attitude reference
Ctrl + u undock from the Space Station
Ctrl + v open view manager
Shift + e initiate / end spacewalk from EVA view
[ yaw left or translate forward
] yaw right or translate backward

149