F
F
T
C
AIR FORCE FLIGHT TEST CENTER
EDWARDS AIR FORCE BASE, CALIFORNIA
AIR FORCE MATERIEL COMMAND
UNITED STATES AIR FORCE
AFFTCTIH9901
WAYNE M. OLSON
Aircraft Performance Engineer
TECHNICAL INFORMATION HANDBOOK
SEPTEMBER 2000
AIRCRAFT PERFORMANCE
FLIGHT TESTING
Approved for public release; distribution is unlimited.
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 07040188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (07040188), 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 222024302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to
comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DDMMYYYY)
September 2000
2. REPORT TYPE
Final
3. DATES COVERED (From  To)
N/A
5a. CONTRACT NUMBER
CRDA #99171FT0
5b. GRANT NUMBER
4. TITLE AND SUBTITLE
Aircraft Performance Flight Testing
5c. PROGRAM ELEMENT NUMBER
5d. PROJECT NUMBER
5e. TASK NUMBER
6. AUTHOR(S)
Olson, Wayne M., Aircraft Performance Engineer
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
Air Force Flight Test Center
412 TW/TSFT
195 E. Popson Avenue
Edwards AFB, California 935246841
8. PERFORMING ORGANIZATION REPORT
NUMBER
AFFTCTIH9901
10. SPONSOR/MONITOR’S ACRONYM(S)
AFFTC
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION / AVAILABILITY STATEMENT
Approved for public release; distribution unlimited.
13. SUPPLEMENTARY NOTES
14. ABSTRACT
This document is intended as a reference source on the topic of aircraft performance flight testing.
Formulas are derived for equations of motion, altitude and airspeed. It covers the various performance
maneuvers, including takeoff, landing, cruise, acceleration, climb, and turn. Specialized tests to calibrate air
data systems and to dynamically determine aircraft lift and drag are discussed. Lift, drag, thrust, and fuel flow
analysis methods are presented. Special topics include gravity models, aerial refueling, terrain following, and
effects of temperature and wind. The text is primarily for conventional jet aircraft, however, many of the
equations and methods are applicable to light civil aircraft.
15. SUBJECT TERMS
aircraft performance models simulation air data takeoff landing cruise performance
acceleration drag turning flight GPS INS lift climb performance
thrust fuel flow jet aircraft calibration atmospheric effects
16. SECURITY CLASSIFICATION OF:
17. LIMITATION
OF ABSTRACT
18. NUMBER
OF PAGES
19a. NAME OF
RESPONSIBLE PERSON
a. REPORT
UNCLASSIFIED
b. ABSTRACT
UNCLASSIFIED
c. THIS PAGE
UNCLASSIFIED
UNLIMITED
284
19b. TELEPHONE NUMBER
(include area code)
Standard Form 298
(Rev. 898)
Prescribed by ANSI Std. Z39.18
iii
PREFACE
The author was employed at the Air Force Flight Test Center (AFFTC), Edwards AFB,
California, from 1968 through 1993 as an aircraft performance flight test engineer. This
document began, but was not finished, prior to his retirement in 1993. He endeavored to
complete the document on his own and this text is the final result of that. He received a lot of
help from the reviewers, which he mentions belowthey each made suggestions that
improved the text vastly.
The intent of this text is that it should provide a highly useful reference source for aircraft
performance flight test engineers. It certainly should not be the only source of information.
The bibliography contains just a few of the sources that the author has found most useful.
Much of the material covered in this handbook can be found in slightly different forms in the
bibliographies listed in the Bibliography section. Even though the Flight Test Engineering
Handbook (listed in the Bibliography Section) was originally written in the 1950s and
updated slightly in the 1960s, it still contains much useful information. The author utilized
Everett Dunlap’s Theory of the Measurement and Standardization of InFlight Performance
of Aircraft extensively as a reference source during his years at Edwards AFB. Also, the
USAF Test Pilot School’s (TPS) Aircraft Performance manual was a valuable source, as well
as the knowledge the author gained while a student at the USAF TPS.
The emphasis here is on performance testing as conducted at Edwards AFB; therefore,
low budget or light aircraft testing is not covered extensively. Very little is said about
instrumentation, except that it is needed and should be as accurate as reasonably possible.
The thrust discussion is kept to a minimum. A number of other possible topics are discussed
lightly, if not at all. Items not necessarily complete are:
1. airspeed calibration in ground effect,
2. test planning,
3. test conduct,
4. how to fly the maneuvers,
5. use of parameter identification,
6. report writing, and
7. cg accelerometer system.
This handbook is pieced together from writing the author has done going back as far as
1975. Much of it is from individual performance office memos which were written to
standalone; therefore, you will see quite a bit of duplication. The same equation appears in
several placesthe author tried to have the major derivation of the equation appear only
once. For those of you who are familiar with the author’s style, you know he is big on theory
and equations. Although it appears that there are a lot of intermediate steps in the derivations,
the extra steps are appropriate to show where all the constants come from.
iv
Early versions of this text had three primary reviewers: Messrs. Mac McElroy, Ron Hart,
and Frank Brown. Mr. McElroy looked at some early versions of this handbook. Messrs. Hart
and Brown reviewed both the draft and final versions of this handbook. Mr. Bill Fish
suggested adding the discussion of the ratio method of standardization and reviewed the
thrust section. Mr. Allan Webb also reviewed the thrust section. Mr. Alan Lawless of the
National TPS and Mr. John Hicks from NASA, Dryden Flight Research Center, provided
significant comments that were implemented into the text. In addition, Mr. Richard Colgren
of LockheedMartin Skunk Works and Captain Timothy Jorris of the AFFTC provided
excellent suggestions that were incorporated.
There were many individual engineers at Edwards AFB that the author would like to
acknowledge in this handbook. Although the list is long, they deserve mentioning. They are:
1. Mr. Jim Pape (who never found out the author did not know the difference between
an aileron and an elevator when he first started working at Edwards AFB).
2. Mr. Willie Allen for teaching the author almost everything he knows about dynamic
performance and flight path accelerometers. Mr. Allen invented the “cloverleaf” airspeed
calibration method, which is discussed in this handbook.
3. Mr. Milton Porter for teaching the author the mathematics that he applied to the
cloverleaf method in a mathematics class at the USAF TPS.
4. Mr. Randy Simpson of the Naval Air Test Center (now called the Naval Air Weapons
Center). The author worked several months with Mr. Simpson on developing dynamic
performance methods in the early 1970s.
5. Mr. Dave Richardson, while reviewing a very early version of this text, pointed out
that the AFFTC and NASA were using dynamic performance methods on the lifting body
research projects years before those of us in the conventional aircraft business.
6. Mr. Jim Olhausen of General Dynamics on the YF16 and F16A, who in the middle
1970s taught the author about using inertial navigation systems (INSs) for performance. As a
result of Mr. Olhausen’s work, the INS became the primary source of flight path acceleration
data on almost every large project at the AFFTC.
7. Mr. Al DeAnda for teaching the author about calibrating airspeed.
8. Mr. Bill Fish for tutoring the author in propulsion (though propulsion is discussed
lightly in this handbook).
9. Mr. Bob Lee  The author worked with Mr. Lee for a short period of time in the early
1970s studying parameter identification.
10. Messrs. Clen Hendrickson, Lyle Schofield, Jim Cooper, Ken Rawlings, Mac
McElroy, Ron Hart, Charlie Johnson, Pete Adolph, Don Johnson, Frank Brown and many
others for helping the author learn about test techniques and other aspects of flight test.
Finally, the author would like to give sincere thanks to Mr. Frank Brown, his successor at
Edwards AFB, for all his help in the preparation of this handbook. In addition, Ms. Virginia
v
O’Brien of Computer Sciences Corporation for the technical editing and final format of this
handbook.
This will not be the final version of this handbook. The AFFTC would appreciate any
suggestions for additional material, clarification of existing material, or any technical errors you
may find. A form to submit proposed changes and/or improvements is included in the back of
this handbook, or if needed, contact either Frank Brown or the author via email with any
comments. Following are addresses and email for each of them.
Frank Brown
412 TW/TSFT
195 E. Popson Ave
Edwards, AFB, CA 935246841
Frank.Brown@edwards.af.mil
Wayne Olson
3003 NE 3
rd
Ave, #222
Camas, WA 986072340
Wayneoperf@home.com
vi
This page intentionally left blank.
vii
TABLE OF CONTENTS
Page No.
PREFACE...................................................................................................................... iii
LIST OF ILLUSTRATIONS........................................................................................... xii
LIST OF TABLES.......................................................................................................... xvii
1.0 OVERVIEW............................................................................................................. 1
1.1 Introduction.......................................................................................................... 1
1.2 Primary Instrumentation Parameters...................................................................... 1
1.3 Ground Tests ....................................................................................................... 2
1.4 Flight Maneuvers.................................................................................................. 3
1.5 Data Analysis ...................................................................................................... 3
2.0 AXIS SYSTEMS AND EQUATIONS OF MOTION.................................................. 5
2.1 Flight Path Axis.................................................................................................... 5
2.2 Body Axis............................................................................................................ 7
2.3 True AOA and Sideslip Definitions........................................................................ 8
2.4 InFlight Forces.................................................................................................... 10
SECTION 2.0 REFERENCE........................................................................................... 12
3.0 ALTITUDE............................................................................................................... 13
3.1 Introduction – Altitude .......................................................................................... 13
3.2 Hydrostatic Equation ............................................................................................ 13
3.3 Geopotential Altitude ............................................................................................ 15
3.4 1976 U.S. Standard Atmosphere ........................................................................... 16
3.5 Temperature and Pressure Ratio ........................................................................... 16
3.6 Pressure Altitude.................................................................................................. 18
3.6.1 Case 1: Constant Temperature .....................................................................18
3.6.2 Case 2: Linearly Varying Temperature..........................................................19
3.7 Geopotential Altitude (H) versus Geometric Altitude (h) ......................................... 23
3.8 Geopotential versus Pressure Altitude  Nonstandard Day....................................... 24
3.9 Effect of Wind Gradient........................................................................................ 25
3.10 Density Altitude.................................................................................................. 26
3.11 Pressure Altitude Error Due to Ambient Pressure Measurement Error................... 28
4.0 AIRSPEED............................................................................................................... 30
4.1 Introduction – Airspeed........................................................................................ 30
4.2 Speed of Sound.................................................................................................... 30
4.3 History of the Measurement of the Speed of Sound................................................ 31
4.4 The Nautical Mile................................................................................................. 32
4.5 True Airspeed...................................................................................................... 32
4.6 Mach Number...................................................................................................... 32
4.7 Total and Ambient Temperature............................................................................ 35
4.8 Calibrated Airspeed.............................................................................................. 35
4.9 Equivalent Airspeed.............................................................................................. 37
4.10 Mach Number from True Airspeed and Total Temperature................................... 37
4.11 Airspeed Error Due to Error in Total Pressure...................................................... 38
5.0 LIFT AND DRAG..................................................................................................... 40
5.1 Introduction.......................................................................................................... 40
viii
TABLE OF CONTENTS (Continued)
Page No.
5.2 Definition of Lift and Drag Coefficient Relationships .............................................. 40
5.3 The Drag Polar and Lift Curve.............................................................................. 41
5.4 Reynolds Number................................................................................................. 42
5.5 Skin Friction Drag Relationships ............................................................................ 43
5.6 Idealized Drag Due to Lift Theories....................................................................... 44
5.7 Air Force Flight Test Center Drag Model Formulation ............................................ 45
5.8 The Terminology ‘Drag Polar’ .............................................................................. 45
SECTION 5.0 REFERENCES......................................................................................... 48
6.0 THRUST................................................................................................................... 49
6.1 Introduction.......................................................................................................... 49
6.2 The Thrust Equation............................................................................................. 50
6.3 InFlight Thrust Deck........................................................................................... 51
6.4 Status Deck......................................................................................................... 51
6.5 Inlet Recovery Factor ........................................................................................... 51
6.6 Thrust Runs ......................................................................................................... 53
6.7 Thrust Dynamics.................................................................................................. 54
6.8 Propeller Thrust ................................................................................................... 54
6.8.1 The Reciprocating Engine at Altitude ............................................................ 55
7.0 FLIGHT PATH ACCELERATIONS......................................................................... 57
7.1 AirspeedAltitude Method..................................................................................... 57
7.2 GPS Method........................................................................................................ 58
7.3 Accelerometer Methods ....................................................................................... 58
7.4 Flight Path Accelerometer Method........................................................................ 58
7.5 Accelerometer Noise............................................................................................ 60
7.6 Inertial Measurement Method ............................................................................... 66
7.7 Calculating Alpha, Beta and True Airspeed............................................................ 66
7.8 Flight Path Accelerations ...................................................................................... 71
7.9 Accelerometer Rate Corrections ........................................................................... 72
7.10 Velocity Rate Corrections ................................................................................... 73
7.11 Calculating p, q, and r.......................................................................................... 73
7.12 Euler Angle Diagram.......................................................................................... 73
8.0 TAKEOFF ................................................................................................................ 75
8.1 General................................................................................................................ 75
8.2 Takeoff Parameters ............................................................................................. 75
8.3 Developing a Takeoff Simulation........................................................................... 78
8.4 Ground Effect...................................................................................................... 80
8.5 Effect of Runway Slope........................................................................................ 87
8.6 Effect of Wind on Takeoff Distance...................................................................... 88
8.7 Takeoff Using Vectored Thrust............................................................................. 88
8.8 Effect of Thrust Component.................................................................................. 92
8.9 EngineInoperative Takeoff................................................................................... 98
8.10 Idle Thrust Decelerations .................................................................................... 102
ix
TABLE OF CONTENTS (Continued)
Page No.
9.0 LANDING................................................................................................................ 103
9.1 Braking Performance ........................................................................................... 103
9.2 Aerobraking......................................................................................................... 106
9.3 Landing Air Phase................................................................................................ 107
9.4 Landing on an Aircraft Carrier .............................................................................. 109
9.5 Stopping Distance Comparison.............................................................................. 112
9.6 Takeoff and Landing Measurement ....................................................................... 113
10.0 AIR DATA SYSTEM CALIBRATION................................................................... 115
10.1 Historical Perspective ......................................................................................... 115
10.2 Groundspeed Course Method .............................................................................. 115
10.3 General Concepts ............................................................................................... 116
10.4 Pacer Aircraft.................................................................................................... 119
10.5 Tower Flyby....................................................................................................... 119
10.6 AccelDecel....................................................................................................... 121
10.7 The Cloverleaf Method  Introduction .................................................................. 124
10.8 The Flight Maneuver........................................................................................... 125
10.9 Error Analysis .................................................................................................... 126
10.10 Air Force Flight Test Center Data Set................................................................ 126
10.11 Mathematics of the Cloverleaf Method............................................................... 132
11.0 CRUISE.................................................................................................................. 135
11.1 Introduction........................................................................................................ 135
11.2 Cruise Tests....................................................................................................... 136
11.3 Range................................................................................................................ 136
11.4 Computing Range from Range Factor .................................................................. 139
11.5 Constant Altitude Method of Cruise Testing ......................................................... 141
11.6 Range Mission.................................................................................................... 141
11.7 Slow AccelDecel .............................................................................................. 142
11.8 Effect of Wind on Range .................................................................................... 142
12.0 ACCELERATION AND CLIMB............................................................................. 144
12.1 Acceleration....................................................................................................... 144
12.2 Climb................................................................................................................. 145
12.3 Sawtooth Climbs................................................................................................. 146
12.4 Continuous Climbs .............................................................................................. 148
12.5 Climb Parameters............................................................................................... 149
12.6 Acceleration Factor (AF).................................................................................... 149
12.6.1 Two Numerical Examples for AF...............................................................150
12.7 Normal Load Factor During A Climb................................................................... 152
12.8 Descent ............................................................................................................. 154
12.9 Deceleration....................................................................................................... 154
SECTION 12.0 REFERENCES ....................................................................................... 154
13.0 TURNING............................................................................................................... 155
13.1 Introduction........................................................................................................ 155
13.2 Accelerating or Decelerating Turns ..................................................................... 155
x
TABLE OF CONTENTS (Continued)
Page No.
13.3 ThrustLimited Turns .......................................................................................... 155
13.4 Stabilized Turns .................................................................................................. 156
13.5 LiftLimited Turns .............................................................................................. 156
13.6 Turn Equations ................................................................................................... 157
13.6.1 Normal Load Factor................................................................................... 157
13.6.2 Turn Radius............................................................................................... 159
13.7 Turn Rate .......................................................................................................... 159
13.8 Winds Aloft........................................................................................................ 160
14.0 DYNAMIC PERFORMANCE................................................................................. 164
14.1 Introduction........................................................................................................ 164
14.2 Roller Coaster .................................................................................................... 164
14.3 Windup Turn...................................................................................................... 167
14.4 SplitS................................................................................................................ 167
14.5 Pullup ................................................................................................................ 170
14.6 Angle of Attack.................................................................................................. 172
14.7 Vertical Wind..................................................................................................... 172
15.0 SPECIAL PERFORMANCE TOPICS ..................................................................... 173
15.1 Effect of Gravity on Performance........................................................................ 173
15.2 Performance Degradation during Aerial Refueling................................................ 176
15.3 Performance Degradation during Terrain Following .............................................. 177
15.4 Uncertainty in Performance Measurements.......................................................... 178
15.5 Sample Uncertainty Analysis............................................................................... 178
15.6 Wind Direction Definition.................................................................................... 179
16.0 STANDARDIZATION............................................................................................ 180
16.1 Introduction........................................................................................................ 180
16.2 Increment Method.............................................................................................. 180
16.2.1 Climb/Descent........................................................................................... 181
16.2.2 Acceleration/Deceleration.......................................................................... 181
16.2.3 Accelerating/Decelerating Turn.................................................................. 182
16.2.4 Cruise....................................................................................................... 182
16.2.5 ThrustLimited Turn................................................................................... 182
16.3 Ratio Method ..................................................................................................... 182
17.0 A SAMPLE PERFORMANCE MODEL.................................................................. 184
17.1 Introduction........................................................................................................ 184
17.2 Drag Model........................................................................................................ 184
17.2.1 Minimum Drag Coefficient ......................................................................... 184
17.3 Skin Friction Drag Coefficient ............................................................................. 188
17.4 Drag Due to Lift................................................................................................. 189
17.5 Thrust and Fuel Flow Model................................................................................ 193
17.6 Thrust Specific Fuel Consumption........................................................................ 193
17.7 Military Thrust.................................................................................................... 195
17.8 Maximum Thrust ................................................................................................ 197
17.9 Cruise................................................................................................................ 198
17.10 Range .............................................................................................................. 200
17.11 Endurance........................................................................................................ 203
xi
TABLE OF CONTENTS (Concluded)
Page No.
17.12 Acceleration Performance................................................................................. 203
17.13 Military Thrust Acceleration.............................................................................. 204
17.14 Maximum Thrust Acceleration........................................................................... 207
17.15 Sustained Turn.................................................................................................. 210
18.0 CRUISE FUEL FLOW MODELING........................................................................ 213
18.1 Thrust Specific Fuel Consumption........................................................................ 215
18.2 Multiple Regression ............................................................................................ 216
SECTION 18.0 REFERENCE ......................................................................................... 219
19.0 EQUATIONS AND CONSTANTS.......................................................................... 220
19.1 Equations ........................................................................................................... 220
19.2 Constants........................................................................................................... 229
APPENDIX A  AVERAGE WINDS AND TEMPERATURES FOR THE
AIR FORCE FLIGHT TEST CENTER.................................................... 231
APPENDIX B  WEATHER TIME HISTORIES............................................................. 237
APPENDIX C  AVERAGE SURFACE WEATHER FOR THE
AIR FORCE FLIGHT TEST CENTER.................................................... 241
BIBLIOGRAPHY........................................................................................................... 245
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS ...................................... 249
INDEX........................................................................................................................... 261
AIRCRAFT PERFORMANCE FLIGHT TESTING CHANGE FORM
xii
LIST OF ILLUSTRATIONS
Figure No. Title Page No.
2.1 Aircraft Axis System...................................................................... 7
2.2 Angle of Attack and Sideslip Definitions........................................... 8
2.3 InFlight Forces ..............................................................................10
2.4 Axis System Angle Diagram............................................................11
3.1 Element of Air ................................................................................14
3.2 Logarithmic Variation of Pressure Ratio...........................................22
3.3 Standard Atmosphere Temperature..................................................23
4.1 True Airspeed versus Calibrated Airspeed........................................36
4.2 True Airspeed Error for 0.001 in. Hg Error.......................................38
5.1 Ratio of Compressible to Incompressible Dynamic Pressure..............41
5.2 Skin Friction Drag Relationships.......................................................44
5.3 Drag Polar .....................................................................................46
5.4 LifttoDrag Ratio versus Lift Coefficient.........................................47
6.1 Turbine Engine Schematic ...............................................................49
6.2 Normal Shock Recovery Factor .......................................................52
6.3 F15 Inlet Schematic .......................................................................53
6.4 Thrust Dynamics from an Air Force Flight Test Center Thrust Stand .54
7.1 Air Force Flight Test Center Nose Boom Instrumentation Unit ..........60
7.2 Longitudinal Load Factor – Unfiltered Data......................................61
7.3 Normal Load Factor – Unfiltered Data.............................................62
7.4 FourPole Butterworth Filter Attenuation Characteristics...................63
7.5 FourPole Butterworth Filter Group Time Delay................................64
7.6 Longitudinal Load Factor – Filtered Data..........................................65
7.7 ThirdOrder Polynomial Fit of Filtered
Longitudinal Load Factor Data.........................................................65
7.8 Euler Angles...................................................................................74
8.1 Takeoff and Landing Forces and Angles ..........................................76
8.2 Predicted Ground Effect Drag.........................................................80
8.3 Lift Ratio InGround Effect .............................................................83
8.4 Takeoff Forces...............................................................................86
8.5 Takeoff Parameters........................................................................86
xiii
LIST OF ILLUSTRATIONS (Continued)
Figure No. Title Page No.
8.6 Effect of Wind................................................................................ 88
8.7 F16 Dimensions............................................................................. 89
8.8 Distance to LiftOff ........................................................................ 91
8.9 Angle of Attack at LiftOff ............................................................. 91
8.10 Effect of Thrust Component on LiftOff Speed................................. 92
8.11 Effect of Thrust Component on Distance to LiftOff ......................... 93
8.12 Delta Tail Lift for Tail Area = 60 ft
2
................................................. 94
8.13 Delta Tail Lift for Tail Area = 80 ft
2
................................................. 95
8.14 Distance to LiftOff versus Airspeed............................................... 96
8.15 Calibrated Airspeed at LiftOff........................................................ 96
8.16 Takeoff Lift Model......................................................................... 97
8.17 Takeoff Drag Model....................................................................... 98
8.18 Takeoff Parameters versus Time..................................................... 99
8.19 Takeoff Forces versus Airspeed......................................................100
8.20 Takeoff Forces versus Airspeed: Engine Inoperative........................101
9.1 Braking Forces ...............................................................................103
9.2 Stopping Distance versus Mu ( µ ) ...................................................104
9.3 Deceleration versus Calibrated Airspeed..........................................104
9.4 Mu versus Groundspeed (Wet Runway) ...........................................105
9.5 Braking Forces versus Calibrated Airspeed......................................106
9.6 Total Resistance Force Comparison.................................................107
9.7 Final Descent Rate versus Initial Descent Rate.................................108
9.8 Landing Air Phase..........................................................................109
9.9 F/A18 with Tailhook Extended........................................................110
9.10 The U.S.S. Nimitz...........................................................................110
10.1 Groundspeed Course – Heading Method ..........................................115
10.2 Groundspeed Method – Direction Method ........................................116
10.3 Flyby Tower Grid............................................................................120
10.4 Altitude versus Grid Reading for Flyby Tower ..................................120
xiv
LIST OF ILLUSTRATIONS (Continued)
Figure No. Title Page No.
10.5 Effect of 10Foot Error in Flyby Tower Altitude................................121
10.6 Pressure Survey.............................................................................123
10.7 AccelDecel Delta H......................................................................123
10.8 AccelDecel Position Error Coefficient ............................................124
10.9 Cloverleaf Flight Maneuver .............................................................126
10.10 Air Force Flight Test Center F15 Pacer ..........................................126
10.11 Position Error .................................................................................129
10.12 Groundspeed – Run 1a....................................................................130
10.13 Groundspeed – Run 1b....................................................................130
10.14 Groundspeed – Run 1c....................................................................131
10.15 True Airspeed................................................................................131
12.1 Specific Excess Power from Acceleration........................................145
12.2 AC119G Aircraft...........................................................................147
12.3 AC119G Sawtooth Climb Data.......................................................147
12.4 AC119G Excess Thrust Data .........................................................148
12.5 Acceleration Factor – Constant Calibrated Airspeed.........................150
12.6 Acceleration Factor – Constant Mach Number.................................152
12.7 Centripetal Acceleration Diagram....................................................153
13.1 Normal Load Factor Vectors In a Turn............................................157
13.2 Banked Turn Diagram.....................................................................158
14.1 Drag Model....................................................................................165
14.2 Roller Coaster Normal Load Factor .................................................166
14.3 Roller Coaster Altitude Time History................................................166
14.4 Roller Coaster Mach Number Time History......................................167
14.5 SplitS Drag Model.........................................................................169
14.6 SplitS Normal Load Factor .............................................................169
14.7 SplitS Mach Number Time History.................................................170
14.8 SplitS Altitude Time History...........................................................170
14.9 Pullup Mach Number Time History..................................................171
xv
LIST OF ILLUSTRATIONS (Continued)
Figure No. Title Page No.
14.10 Pullup Altitude Time History............................................................171
17.1 Subsonic Drag Increment ................................................................185
17.2 Transonic Drag Increment...............................................................185
17.3 Supersonic Drag Increment .............................................................186
17.4 Summary of Delta Drag Coefficient.................................................188
17.5 Skin Friction Drag Coefficient..........................................................188
17.6 Drag Due to Lift Slope....................................................................190
17.7 Drag Model at 0.8 Mach Number ....................................................191
17.8 Subsonic Drag Model......................................................................192
17.9 Drag Model – All Mach Numbers....................................................192
17.10 Thrust Specific Fuel Consumption....................................................194
17.11 Military Referred Net Thrust ...........................................................196
17.12 Military Thrust................................................................................196
17.13 Referred Net Thrust for Maximum Thrust........................................197
17.14 Maximum Thrust ............................................................................198
17.15 Range Factor..................................................................................200
17.16 Maximum Range Factor ..................................................................201
17.17 Range Factor – Altitude Effect........................................................201
17.18 Range Factor – Variation with Temperature.....................................202
17.19 Fuel Flow  Endurance....................................................................203
17.20 Military Thrust Specific Excess Power.............................................205
17.21 Military Thrust – Specific Excess Power, Temperature Effect...........205
17.22 Military Thrust – Thrust and Drag at 10,000 Feet..............................206
17.23 Drag at 10,000 Feet – Temperature Variation...................................207
17.24 Maximum Thrust Specific Excess Power .........................................208
17.25 Maximum Thrust Specific Excess Power Temperature Effect at
30,000 Feet.....................................................................................208
17.26 Acceleration Time – Variation with Thrust .......................................210
17.27 Maximum Thrust – Sustained Turn Normal Load Factor ...................211
xvi
LIST OF ILLUSTRATIONS (Concluded)
Figure No. Title Page No.
18.1 C17A Aircraft...............................................................................213
18.2 Thrust Specific Fuel Consumption....................................................215
18.3 Percentage Error in Thrust Specific Fuel Consumption......................218
18.4 Range Factor Variation with Altitude ...............................................219
A1 Delta Temperature at 10,000 Feet....................................................233
A2 Delta Temperature at 20,000 Feet....................................................233
A3 Delta Temperature at 30,000 Feet....................................................234
A4 Delta Temperature at 40,000 Feet....................................................234
A5 Delta Temperature at 50,000 Feet....................................................235
A6 Wind Direction ...............................................................................235
A7 Windspeed.....................................................................................236
A8 Geometric Height minus Pressure Altitude........................................236
B1 Delta Temperature Time History.....................................................239
B2 Wind Direction Time History...........................................................240
B3 Windspeed Time History.................................................................240
C1 Average Maximum and Minimum Surface Temperatures ..................243
xvii
LIST OF TABLES
Table No. Title Page No.
3.1 1976 U.S. Standard Atmosphere...................................................... 17
3.2 Standard Atmosphere Pressure and Temperature ............................. 17
3.3 Edwards Average Weather Data for January................................... 25
3.4 Energy Altitude Effect of Wind Gradient .......................................... 26
3.5 Pressure Error Versus Altitude Error ............................................... 29
5.1 Reynolds Number Variation with Mach Number and Altitude ............ 42
7.1 Summary of Statistics for Longitudinal Load Factor........................... 66
8.1 Takeoff Events............................................................................... 87
8.2 Effect of Runway Slope .................................................................. 87
8.3 Forces at LiftOff Speed................................................................. 97
8.4 Takeoff Parameters at Flight Events ................................................100
8.5 Takeoff Parameters at Significant EventsEngineInoperative............101
9.1 Ground Effect Parameters for F/A18 Carrier Landing......................111
9.2 Change in True Airspeed During Landing Due to Ground Effect........112
9.3 Dry, Wet, and Aerobraking Data Summary......................................113
9.4 Integration of Braking Results..........................................................113
10.1 Aircraft Average Measurements and Parameters .............................128
10.2 Inertial Speeds (GPS)......................................................................128
10.3 Outputs ..........................................................................................129
11.1 B52G Cruise Data .........................................................................136
11.2 Range Factor Versus Altitude for B52G..........................................140
12.1 Climb Ceiling Definitions .................................................................146
14.1 Pullup and SplitS Initial and End Conditions......................................172
15.1 Effect of Latitude on Gravity at Sea Level........................................174
15.2 Effect of Altitude on Gravity............................................................175
15.3 Effect of Heading and Speed on Normal Load Factor .......................175
15.4 Effect of Heading on Drag Coefficient.............................................176
15.5 Parameter Uncertainties..................................................................178
17.1 Tabulated Drag Rise Data...............................................................187
17.2 Range Factor Variation with Altitude ...............................................202
17.3 Range Factor Variation with Temperature........................................203
17.4 Drag Variation with Temperature ....................................................207
1
1.0 OVERVIEW
1.1 Introduction
Aircraft performance flight testing is different things to different people. It involves ground
tests such as calibrating instruments, weighing the aircraft, and static thrust runs. Taxi tests are
performed prior to first takeoff. Then, there is the collection of data during all phases of flight.
The phases of flight include takeoff, acceleration to climb speed, climb, acceleration, cruise,
deceleration, descent, and landing. During flight, the aircraft will also maneuver in sustained,
accelerating or decelerating turns. Specialized maneuvers called dynamic maneuvers are used
to efficiently collect aircraft lift and drag data. Aircraft airspeed, altitude, and temperature
measurement systems will be calibrated in flight. All data collected will be reduced to enable
analysis of specific maneuvers such as cruise and to verify and update aircraft mathematical
models for lift, drag, thrust, and fuel flow. Simulation and curve fitting may be utilized during
the data analysis process.
1.2 Primary Instrumentation Parameters
In a performance evaluation, there can be hundreds of instrumentation measurements.
However, only a few can be considered primary. We will make a list as follows:
Total pressure. A measurement of the total pressure (in typical units of pounds per square
foot) experienced by the aircraft. For flight test aircraft, this is often from a nose boom.
Ambient (or static) pressure. An attempt to measure the atmospheric ambient pressure (in
same units as total pressure). This is subject to errors called position errors. The terminology is
due to the fact that there is some ‘position’ on the surface of the aircraft where the ambient
pressure error is zero or minimal. The bad news is that for any given static source location, the
position error varies with speed, altitude, and attitude.
Total temperature. A temperature probe is used to measure the total temperature of the air.
From measured total pressure, ambient pressure and total temperature we can calculate the
true airspeed of the aircraft. True airspeed is the physical speed of the aircraft with respect to
the moving air mass. From total and ambient pressure then we compute the indicated airspeed.
Indicated airspeed is a measure of the differential pressure. Differential pressure is simply total
pressure minus ambient pressure. Since we have position error in the ambient pressure, we will
apply corrections to ambient pressure to be able to go from indicated airspeed to the corrected
values for calibrated and true airspeed.
Aircraft gross weight. This is not a single measurement, but a calculation usually based
upon a set of fuel tank quantity measurements in flight. The fuel tank quantity weights are
simply added to a known empty weight of the aircraft. The empty weight will be computed for
each flight based upon the particular configuration for that flight. The aircraft will also be
weighed at various times during the program to verify the calculations.
Longitudinal flight path acceleration. We will compute the longitudinal acceleration of the
aircraft parallel to the flight path. The flight path is determined by the true airspeed vector. On
most aircraft programs, we use inertial navigation system (INS) data to compute the
2
longitudinal acceleration. The airspeedaltitude method or GPS are also used. By dividing
longitudinal acceleration by the acceleration of gravity, we get the longitudinal load factor.
Then, multiply the longitudinal load factor by the gross weight to obtain the excess thrust. If
there is one fundamental equation of aircraft performance, it would be the following:
Drag = Net Thrust – Excess Thrust
where:
Drag = the net aerodynamic resistance parallel to the velocity vector.
Normal acceleration: The acceleration perpendicular to the flight path is the normal
acceleration. Divide normal acceleration by gravity to obtain normal load factor. Lift is the
net aerodynamic force perpendicular to the velocity vector. If we ignore the small component
of thrust perpendicular to the velocity vector, then we get a second fundamental formula.
However, keep in mind this one is only approximately correct, while the first one is exact.
Lift = (Normal Load Factor) x Weight
Thrust. The propulsive force provided by the engine. In this handbook, we will discuss
only turbine engines. However, most of the equations of motion in this handbook are
applicable to aircraft with other types of propulsion. Thrust is produced during the process of
air accelerating through the engine. The air entering the inlet is nearly brought to a stop and
then accelerated through various turbine stages. The combustion process dramatically increases
the temperature of the air and the air (plus the fuel) exits the tail pipe at a much higher
velocity. This change in momentum and a pressure difference between the inlet and exit are the
primary factors that produce thrust. Thrust is computed from a variety of measured engine and
atmospheric parameters.
1.3 Ground Tests
Instrumentation calibration. The installation and calibration of all aircraft instruments
should occur prior to flight. Much of the instrumentation can be checked after it is installed in
the aircraft. The output of the total and ambient pressure probes can be groundtested using
precision pressure monitors.
Aircraft weight and cg. The aircraft should be weighed with zero fuel and with various
amounts of fuel to check the numbers provided by the contractor. The center of gravity (cg)
can be determined in a weight facility where separate scales are available for the main and
nose gear.
Static thrust. The installed thrust of the engines can be measured directly on the ground on
a static thrust stand. The principle of a thrust stand is quite simple. The aircraft sits on a pad
and is connected by cables to a load cell that measures load (thrust) directly in pounds of force.
By operating the engine at various throttle settings, a comparison of thrust at zero speed over a
range of power settings can be made with predictions.
Taxi tests. While taxiing on the ground, the aircraft is tested. Taxi means simply to move
the aircraft under its own power on the ground without achieving flight. The first taxi tests
3
would be accomplished in the lowest power setting called idle. The idle taxi tests, combined
with the static thrust data, will quantify idle thrust at low speeds. Taxi tests at higher throttle
settings and approaching liftoff speeds will give an early indication of thrust and drag on the
ground. The final test, prior to first takeoff, will be to rotate the aircraft to liftoff attitude.
1.4 Flight Maneuvers
Takeoff tests are performed to determine the distance required to liftoff and to clear an
obstacle. In USAF testing, the obstacle clearance height is 50 feet, while in civilian testing, the
height is 35 feet for heavy aircraft and 50 feet for light aircraft. Liftoff is usually defined as
when lift first becomes greater than weight. For multiengine aircraft, engineout testing is also
performed wherein one engine’s power is reduced to idle to simulate an engine failure during
takeoff.
Climb tests are flown to determine time, distance, and fuel used to climb to a cruise
altitude. In addition, rate of climb versus altitude is determined.
Cruise testing is conducted to evaluate aircraft range. The aircraft is flown in stabilized
flight over a range of speed and altitude conditions in order to determine the best speed and
altitude to achieve maximum range. However, with modern analysis methods, the optimum
range conditions are usually determined through analysis of drag and thrust/fuel flow models,
which are verified and updated using cruise and other data.
Acceleration tests are conducted during level 1g flight at fixed throttle settings. These
tests are used in conjunction with climb tests to determine the optimum climb profiles. They
are also used to update thrust and fuel flow models for fixed throttle settings over a range of
altitudes and ambient temperature conditions. Excess thrust (thrust minus drag) is measured
versus speed at various altitudes.
Turning performance is conducted to both determine ability of the aircraft to turn and to
assist in generating aircraft lift and drag models at higher lift and angleofattack values than
what are obtainable in 1g flight.
Deceleration and descent tests are conducted to determine ability of the aircraft to
decelerate and the fuel used in descent maneuvers. In addition, this data can be used to assist in
generating aircraft thrust/fuel flow and drag models.
Landing tests are used to measure the distance to land starting from clearing an obstacle (as
in the takeoff test). Braking tests performed during the landings or as separate tests, will
evaluate stopping performance as well as the ability of the brakes to withstand the high
temperatures associated with maximum performance braking.
1.5 Data Analysis
Thrust. Engine thrust is evaluated at fixed throttle settings. For military aircraft, these
settings are usually designated IDLE, MIL (military) and MAX (maximum). Idle is the
minimum throttle setting, MIL is the maximum throttle setting without the use of afterburner,
and MAX is the Maximum throttle setting with the use of afterburner. Thrust at these fixed
throttle positions is primarily a function of flight conditions (speed, altitude, and temperature).
4
A secondary function is angle of attack (angle between the aircraft body xaxis and the
airspeed vector). Thrust is not measured directly, but rather computed from flight conditions
and engine parameter measurements. The engine parameters needed usually include pressure,
temperature, and rpm (revolutions per minute). Thrust is then computed using an engine
manufacturerprovided computer program as modified by the airframe contractor to include
installation effects. This is designated an inflight thrust deck. A second computer program is
usually provided÷a prediction deck, which will predict thrust without knowing any engine
parameters (just flight conditions and throttle setting). The flight test data analyst will compare
the inflight thrust deck data to the prediction deck data. Then, analysis will be performed to
attempt to ‘model’ this data.
Fuel flow. Engine fuel flow will be measured, modeled, and plotted versus thrust and as a
function of flight conditions. Fuel flow data will be obtained both during the fixed throttle
maneuvers (climb, accel, and turn) and during cruise testing. Fixed throttle refers to a specified
throttle position like MIL, MAX or IDLE.
Lift. Lift in the form of a nondimensional lift coefficient will be determined and modeled
versus angle of attack and Mach number.
Drag. Drag will be computed from thrust and excess thrust and modeled versus lift in
nondimensional coefficient form.
5
2.0 AXIS SYSTEMS AND EQUATIONS OF MOTION
2.1 Flight Path Axis
The true airspeed vector defines the flight path (or wind) axis. The inertial velocity vector
defines the inertial flight path axis. In this text, when the singular axis is used, we are usually
referring to the longitudinal or x component of the wind axis system. The component of
aerodynamic force parallel to the flight path axis is defined as drag. Lift is the component of
aerodynamic force perpendicular to the drag (or flight path) axis. The component of aircraft
acceleration parallel to the flight path is the longitudinal acceleration (
x
A ). The longitudinal
load factor (
x
N ) is simply the
x
A divided by the acceleration of gravity ( g ). In conventional
aircraft performance, g is assumed a constant at the reference gravity and given the value of
32.174 ft/sec² (foot per second squared). The symbol
0
g will be used to denote the reference
gravity. The effect of assuming a constant g is dealt with in the gravity section.
To derive the equations of motion we could start with the following energy relationship:
E KE PE · + (2.1)
where:
E = total energy (footpounds),
KE = kinetic energy (footpounds), and
PE = potential energy (footpounds).
Then, assuming zero wind:
2
0
0.5
t
t
W
KE V
g
 `
· ⋅ ⋅
. ,
(2.2)
0 t
W m g · ⋅ (2.3)
t
PE W H · ⋅ (2.4)
where:
m = aircraft mass (slugs), [(pounds force)(seconds)
2
/(foot)],
t
W = aircraft gross weight (pounds),
H = geopotential altitude (feet), and
t
V = true airspeed (feet/sec).
Note: It is assumed that tapeline (or geometric) altitude ( h ) and geopotential altitudes ( H )
are identical. The small difference of these two altitude parameters is discussed in the altitude
section.
6
Adding the potential and kinetic energy relationships (2.2) and (2.4) and dividing by
t
W
yields the following:
( )
2
0
/
2
t
t
t t
PE KE
V
E W H
g
W W
]
· + · +
]
⋅
]
(2.5)
The energy per unit weight ( /
t
E W ) is called energy altitude (or energy height) (
E
H ).
( )
2
0
2
t
E
V
H H
g
· +
⋅
(2.6)
Taking the derivative with respect to time (and ignoring wind) yields:
0
/
t t
E
V dV
dH dt dH dt
g dt
]  `  `
· + ⋅
]
. , . ,
]
(2.7)
The derivative of
E
H with respect to time is called specific excess power and given the
symbology of
s
P . The Cambridge Air and Space Dictionary (Reference 2.1) gives the
following definition of specific excess power: “Thrust power available to an aircraft in excess
of that required to fly at a particular constant height and speed, thus being usable for climbing,
accelerating or turning.”
Equation 2.7 then becomes:
( )
0
t
s E t
V
P H H V
g
]  `
· · + ⋅
]
. , ]
! ! !
(2.8)
Dividing by
t
V yields:
( ) ( ) ( )
0 s t E t t t
P V H V H V V g · · +
! ! !
(2.9)
Envision an accelerometer aligned perfectly with the longitudinal flight path axis and
calibrated in units of g. The accelerometer would be sensitive to both aircraft change in
velocity ( /
t
dV dt ) and a component of gravity ( ( ) / /
t
dH dt V ). Equation (2.9) then becomes:
0 x t t
N H V V g · +
! !
(2.10)
In performance analysis, the axis system of interest is the flight path axis and not the body
or earth axis, so the subscript f (f for flight path) is usually deleted on the flight path axis load
factors. That is, we use
x
N rather than
f
x
N or even
w
x
N (subscript w is for wind axis). Other
references may use other symbologies.
7
2.2 Body Axis
The aircraft axis system (Figure 2.1) is called the body axis system. The Xaxis is defined
through the center of the fuselage with positive being forward. The Yaxis is positive out the
right wing and the Zaxis is positive down. The XYZ body axis system is an orthogonal axis
system usually originating at the center of mass of the aircraft.
Figure 2.1 Aircraft Axis System
If the acceleration of the vehicle in the body axis is known, then the flight path acceleration
can be computed by transforming first through the angle of attack and then through the sideslip
angle. The relationships for α and β as a function of the body axis true airspeed components
are as follows:
( )
1
tan
bz bx
V V α
−
· (2.11)
( )
1
sin
by t
V V β
−
· (2.12)
( )
2 2 2
t bx by bz
V V V V · + + (2.13)
where:
bx
V = body axis x component of the true airspeed,
by
V = body axis y component of the true airspeed,
bz
V = body axis z component of the true airspeed, and
t
V = true airspeed.
8
2.3 True AOA and Sideslip Definitions
The following illustration, shows angle of attack ([AOA] or α ) and angle of sideslip
([AOSS] or β ) in relation to the body axis velocities. The following is the equivalent
symbology for Figure 2.2.
a.
cg bx
U V ·
b.
cg by
V V ·
c.
cg bz
W V ·
Note: Positive directions are shown.
Figure 2.2 Angle of Attack and Sideslip Definitions
AOA (α ) is the angle between the Xbody axis and the projection of the true airspeed
vector ( cos
t
V β ⋅ ) on the XZ body axis plane. AOSS ( β ) is the angle between the velocity
vector and the XZ body plane.
In three dimensions, the α transformation matrix from the body axis to the flight path axis
is as follows:
[ ]
cos 0 sin
0 1 0
sin 0 cos
α α
α
α α
]
]
·
]
] −
]
(2.14)
In three dimensions, the β transformation matrix from the body axis to the flight path axis
is as follows:
9
[ ]
cos sin 0
sin cos 0
0 0 1
β β
β β β
]
]
· −
]
]
]
(2.15)
The transformation of the acceleration from the body axis to the flight path axis is as
follows (a subscript f [for flight path] will be dropped for the flight path axis):
cos sin 0 cos 0 sin
sin cos 0 0 1 0
0 0 1 sin 0 cos
x bx
y by
z bz
A A
A A
A A
β β α α
β β
α α
¹ ¹ ] ] ¹ ¹
¹ ¹ ¹ ¹
] ]
· − ⋅ ⋅
' ' ' '
] ]
¹ ¹ ¹ ¹
] ] −
¹ ¹ ] ] ¹ ¹
(2.16)
Multiplying the equation 2.16 for the longitudinal load factor in the flight path axis yields
equation 2.17.
cos cos sin cos sin
x bx by bz
A A A A β α β β α · ⋅ ⋅ + ⋅ + ⋅ ⋅ (2.17)
The vast majority of performance maneuvers produce very low sideslip and lateral
acceleration such that equation 2.17 may be approximated by equation 2.18 assuming zero
sideslip.
cos sin
x bx bz
A A A α α ≅ ⋅ + ⋅ (2.18)
In matrix shorthand, equation 2.16 is as follows:
{ ¦ [ ] [ ]{ ¦
b
A A β α · ⋅ (2.19)
where:
,
,
x y z
A A A
= three components of flight path accelerations, and
, ,
bx by bz
A A A = three components of body axis accelerations.
Usually, analysis is performed using the flight path axis load factors, as shown in equation
2.20 through 2.22, rather than the above flight path accelerations.
0
/
x x
N A g · (2.20)
0
/
y y
N A g · (2.21)
0
/
z z
N A g · − (2.22)
Note the sign change on the Z component.
The topic of axis transformations is dealt with in more detail in the accelerometer section.
There, we will deal with inertial axis (north, east, down), flight path axis, and with rate
10
corrections to accelerations and velocities in the body axis. Transformations are made to the
body axis where the rate corrections are applied.
2.4 InFlight Forces
Figure 2.3 illustrates the X and Z forces acting on an aircraft in flight. Figures 2.3 and 2.4
illustrate the basic forces and angles of a typical aircraft in flight. It is, however, simplified in
that all forces are acting through a single point. This is called the point mass model. Most
conventional aircraft simulations utilize this simplification. A more complex model would
distribute the lift and drag forces between the wing and tail. The tail may be a part of the wing
as in an aircraft like the French Mirage. What we might otherwise call the trailing edge flap of
the wing provides the pitching moment that a tail usually would.
Figure 2.3 InFlight Forces
The flight path axis is defined by the true airspeed (
t
V ) vector.
a. D  drag acting parallel to the flight path;
b. L  lift acting perpendicular to the flight path;
c. α  angle of attack  angle between xbody axis and the flight path axis;
d. γ  flight path angle  angle between horizontal and the flight path;
e. θ  pitch attitude  angle between horizontal and xbody axis (not shown above);
f.
g
F  gross thrust – acting through the engine axis;
g.
e
F  net propulsive drag – acting through the flight path axis; and
h.
t
i  thrust incidence angle (not shown) – angle above the xbody axis through which
the gross thrust acts; often equals zero.
11
Figure 2.4 Axis System Angle Diagram
Summing forces in the longitudinal or Xflight path axis:
( )
0
0
t
x x x x t ex
x
W
F m A N g N W F
g
 `
· ⋅ · ⋅ ⋅ · ⋅ ·
. ,
∑
(2.23)
where:
ex
F = excess thrust.
[ cos( ) ]
ex g t e
F F i F D α · ⋅ + − − (2.24)
Some airframe manufacturers will define α as the angle between the flight path axis and
the wing axis. However, most will define α as the angle between the flight path axis and the
xbody axis, which is the definition used in this handbook.
The true airspeed velocity vector and the inertial (or ground) speed vector will, in general,
be in a different direction and a different magnitude. The vector relationship between true
airspeed and groundspeed is simply airspeed equals groundspeed plus windspeed. However,
this is a three dimensional relationship that we can represent in vector notation as follows:
t g w
V V V · +
" " "
(2.25)
12
where:
true airspeed vector
t
V ·
"
,
ground speed vector
g
V ·
"
, and
wind speed vector
w
V ·
"
.
Wind direction, by meteorological convention, is the direction from which the wind is
blowing. For instance, let’s say you are flying due north, with zero sideslip, at 500 knots.
Heading is the direction the aircraft is pointing. Assume there is a 100 knot wind at 0 degrees.
That would mean the wind is 100 knots blowing from due north. Or in this case, a pure
headwind of 100 knots. If you have a 100knot headwind and a 500knot true airspeed then the
groundspeed is 400 knots. Airspeed equals groundspeed plus wind (plus is italicized to place
emphasis). There is, in the aero community, some controversy as to the sign convention. This
author considers plus to be the ‘correct’ sign. However, if one uses a negative sign and is
consistant with definitions, the results will come out the same.
Summing forces in the normal or Zflight path axis:
( )
0
0
t
z z z z t
W
F m A N g N W
g
 `
· ⋅ · ⋅ ⋅ · ⋅
. ,
∑
(2.26)
sin( )
z t g t
N W L F i α ⋅ · + ⋅ + (2.27)
where:
normal load factor
z
N ·
, and
lift L · .
The propulsive drag (
e
F ) is only in the longitudinal flight path axis so that its contribution
normal to the flight path is zero.
SECTION 2.0 REFERENCE
2.1 Walker, P.M.B., ed. 1995. Cambridge Air and Space Dictionary. Cambridge University Press.
13
3.0 ALTITUDE
3.1 Introduction – Altitude
There are several forms of altitude of interest in aircraft performance. For this text,
generally, all units will be in feet. The first altitude is geometric (or tapeline) altitude ( h ).
Geometric altitude is the physical, linear altitude measured from mean sea level. Mean sea
level is defined (from Britannica¹) as the height of the sea surface averaged over all stages of
the tide over a long period of time. The length of a foot of geometric altitude does not vary as a
function of temperature or gravity variation with altitude. In the early days of flight, the
technology was not available to measure altitude onboard an aircraft. However, they could
measure the outside ambient pressure. A standard atmosphere was defined which allowed the
computation of an altitude that was proportional to the ambient pressure. That altitude is the
pressure altitude, which we will denote with the symbology
C
H , where
c
stands for calibrated.
In order to derive a relationship between pressure and pressure altitude, it became necessary to
define another altitude called geopotential altitude ( H ). The length of geopotential altitude
foot varies with increasing altitude proportional to the change in gravity with altitude. The
gravity model that has been used to define the geopotential altitude is a simplified model based
upon reference gravity at sea level (
0
g = 32.174 ft/sec
2
) and gravity varying with altitude as
per the inverse square gravity relationship.
For the standard atmosphere model,
C
H and H are identical by definition. This requires
that sea level pressure is exactly the standard atmosphere value and that temperature is
precisely standard day at all altitudes (not just at the altitude being considered). As will be
shown later, the difference between h and H at 50,000 feet is less than 200 feet, but this
difference grows in proportion the square of altitude from the center of earth, where the radius
of the earth is over 20 million feet. Finally, an altitude commonly used to compute piston
powered light aircraft performance is density altitude (
d
H ). Density altitude is useful for light
aircraft primarily because engine performance is generally proportional more to density than to
pressure for internal combustion engines. Density altitude is proportional to atmospheric
density, just as pressure altitude is proportional to atmospheric pressure. Density altitude and
pressure altitude is the same on a standard day at the altitude being considered. In this case, it
is not required that temperatures be standard at all altitudes as was the case for H and H
c
being
identical.
3.2 Hydrostatic Equation
We will derive the relationship between atmospheric pressure and altitude. Envision a
cubic element of air with unit horizontal dimensions ( dx and dy ) and a height equal to dh .
The pressure on the bottom of the element is P . The pressure on the top of the element is
P dP + . The equation for static equilibrium of the element of air is as follows (the unit
dimension into the page ( dy ) is not shown in Figure 3.1):
W g dx dy dz ρ · ⋅ ⋅ ⋅ ⋅ ·weight of the element of air (3.1)
14
W
P dP +
P
dh
dx
Figure 3.1 Element of Air
( ) P dP P g dx dy dz P g dh ρ ρ + · − ⋅ ⋅ ⋅ ⋅ · − ⋅ ⋅ (3.2)
Since dx and dy are of unit length, and the height ( dz ) is equal to dh ,
dP g dh ρ · − ⋅ ⋅ (3.3)
where:
P = pressure,
ρ = density,
g = acceleration of gravity,
h = height, and
dh = height increment.
Using the inverse square gravity law:
( )
2
0
0
0
r
g g
r h
]
· ⋅
]
+
]
]
(3.4)
15
where:
0
r = reference radius of the earth (20,855,553 ft),
= 6,356,772 meters,
0
g = reference gravity (32.17405 ft/sec²), and
= 9.80665 m/sec
2
(exactly by international agreement).
Introducing the ideal gas equation of state:
P R T ρ · ⋅ ⋅ (3.5)
Solving for ρ in 3.5:
( )
P
R T
ρ ·
⋅
(3.6)
where:
T = ambient temperature, and
R = gas constant = 3,089.8136 ft²/(sec²°K).
Value for R is converted from metric units using the 1976 U.S. Standard Atmosphere.
Substituting 3.4 and 3.6 into 3.3:
2
0
0
0
r P
dP g dh
R T r h
]
 `
 `
] · − ⋅ ⋅ ⋅
⋅ +
. , ]
. ,
]
(3.7)
( ) ( ) ( )
2
0 0 0
/ 1/ dP P g R T r r h dh ] · − ⋅ ⋅ + ⋅
]
(3.8)
It is not a simple matter to integrate the above equation exactly. The concept of a
geopotential altitude was introduced to allow for the integration.
3.3 Geopotential Altitude
Geopotential altitude is developed from equation 3.9.
0
g dh g dH ⋅ · ⋅ (3.9)
where:
g = gravity at altitude h ,
h = tapeline (or geometric) altitude, and
H = geopotential altitude.
16
A tapeline foot is the same physical length independent of height while a geopotential foot
expands with increasing altitude linearly with the corresponding decrease in gravity.
0
g
dH dh
g
 `
· ⋅
. ,
(3.10)
Substituting 3.10 into 3.3 and using 3.6:
( )
0 0
P
dP g dH g dH
R T
ρ
]
· − ⋅ ⋅ · − ⋅ ⋅
]
⋅
]
(3.11)
( ) ( )
0
/ / dP P g R dH T · − ⋅ (3.12)
The above formula can be integrated if T either is a constant or is linearly varying with
geopotential altitude ( H ). This means you can look up the integration formula in a table of
integrals. A standard atmosphere model has been defined which contains only constant or
linear temperature segments. The first standard atmosphere, defined by the French in 1919,
contained just one segment. The constants in that segment are still the same today (as of 1976).
This standard atmosphere purports to represent an average temperature model of the earth’s
atmosphere throughout the world and during the various seasons.
3.4 1976 U.S. Standard Atmosphere
The 1976 U.S. Standard Atmosphere model is (as of the writing of this handbook) the
accepted temperature and pressure profile model in the United States. The profile is presented
in Tables 3.1 and 3.2. The region up to about 17 kilometers (56,000 feet) is known as the
troposphere. Quoting from Britannica¹ Online: “troposphere  a term derived from the Greek
words tropos, ‘turning’ and sphaira, ‘ball’.” The temperature decreases rapidly with altitude in
this region. The rising warm air meets the sinking cold air and the air tends to “turn over” like
a “ball” – hence the term troposphere. One would pause between layers, hence, the transition
to the next layer is called the tropopause. To about 50 kilometers (164,000 feet), the
temperature rises slowly in a region called the stratosphere. Altitudes higher than 50 kilometers
are above the region of conventional aircraft performance, so we will not discuss those.
However, the temperatures for the model atmosphere are included in Tables 3.1 and 3.2 to a
geometric altitude of 86 kilometers.
3.5 Temperature and Pressure Ratio
We will define temperature ratio (θ ) and pressure ratio (δ ). These are, respectively, the
ratio of ambient temperature to standard temperature at sea level and the ratio of ambient
pressure to standard pressure at sea level. The formulas are as follows:
288.15
SL
T T
T
θ · · (3.13)
17
2116.22
SL
P P
P
δ · · (3.14)
where:
T = units of degrees K, and
P = units of pounds/foot
2
.
Table 3.1
1976 U.S. STANDARD ATMOSPHERE
Geopotential
Height
(m)
Geopotential
Height
(ft)
Temperature
Gradient
(°K/1,000 ft)
Temperature
(°K)
Pressure
(lbs/ft
2
)
0 0 1.9812 288.15 2,116.2166
11,000 36,089 0.0000 216.65 472.6805
20,000 65,617 0.3048 216.65 114.3454
32,000 104,987 0.8534 228.65 18.1289
47,000 154,199 0.0000 270.65 2.31632
51,000 167,323 0.8534 270.65 1.39805
71,000 232,940 0.6096 214.65 0.082632
84,852 278,386 N/A 186.95 0.0077983
Notes: 1. The temperature gradient and base temperature in the first segment of the standard
atmosphere has remained unchanged since the 1925 U.S. Standard Atmosphere.
2. The standard atmosphere is defined in metric units. The exact conversion factor from
meters to feet is to divide meters by 0.3048.
3. The highest altitude in the table is an even 86,000 meters geometric (tapeline) altitude.
Table 3.2
STANDARD ATMOSPHERE PRESSURE AND TEMPERATURE
Geopotential
Altitude ( H )
(ft)
Ambient
Pressure ( P )
(lbs/ft
2
)
Pressure
Ratio (δ )
Ambient
Temperature (T )
(°K)
Temperature
Ratio (θ )
0.00 2116.22 1.0000 288.15 1.0000
5,000 1760.80 0.8320 278.24 0.9656
10,000 1455.33 0.6877 268.34 0.9312
15,000 1194.27 0.5643 258.43 0.8969
20,000 972.49 0.4595 248.53 0.8625
25,000 785.31 0.3711 238.62 0.8281
30,000 628.43 0.2970 228.71 0.7937
35,000 497.95 0.2353 218.81 0.7594
36,089.24 472.68 0.2234 216.65 0.7519
40,000 391.68 0.1851 216.65 0.7519
45,000 308.01 0.1455 216.65 0.7519
50,000 242.21 0.1145 216.65 0.7519
55,000 190.47 0.09001 216.65 0.7519
60,000 149.78 0.07078 216.65 0.7519
65,000 117.78 0.05566 216.65 0.7519
18
Table 3.2 (Concluded)
STANDARD ATMOSPHERE PRESSURE AND TEMPERATURE
Geopotential
Altitude ( H )
(ft)
Ambient
Pressure ( P )
(lbs/ft
2
)
Pressure
Ratio (δ )
Ambient
Temperature (T )
(°K)
Temperature
Ratio (θ )
65,616.8 114.350 0.05403 216.65 0.7519
70,000 92.684 0.04380 217.99 0.7565
75,000 73.054 0.03452 219.51 0.7618
80,000 57.674 0.02725 221.03 0.7671
85,000 45.608 0.02155 222.56 0.7724
90,000 36.123 0.01707 224.08 0.7777
95,000 28.656 0.01354 225.61 0.7820
100,000 22.768 0.01076 227.13 0.7882
The numbers in Tables 3.1 and 3.2 represent the model atmosphere. On any given day,
there will be variation from that model (refer to Appendix A for what the average variation is
for data taken above Edwards AFB).
3.6 Pressure Altitude
3.6.1 Case 1: Constant Temperature
0
T T · (3.15)
Substituting 3.15 into the relationship 3.12:
( ) ( )
0 0
/ dP P g R dH T · − ⋅ (3.16)
We will integrate using a table of integrals and relationships for natural logarithms. Since
0 0
, g R and T are each constant:
( ) ( )
( )
0 0
0 0
0 0
ln( ) ln( )
g g dP
P P dH H H
P R T R T
 `  `
− −
· − · ⋅ · ⋅ −
⋅ ⋅
. , . ,
∫ ∫
(3.17)
Solving for P in 3.17:
( )
( ) { ¦
0
0
0
0
g
H H
R T
P P e
¹ ¹
− ⋅ −
' '
⋅
¹ ¹
· ⋅ (3.18)
Solving for H :
( )
0
0
0 0
ln
R T
P
H H
g P
] ⋅
 `
· − ⋅
]
. ,
]
(3.19)
19
For the segment of the atmosphere from 11,000 meters (36,089 feet) to 20,000 meters
(65,617 feet):
a.
0
T = 216.65 °K (69.7 °F or 56.5 °C),
b.
0
P = 472.68 pounds/ft
2
0
at H H · , and
c.
0
H = 36,089.24 feet (11,000 m).
3.6.2 Case 2: Linearly Varying Temperature
Assume a temperature that varies linearly with altitude as follows:
( )
0 0
T T a H H · + ⋅ − (3.20)
where:
0
T = base temperature,
0
H = base geopotential altitude, and
a = temperature gradient (deg K/foot).
Substituting, again, into the relationship (3.12)
( ) ( )
0
/ / dP P g R dH T · − ⋅ :
[ ] ( )
0
0 0
/
g
dP P dH
R T a H H
¹ ¹
¹ ¹
· − ⋅
' '
]
⋅ + ⋅ −
¹ ¹
] ¹ ¹
(3.21)
Integrating from a table of integrals:
( )
( )
1
ln
dx
a bx
a b x b
· ⋅ +
+ ⋅
∫
Then using the relationship ln( ) ln( ) ln( / ) u v u v − · :
( )
( ) ( )
0 0
0
0
0
ln ln
T a H H
g
P
P R a
T
]
+ −
¹ ¹
 `
· − ⋅ ] ' '
⋅
. ,
¹ ¹ ]
]
(3.22)
Solving for P :
( )
( )
0
0 0
0
1
g
R a
a
P P H H
T
]
−
]
⋅
]
]  `
· ⋅ + ⋅ −
]
. , ]
(3.23)
Or solving for H :
20
( )
0
0
0
0
1
R a
g
T
P
H H
P a
− ⋅
]
 `  `
]
· + − ⋅
. , . , ]
]
(3.24)
For the first segment of the standard atmosphere (zero to 11,000 meters; zero to 36,089.24
feet), substituting constants (from the international standard atmosphere) [for English units]:
( )
0
1.9812/1000
6.8755856 6
288.15
a
E
T
− · · − (round to 6.87559 6 E − ) (3.25)
( )
( )
0
32.17405
5.255876
3089.8136 1.9812/1000
g
R a
− · ·
⋅
] ⋅
]
(round to 5.2559) (3.26)
( )
5.2559
0
1 6.87559 6
P
E H
P
· − − ⋅ (3.27)
Solving for H :
( )
( )
1 5.2559
0
1
6.87559 6
P
P
H
E
]
 `
−
]
. ,
]
·
−
(3.28)
Equation 3.26 is the definition of pressure altitude for altitudes from zero to 36,089 feet
(zero to 11,000 meters).
Using the pressure ratio (δ) as defined in equation 3.14.
SL
P
P
δ · (3.29)
where:
SL
P = standard sea level pressure = 101,325 pascals (exactly, by international agreement).
The unit pascal has been defined as a newton of force per square meter. A newton has units
of (kg m/sec
2
). One newton is equal to 0.2248195 pounds force.
In various English units:
SL
P = 2,116.2166 pounds/ft² (usually rounded to 2,116.22);
≅ 760 mm Hg;
= 1,013.25 millibar (mb); and
= 29.92 in. Hg
21
Substituting 3.29 into 3.28:
( )
( )
1 5.2559
1
6.87559 6
C
H
E
δ
]
−
]
·
−
(3.30)
The above is for zero to 36,089 feet pressure altitude.
The symbol
C
H is used for pressure altitude to distinguish it from the geopotential altitude
( H ). Pressure altitude and geopotential altitudes are only identical for the model atmosphere.
Similarly:
( )
5.2559
1 6.87559 6
C
E H δ · − − ⋅ (3.31)
For the temperature ratio (θ ), using equation 3.20 and substituting constants (from the
international standard atmosphere):
0
1.9812
1 6.87559 6
288.15 288.15 1, 000
T T
H E H θ · · − ⋅ · − − ⋅ (3.32)
The second segment of the standard atmosphere (11,000 to 20,000 meters) (36,089 to
65,617 feet) is a constant temperature (T =56.5 degrees C) segment. The standard atmosphere
is defined in metric units. English units require the conversion factor of 0.3048 meters per foot.
For instance, the 11,000meter point is 36,089.24 feet.
For the altitude segment between 36,089 feet and 65,617 feet:
( )
0 0
32.17405
/( ) 4.806343 5
3089.8136 216.65
g R T E ⋅ · · −
⋅
(3.33)
0
0
20, 805.84
R T
g
⋅  `
·
. ,
Computing δ for 36, 089.24 H · feet using the δ formula for the first segment of the
atmosphere (equation 3.31):
[ ] ( ) { ¦ 4.806343 5 36089.24
0.22336
C
E H
e δ
− − ⋅ −
· ⋅ (3.34)
For the temperature ratio (θ ), using equation 3.20 and substituting constants (from the
international standard atmosphere):
0
1.9812
1 6.87559 6
288.15 288.15 1, 000
T T
H E H θ · · − ⋅ · − − ⋅ (3.35)
22
The equations for any segment of the 1976 U.S. Standard Atmosphere can be derived by
simply applying the above equations since all segments of the standard atmosphere are either
constant temperature or linearly varying temperature versus pressure altitude.
The standard atmosphere pressure ratio versus pressure altitude is nearly a straightline
logarithmic function as can be seen in Figure 3.2.
Log(delta) versus Pressure Altitude [K Feet]
6
5
4
3
2
1
0
0 50 100 150 200 250 300
Pressure Altitude (ft*1,000)
L
o
g
(
d
e
l
t
a
)
:
d
e
l
t
a
=
P
a
/
P
a
s
l
Figure 3.2 Logarithmic Variation of Pressure Ratio
The logarithm in Figure 3.2 is base 10. As can be seen, at each 50K point the atmospheric
pressure decreases by a factor of 1/10th. For instance at 50K the pressure ratio is 0.1145, at
100K it is 0.01076, at 150K it is 0.00010946, etc. As discussed earlier, all the segments of the
standard atmosphere are either constant temperature or linearly varying with altitude. Figure
3.3 illustrates the linear temperature segments.
23
Standard Atmosphere Temperature
0
50
100
150
200
250
300
180 200 220 240 260 280 300
Standard Temperature (deg K)
P
r
e
s
s
u
r
e
A
l
t
i
t
u
d
e
(
f
t
*
1
0
0
0
)
Figure 3.3 Standard Atmosphere Temperature
3.7 Geopotential Altitude (H) versus Geometric Altitude (h)
Using the inverse square gravity law and the definition of H:
( )
2
0
0
0
r
g g
r h
]
· ⋅
]
+
]
]
(3.36)
0
g dh g dH ⋅ · ⋅ (3.37)
Substituting 3.36 into 3.37 and solving for dH :
( )
0
2
0
r
dH dh
r h
]
] · ⋅
+ ]
]
(3.38)
Integrating gives the relationship between H and h (or tapeline). From a table of integrals:
( )
( )
2
1 dx
b a bx
a bx
· −
+
+
∫
In our case,
0
a r · , 1 b · and x h · .
Factoring out the
2
0
r term in the numerator:
24
( )
( )
2 2
0 0 2
0 0 0
0
1 1
h
dh
H r r
r h r
r h
]
· ⋅ · ⋅ − +
]
+
+ ]
]
∫
(3.39)
Multiply the first term in square brackets by
0
0
r
r
and the second term by
( )
( )
0
0
r h
r h
+
+
.
( )
( )
( )
2 0
0
0
0 0 0 0
r h
r
H r
r h r r h r
] +
−
· ⋅ +
]
+ ⋅ + ⋅
]
]
(3.40)
By factoring terms, we get:
( )
0
0
r
H h
r h
]
· ⋅
]
+
]
]
[ ]
0
20,855, 553 feet r · (3.41)
At 50,000 feet tapeline altitude (the upper limit of most conventional aircraft performance
testing), H computes to be 49,881 feet, for a difference of only 119 feet, or 0.24 percent.
3.8 Geopotential versus Pressure Altitude  Nonstandard Day
A standard temperature may exist at a given altitude on a test day but there would never be
a standard atmosphere at all altitudes except in computer models.
Using the basic / dP P relationship (3.16):
0
( / ) ( / )
C STD
dP P g R dH T · − ⋅ standard day (3.42)
0
( / ) ( / ) dP P g R dH T · − ⋅ test day (3.43)
There can be a significant difference between having a standard atmosphere and achieving
standard temperature at a given altitude. The pressure levels at a given pressure altitude are by
definition the same whatever the temperature. Therefore, we could equate the right sides of
equations 3.39 and 3.40.
/ /
C STD
dH T dH T · (3.44)
where:
test day
T T · .
C
STD
T
dH dH
T
 `
· ⋅
. ,
(3.45)
Since dh dH ≅ (i.e., ∆tapeline ≅ ∆geopotential):
25
C
STD
T
dh dH
T
 `
· ⋅
. ,
(3.46)
Or in a climb, for instance:
C
STD
T
h H
T
 `
· ⋅
. ,
! !
= rate of climb (3.47)
Sample calculation:
Assume a climb through 30,000 feet with /
C
dH dt = 1,000 ft/min = rate of change of
pressure altitude. Then, presume a test day temperature that is 10.0 degrees C hotter than
standard day. Standard day temperature at 30,000 feet is 228.7 degrees Kelvin (K).
Inserting these values into 3.45:
( ) 228.7 10.0
1, 000 1, 043.7
228.7
h
] +
· ⋅ ·
]
]
!
(3.48)
The physical rate of climb (the derivative of tapeline altitude) is 4.4 percent higher than the
rate of change of pressure altitude for being 10 degrees C hotter than standard day. Average
temperatures for the Air Force Flight Test Center (AFFTC) at altitudes from 10,000 feet every
10,000 feet to 50,000 feet can be found in Appendix A. As can be seen, it is not uncommon to
be off standard day by 10 degrees C or more.
3.9 Effect of Wind Gradient
Average windspeed and direction data for the AFFTC, as a function of altitude for each
month, can be found in Appendix A. This is average data for a time span of over 30 years. To
illustrate the effect of wind on climb performance we will take data from January at pressure
altitudes of 13,801 feet (600 mb [millibar]) and 23,574 feet (400 mb). Standard sea level
pressure in millibars is 1013.25. We will conduct calculations for a climb speed of 280 knots
calibrated airspeed (
C
V ). This is typical for F16 and large transport aircraft. Table 3.3
contains the average meteorological data and computed variables.
Table 3.3
EDWARDS AVERAGE WEATHER DATA FOR JANUARY
Pressure
Altitude
(ft)
Geometric
Altitude
(ft)
Standard
Temperature
(deg K)
Delta
Temperature
(deg K)
Ambient
Temperature
(deg K)
Windspeed
(kts)
13,801 14,065 260.8 3.2 264.0 28.7
23,574 23,937 241.4 1.0 242.4 43.5
26
Now, we wish to compute the change in energy altitude for climbing directly into the wind
(headwind) and with the wind (tailwind). The inertial energy altitude, as derived in the first
section, is as follows:
( )
2
0
2
g
E
V
H h
g
· +
⋅
(3.49)
Table 3.4 shows the values of groundspeed and energy altitude for a headwind, tailwind,
and zero wind. In each case, the calibrated airspeed is the same at 280 knots.
Table 3.4
ENERGY ALTITUDE EFFECT OF WIND GRADIENT
Altitude
( h )
(ft)
Airspeed
(
t
V )
(kts)
Headwind
(
g
V )
(kts)
Tailwind
(
g
V )
(kts)
No Wind
(
E
H )
(ft)
Headwind
(
E
H )
(ft)
Tailwind
(
E
H )
(ft)
14,065 343.4 314.7 372.1 19,285 18,449 20,194
23,937 396.5 353.0 440.0 30,897 29,453 32,507
Calculating the delta energy altitudes:
a. Zero Wind
E
H ∆ = 30,89719,285 = 11,612 feet,
b. Headwind
E
H ∆
= 29,45318,449 = 11,004 feet, and
c. Tailwind
E
H ∆
= 32,50720,194 = 12,312 feet.
Comparing these numbers, on an average day over Edwards AFB in January, the change in
energy altitude is 1,308 feet greater flying with a tailwind than flying into a headwind. This is
over the geometric altitude range of 14,065 to 23,937 feet. This is 11.9 percent compared to the
headwind number or 6.0 percent compared to zero wind. In making this comparison we have
ignored the flight path angle. The airspeed vector is inclined with respect to the horizontal by
the flight path angle while the winds are in the horizontal plane.
When climb performance is measured using the altimeter (pressure altitude) large errors
could be induced due to wind gradients. This is why opposite heading climb data are obtained
("sawtooth climbs"). The wind gradient effect can now be accounted for using GPS or INS
data.
3.10 Density Altitude
Density altitude is nothing more than an altitude on a test day that produces an equivalent
density on a standard day. The density altitude parameter has been used primarily for
reciprocating engines, whose power output is generally proportional to air density (i.e., density
altitude). Since the reciprocating engine is generally flown at altitudes below 11 km
(kilometer); the pressure and temperature ratio equations for the first segment of the
atmosphere are appropriate. The relations (equations 3.31 and 3.32) were derived above in the
altitude portion of this section.
27
( )
5.2559
1 6.87559 6
C
E H δ · − − ⋅
( ) 1 6.87559 6
C
E H θ · − − ⋅
The first formula (δ ) is valid for standard or any nonstandard day. That is, pressure ratio is
a function of pressure altitude only and vice versa. On the other hand, the temperature ratio
(θ ) formula is valid only for standard temperatures.
We can compute density ratio (σ ) for a standard day, by taking the ratio of the above
formulas.
( )
( )
( )
5.2559
4.2559
1 6.87559 6
1 6.87559 6
1 6.87559 6
C
C
C
E H
E H
E H
δ
σ
θ
− − ⋅
· · · − − ⋅
− − ⋅
(3.50)
The above σ formula is valid only for standard day. However, one could define the
density altitude (
d
H ) as being directly proportional to density as defined by equation 3.50.
( )
4.2559
1 6.87559 6
d
E H σ · − − ⋅
Let’s give an example. We are at 10,000 feet pressure altitude at 100 degrees F. The
pressure ratio is:
5.2559
(1 6.87559 6 10, 000) 0.6877 E δ · − − ⋅ ·
On a standard day, the temperature would have been:
( ) 1 6.87559 6 10, 000 0.9312 E θ · − − ⋅ ·
( ) 288.15 288.15 0.9312 268.3 268.3 273.15 1.8 32 23.3 T F θ · ⋅ · ⋅ · · − ⋅ + · °
The standard day σ is:
0.6877
0.7384
0.9312
σ · ·
Solving for
d
H
[ ]
( )
[ ] 1 4.2559
1 4.2559
1
1
6.87559 6 6.87559 6
d
H
E E
δ
σ
θ
] ]
−
]
] − ]
] ]
] · ·
− −
]
]
]
(3.51)
For the test day temperature of 100 degrees F:
28
( ) 459.67 100
1.0790
518.67
θ
+
· ·
The σ for the test day would be:
0.6877
0.6373
1.0790
δ
σ
θ
· · ·
Then, computing
d
H we get:
1/ 4.2559
0.6877
1 / 6.87559 6
1.0790
d
H E
]
 `
· − −
]
. ,
]
]
(3.52)
14, 607
d
H · feet versus 10,000 feet for
C
H (pressure altitude).
Equation 3.52 shows the density (or σ ) at 100 degrees F at 10,000 feet pressure altitude is
the same as at 14,607 feet pressure altitude on a standard day for that altitude. To check on our
calculations, calculate the standard density ratio for 14,607 feet as follows:
a.
( )
5.2559
1 6.87559 6 14, 607 0.5733 E δ · − − ⋅ · ,
b. (1 6.87559 6 14, 607) 0.8996 E θ · − − ⋅ · , and
c.
0.5733
0.6373
0.8996
δ
σ
θ
· · · .
It checks! The density ratio for 100 degrees F at 10,000 feet pressure altitude is identical to
the density ratio at a density altitude of 14,607 feet.
3.11 Pressure Altitude Error Due to Ambient Pressure Measurement Error
At Edwards AFB, the field elevation (geometric height) of the main runway (22/04) is
2,300 feet. With standard atmospheric conditions, the pressure altitude would also be 2,300
feet. That requires more than just being at standard temperature. As we have derived, pressure
altitude is only a function of ambient pressure and is independent of ambient temperature.
Using the standard atmosphere model formulas, we can compute what a 1foot change in
altitude will produce in ambient pressure. Table 3.5 shows the resultant pressure error for a 1
foot error in pressure altitude.
29
Table 3.5
PRESSURE ERROR VERSUS ALTITUDE ERROR
C
H
(ft)
δ P
(psf)
P ∆
(psf)
P
(in. Hg)
P ∆
(in. Hg)
P
(millibar)
P ∆
(millibar)
0.0 1.00000 2116.22 0.076 29.921 0.0011 1,013.250 0.037
2,300 0.91963 1946.15 0.071 27.516 0.0010 931.820 0.034
10,000 0.68770 1455.33 0.056 20.577 0.0008 696.820 0.027
20,000 0.45954 972.49 0.041 13.750 0.0006 465.630 0.020
30,000 0.29695 628.43 0.029 8.885 0.0004 300.890 0.014
40,000 0.18509 391.68 0.019 5.538 0.0003 187.540 0.009
50,000 0.11446 242.21 0.012 3.425 0.0002 115.972 0.006
Note: The pressure errors are carried to one extra digit than the pressure magnitude.
Data recording system resolution is a limitation for any parameter, but let us use pressure
altitude as an illustration. Looking at the inches of mercury column, one can see that better
than 1/1000th of an inch of mercury accuracy would be required to achieve 1foot accuracy in
pressure altitude. It turns out that such accuracy level instrumentation is available. There are
two other limiting factors on altitude accuracy. First, is the number of digits recorded in the
data stream. The data recording is an 8, 10, 12, 14, or 16 “bit” system. An 8bit system breaks
full scale into
8
2 (or 256) parts. If full scale were 30 in. Hg, then the resolution of ambient
pressure would be 30/256=0.117 in. Hg. At sea level, this would be an altitude error of
0.117 in. Hg/(0.0011 in. Hg/ft)=107 feet. Clearly, this is unacceptable for performance testing.
For higher bit resolution the following numbers are computed:
a.
10
2 = 1,024 P ∆ = 30/1,024= 0.029 in. Hg
C
H ∆ =0.029/0.0011=26 feet
b.
12
2 = 4,096 P ∆ = 30/4,096= 0.0073 in. Hg
C
H ∆ =0.0073/0.0011=6.6 feet
c.
14
2 =16,384 P ∆ = 30/16,384= 0.0018 in. Hg
C
H ∆ =0.0018/0.0011= 1.6 feet
d.
16
2 = 65,536 P ∆ = 30/65,536= 0.0005 in. Hg
C
H ∆ = 0.0005/0.0011= 0.5 feet
Therefore, it appears that at least at sea level, a 14bit system will get us to our goal of 1foot
accuracy. However, let us see what happens at 50,000 feet. We have the same value for
14
2 =16,384:
a. P ∆ =30/16384=0.0018
C
H ∆ =0.0018/0.0002=9.0 ft
Therefore, our error due to recording system resolution is substantially larger at the higher
altitudes. However, a 9foot error at 50,000 feet is considered acceptable. The AFFTC pacer
aircraft use a 16bit system. The second limiting factor on altitude accuracy is the ‘position error,’
discussed in the air data calibration section.
30
4.0 AIRSPEED
4.1 Introduction – Airspeed
Aircraft speed can be expressed in several forms. For this text, generally, the units will be
in either knots (nautical miles per hour) or feet per second, except for Mach number ( M ),
which is dimensionless. Groundspeed (
g
V ) is the physical speed relative to the ground and is
usually expressed as a vector relationship with north, east, and down components. This is due
to obtaining groundspeed from INS (inertial navigation system) or GPS (global positioning
system) data sources. True airspeed (
t
V ) is the physical speed of the aircraft with respect to the
moving air mass. This is usually a scalar quantity, though components of true airspeed can be
computed using axis transformations using INS velocities and angles and windspeeds.
Windspeed (
w
V ) is the speed of the air mass (wind) with respect to the ground. This is also a
vector quantity with north, east and down components. The Mach number ( M ) is the ratio of
true airspeed to the local speed of sound. Mach numbers less than 1 are referred to as subsonic
and those greater than 1 are supersonic. The speed of sound is a function of the square root of
the ambient temperature. Calibrated airspeed (
C
V ) is the speed displayed on a typical cockpit
airspeed indicator. It is a function of only one parameter÷differential (or impact) pressure.
Impact pressure is the difference between total and ambient pressure. The
c
(calibrated) has
two meanings. The first is that calibrated airspeed is ‘calibrated’ to sea level in the sense that it
will be exactly equal to true airspeed at sea level, standard day, but only at that condition. The
second is calibrated versus indicated. A pneumatic instrument (physically driven from pressure
inputs) displays an ‘indicated’ value. The value has instrument and position errors. The
instrument errors are errors due to the instrument itself. Position errors are those due to the
location of pressure probes. There may be some ideal location to place probes where the errors
are zero. However, in the real world, there is no such position so there will always be position
errors of some magnitude. Once instrument and position error corrections are applied, the
indicated airspeed becomes calibrated airspeed.
In aircraft equipped with an ADC (air data computer), those corrections are usually already
applied in the ADC so that the displayed airspeed is calibrated airspeed. Calibrated airspeed, as
mentioned above, is a function only of the impact pressure. That pressure is also designated
compressible dynamic pressure. A measure of airspeed that is a function of incompressible
dynamic pressure is called equivalent airspeed (
e
V ). Structural analysis is often in terms of
incompressible dynamic pressure, so that equivalent airspeed is a useful speed for structural
testing. At sea level, standard day, calibrated airspeed and equivalent airspeed are equal (or
equivalent), but only at that condition.
4.2 Speed of Sound
The speed of sound is computed by the following formula:
( ) a R T γ · ⋅ ⋅ (4.1)
31
where:
a = speed of sound (ft/sec),
γ = 1.40 (ratio of specific heats), and
R = 3,089.8136 ft²/(sec² °K) (from the 1976 U.S. Standard Atmosphere).
For a sea level standard day, T = 288.15 °K. Then,
[ ] 1.40 3089.8136 288.15 a · ⋅ ⋅ (4.2)
= 1,116.4505 ft/sec (usually rounded to 1116.45)
= 661.4788 knots (usually rounded to 661.48)
For the speed of sound at temperatures other than standard sea level,
( )
SL SL
SL
R T
a T
a T
R T
γ
γ
⋅ ⋅
· ·
⋅ ⋅
(4.3)
Then, define θ as the ratio of test day temperature to standard day temperature at sea level.
SL
a a θ · ⋅ (4.4)
4.3 History of the Measurement of the Speed of Sound
From Britannica¹ Online, the speed of sound in air was first measured by the French
scientist Pierre Gassendi in the 1600s at 478.4 meters per second. He “measured the time
difference between spotting the flash of a gun and hearing its report over a long distance.”
Very clever! In the 1650s, two Italians (Giovanni Borelli and Vincenzo Viviani) obtained a
much more accurate value of 350 meters per second. The first precise value was obtained at the
Academy of Sciences in Paris in 1738 at 332 meters per second. Britannica¹ reports a value of
331.45 meters per second was obtained in 1942, which was amended to 331.29 meters per
second in 1986. These values were at 0 degrees C.
In 1942, NACA (National Advisory Committee for Aeronautics) published Report No. 1235.
In that report, they specified the speed of sound at sea level standard day as 1116.89 feet/second.
Converting the NACA number to meters per second and to 0 degrees C:
a.
273.15
1116.89 0.3048 331.45
288.15
a · ⋅ ⋅ · meters/second
In 1962 and again in 1976, the ICAO (International Civil Aviation Organization) agreed
upon constants for use in a standard atmosphere. The speed of sound is not directly defined,
but could be computed from the other constants. The speed of sound at sea level in English and
metric units is as follows:
a. 1116.4505
SL
a · ft/sec = 340.2941 m/sec
32
4.4 The Nautical Mile
The nautical mile (nm) has been set, by international agreement, to exactly 1,852 meters.
The conversion factor from feet to meters is also an exact number÷0.3048 meters per foot.
Therefore, we can compute the number of feet per nautical mile.
a. 1, 852/ 0.3048 6, 076.1155 feet NM · ·
Since a knot is 1 nm per hour, the conversion from knots to feet per second is as follows:
a.
6, 076.115 Hour
feet/sec 1.6878 knots
Hour 3, 600 sec
NM
· ⋅ ·
An early definition of a nautical mile was an even 6,080 feet. It is called the British
nautical mile. With that definition, the conversion factor becomes:
a.
6, 080. Hour
feet/sec 1.6889 knots
Hour 3, 600 sec
NM
· ⋅ ·
One would see the above conversion factor in textbooks published prior to the U.S.
standard atmosphere of 1959, which had many of the same constants as the 1962 and 1976
atmospheres. Using the 1942 speed of sound and the early knots to feet per second conversion
one gets:
a. 1,116.89/1.6889 661.31knots
SL
a · ·
With the modern (as of this writing) values:
b. 1,116.45/1.6878 661.48 knots
SL
a · ·
4.5 True Airspeed
True airspeed (
t
V ) is the physical speed of the vehicle relative to the moving air mass. The
true airspeed is a vector quantity. The relationship between true airspeed and the speed with
respect to the ground (
g
V ) is:
t g w
V V V · +
" " "
= true airspeed vector (4.5)
where:
w
V ·
"
windspeed vector.
4.6 Mach Number
Mach number ( M ) is defined as the ratio of true airspeed to the local speed of sound.
t
V
M
a
· (4.6)
33
We could compute Mach number from Pitotstatic theory with the simple expression for
differential pressure (
C
q ) versus total pressure detected by a Pitot tube (
t
P
′
) and ambient
pressure ( P ). The prime on the total pressure is to denote a measurement behind a normal
shock (for M ≥1). For M <1, the free stream total pressure (
t
P ) and the measured total
pressure (
t
P
′
) are identical. Differential pressure is also compressible dynamic pressure and
often designated impact pressure.
C
q =
t
P
′
− P
Or dividing both sides by P :
C
q
P
=
t
P
′
P
− 1
(4.8)
Using Bernoulli’s Equation for 1 M < :
( )
( ) 1
2
1
1 1
2
C
q
M
P
γ γ
γ
− ]
]
]  ` −
· + ⋅ −
]
. , ]
(4.9)
And the Rayleigh Supersonic Pitot Equation for 1 M ≥ :
( )
( )
( )
( )
( )
( ) 1/ 1
1
2
1 1
2
1
1
1
2
1 2
C
q
M
P
M
γ
γ γ
γ
γ
γ
γ γ
− ]
]
− ]
]
− ]
]
]
+ ]  ` +
]
· ⋅ ⋅ −
]
]
. , ]
− + ⋅ ⋅
]
]
(4.10)
Substituting γ =1.40 for M <1:
( )
3.5
2
1 0.2 1
c
q
M
P
· + ⋅ − (4.11)
Solving for M in equation 4.11:
[ ] 1 3.5
5 1 1
C
q
M
P
] ¹ ¹
¹ ¹
 `
· ⋅ + −
] ' '
. ,
¹ ¹ ] ¹ ¹ ]
(4.12)
For 1 M ≥ :
( )
( )
2.5
3.5
2
2
2.4
1.2 1
0.4 2.8
C
q
M
P
M
¹ ¹
]
¹ ¹
] · ⋅ ⋅ −
' '
− + ⋅ ]
¹ ¹
]
¹ ¹
(4.13)
34
Multiply by 1= (2.50/2.50)
2.5
and collect terms. Multiply the first term {
2 3.5
(1.2 ) M ⋅ } by
2.5
2.50 and divide the second term in the { } brackets by the same
2.5
2.50 factor.
( )
( )
( )
2.5
2 3.5 3.5 2.5
2.5
2
2.4
1.2 2.5 1
0.4 2.5 2.8 2.5
C
q
M
P
M
⋅
· ⋅ ⋅ ⋅ −
− ⋅ + ⋅ ⋅
(4.14)
( )
7
3.5 2.5 2.5
2.5
2
1.2 2.5 2.4 1
7 1
M
M
· ⋅ ⋅ ⋅ −
⋅ −
(4.15)
3.5 2.5 2.5
1.2 2.5 2.4 166.9215801 ⋅ ⋅ · (round to 166.9216)
( )
7
2.5
2
166.9216 1
7 1
C
q
M
P
M
]
]
· ⋅ −
]
⋅ −
]
(4.16)
Note that one produces the identical value for /
C
q P when M = 1.0 is inserted into either
the subsonic (equation 4.11) or supersonic (equation 4.16) formula. For example:
a.
1.0
/ 0.892929
C
M
q P
·
·
Solving for M in the supersonic formula (4.16), first add 1 to both sides, then multiply
both sides by the term
( )
2.5
2
7 1 M ⋅ − .
( )
2.5
2 7
1 7 1 166.9216
C
q
M M
P
 `
+ ⋅ ⋅ − · ⋅
. ,
Then, divide both sides by
( )
2.5
2
7 M ⋅ .
( )
( )
[ ]
2.5 7
2
2
2 2 2.5 2.5
166.9216
7 1
1 1.287560
7
7
C
M
M
q
M
P
M
M
⋅
⋅
] ⋅ −
 `
+ ⋅ · · ⋅
]
. , ⋅
⋅
]
Finally, solve for M from the M on the right side.
2.5
2
1
0.881285 1 1
7
C
q
M
P
M
]
 `
 `
]
· ⋅ + ⋅ −
] . , ] ⋅
] . ,
]
]
(4.17)
As can be seen, M appears on both sides of the equation. One method to approach the
supersonic M calculation in a computer algorithm is first determine if M is indeed greater
than 1.0 by calculating M from the subsonic equation (4.12). If M is greater than 1.0 at that
point, then use the value of M from the subsonic equation as the initial condition in the
35
supersonic equation. Then perform a simple iteration until M converges to a value  usually in
just a few iterations.
4.7 Total and Ambient Temperature
A total temperature probe is used to measure total temperature (
t
T ). Assuming this probe is
in the freestream with no heat loss (adiabatic), then the relationship between total temperature
and ambient temperature (T ) is as follows:
( )
( )
2 2
1
1 1 0.2
2
t
T T M T M
γ  ` −
· ⋅ + ⋅ · ⋅ + ⋅
. ,
(4.18)
4.8 Calibrated Airspeed
Historically, airspeed indicators were constructed with a single pressure input being the
differential pressure (
C
q ). The gauge is “calibrated” to read true airspeed at sea level standard
pressure and temperature. The subsonic and supersonic Mach number equations are used with
the simple substitutions of ( /
C SL
V a ) for M and
SL
P for P . However, the condition for which
the equations are used is no longer subsonic ( M <1) or supersonic ( M >1) but rather
calibrated airspeed being less or greater than the speed of sound (
SL
a ), standard day, sea level
(661.48 knots).
For
C SL
V a < :
3.5
2
1 0.2 1
C C
SL SL
q V
P a
]
 `
· + ⋅ −
]
. ,
]
(4.19)
(1 3.5)
5 1 1
C
C SL
SL
q
V a
P
¹ ¹ ]
¹ ¹
 `
· ⋅ ⋅ + −
' ' ]
. ,
¹ ¹ ] ¹ ¹
(4.20)
For
C SL
V a ≥ :
( )
( )
7
2.5
2
166.9216
1
7 1
C SL
C
SL
C SL
V a
q
P
V a
⋅
· −
]
⋅ −
]
(4.21)
Solving for
C
V and noting that the formula is similar in form to the M equation, we will
leave out intermediate steps.
36
2.5
2
1
0.881285 1 1
7
C
C SL
SL
C
SL
q
V a
P
V
a
¹ ¹
]
]
¹ ¹
]
]
¹ ¹
 `
] ] · ⋅ ⋅ + ⋅ −
' '
. , ] ]  `
¹ ¹
⋅
] ]
¹ ¹
. , ]
]
¹ ¹
(4.22)
Notice the differences between equations 4.22 and 4.17. We will leave it to the reader to
make that comparison.
Note that
C
V occurs on both sides of equation 4.22. The solution is simply to use the
subsonic formula to obtain a first iteration, then successively iterate on the above equation. It
will converge in just a few steps. It should be emphasized that the supersonic formula is
C SL
V a > and not 1 M > .
Figure 4.1 illustrates the difference of true airspeed versus calibrated airspeed. In summary,
the true airspeed is the physical speed of the aircraft with respect to the moving air mass, while
the calibrated air speed is directly proportional to compressible dynamic pressure. The two
measures of airspeed are identical at sea level, standard day.
True Airspeed (standard day) versus Calibrated Airspeed
0
200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,000
0 100 200 300 400 500 600 700 800
Calibrated Airspeed (kts)
T
r
u
e
A
i
r
s
p
e
e
d
(
S
t
a
n
d
a
r
d
D
a
y
)
(
k
t
s
)
H = Sea Level
H = 10,000 ft
H = 20,000 ft
H = 30,000 ft
H = 40,000 ft
H = 50,000 ft
Note: At 50,000 feet, calibrated airspeed is about ½ of true airspeed.
Figure 4.1 True Airspeed versus Calibrated Airspeed
37
4.9 Equivalent Airspeed
Equivalent airspeed is defined from the incompressible dynamic pressure formula.
2 2
0
0.5 0.5
t e
q V V ρ ρ · ⋅ ⋅ · ⋅ ⋅ (4.23)
0
;
SL
SL
ρ
ρ ρ σ
ρ
· · (4.24)
2 2
e t
V V σ · ⋅ (4.25)
e t
V V σ · ⋅ (4.26)
For the performance engineer, there is no practical reason to use equivalent airspeed for
anything. However, structural analysis is often performed in terms of equivalent airspeed
(since it is a direct function of the incompressible dynamic pressure), so the performance
engineer needs to be able to convert
e
V to parameters that are more useful. Besides equation
4.26, another useful equation is derived. Since Mach number is
( )
t t
SL
V V
M
a
a θ
· ·
⋅
(4.27)
And
δ
σ
θ
· , then
( )
( ) e t SL
V V a M
δ
σ θ
θ
· ⋅ · ⋅ ⋅
( )
e
SL
V
M
a δ
·
⋅
(4.28)
Therefore, the equation 4.28 is a handy conversion between
e
V and M . Notice that it is not a
function of temperature.
4.10 Mach Number from True Airspeed and Total Temperature
If one has an accurate direct measure of
t
V , then M can be computed with the additional
measurement of total temperature (
t
T ). The direct
t
V measure could come from laser
velocimetry. For example:
288.15
t SL
T
V a M
 `
· ⋅ ⋅
. ,
(4.29)
38
( )
( )
288.15
661.48
t
V
M
T
⋅
·
⋅
(4.30)
Recalling the total temperature equation 4.18,
( )
2
1 0.2
t
T T M · ⋅ + ⋅ and solving for T :
( )
2
1 0.2
t
T
T
M
·
+ ⋅
(4.31)
Then, one would iterate between the and M T equations (4.30 and 4.31). An initial estimate of
standard day might be chosen for the initial value of T for the iteration.
In this case, M is a function of ambient temperature (T ). This is due to the way we have
chosen to compute M using a measurement of
t
V . At the time of this writing, the technology to
directly measure true airspeed was not generally available so one must rely on computing M
from total (
t
P ) and ambient ( P ) pressure measurements.
4.11 Airspeed Error Due to Error in Total Pressure
An error analysis was presented at the end of the altitude section. That error analysis
showed the effect of an error in ambient pressure on pressure altitude. A similar analysis can
be performed for an error in total pressure and its effect on the calculation of true airspeed.
Figure 4.2 shows that effect for an error of 0.001 in. Hg in the total pressure measurement.
Effect of 0.001 InHg Error in Total Pressure
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 200 400 600 800 1000 1200 1400 1600 1800
True Airspeed (kts)
D
e
l
t
a
T
r
u
e
A
i
r
s
p
e
e
d
(
k
t
s
)
H = 10,000 ft
H = 30,000 ft
H = 50,000 ft
Figure 4.2 True Airspeed Error for 0.001 in. Hg Error
39
We have summarized the functional relationships derived in the altitude and airspeed
sections as functions of three basic measurements: total pressure (
t
P
′
), ambient (or static)
pressure ( P ), and total temperature (
t
T ).
a. ( )
C
H f P · pressure altitude,
b. ( )
C C
V f q · calibrated airspeed,
c.
C t
q P P
′
· − compressible dynamic pressure,
d. ( , )
t
M f P P
′
· Mach number. Note that Mach number is obtained without a
measurement of temperature,
e. ( , )
t
T f T M · ambient temperature, and
f. ( , )
t
V f M T · true airspeed.
40
5.0 LIFT AND DRAG
5.1 Introduction
The aerodynamic force axis system used for aircraft performance is defined by the true
airspeed vector. Assuming zero sideslip angle ( β ), the force parallel to true airspeed (
t
V ) is
the retarding force drag ( D). Octave Chanute in his 1897 book, Progress in Flying Machines
(Reference 5.1), uses the terminology resistance for what we now refer to as drag. The force
perpendicular to the true airspeed vector is the lift ( L ) force.
5.2 Definition of Lift and Drag Coefficient Relationships
Lift and drag are referenced to incompressible dynamic pressure and a reference area so
that the coefficients are nondimensional. In aircraft applications, the area is a reference wing
area. The constants in the following equations are derived from the 1976 U.S. Standard
Atmosphere (which are the same as in the 1962 U.S. Standard Atmosphere below 65,000 feet).
The lift and drag coefficients are defined as follows:
( ) /
D
C D q S · ⋅ drag coefficient (5.1)
( ) /
L
C L q S · ⋅ lift coefficient (5.2)
where:
D = drag (pounds),
L = lift (pounds),
q = incompressible dynamic pressure (pounds/feet²), and
S = reference wing area (feet²).
Defining q :
2 2
0.5 0.7
t
q V P M ρ · ⋅ ⋅ · ⋅ ⋅ (5.3)
To show how the above equivalence is developed, we use formulas we previously derived.
a.
( )
P
R T
ρ ·
⋅
,
b.
t
V R T M γ · ⋅ ⋅ ⋅ , and
c.
( )
( )
2 2
0.5 0.5 0.5 1.4 0.7
t
P
q V R T M M P M
R T
ρ γ · ⋅ ⋅ · ⋅ ⋅ ⋅ ⋅ · ⋅ ⋅ · ⋅ ⋅
⋅
.
Figure 5.1 illustrates the difference between the compressible (
C
q ) and incompressible
( ) q dynamic pressure.
41
Ratio of Compressible to Incompre ssibl e Dynamic Pressure
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
0.0 0.5 1.0 1.5 2.0 2.5
Mach Number
q
c
/
q
b
a
r
Figure 5.1 Ratio of Compressible to Incompressible Dynamic Pressure
More convenient forms for
D
C and
L
C are as follows:
2116.2166 P δ · ⋅ (usually rounded to 2116.22) (pounds per ft
2
)
2 2
0.7 2116.22 1481.3516 q M M δ δ · ⋅ ⋅ ⋅ · ⋅ ⋅ (5.4)
( )
2
0.00067506
D
C D M S δ · ⋅ ⋅ ⋅ (5.5)
(The constant is usually rounded to 0.000675)
A drag coefficient of 0.0001 is defined as one drag count.
( )
2
0.00067506
L
C L M S δ · ⋅ ⋅ ⋅ (5.6)
5.3 The Drag Polar and Lift Curve
The drag polar and lift curve are usually presented as a function of lift coefficient and
Mach number as follows:
a. ( , ) drag polar
D L
C f C M · , and
b. ( , ) lift curve
L
f C M α · .
This is typically for a reference logitudinal center of gravity and Reynolds number or altitude.
42
5.4 Reynolds Number
Reynolds number is defined as follows:
t
V l
RN
ρ
µ
⋅ ⋅
· (5.7)
where:
RN = Reynolds number,
l = characteristic length (feet) ( l is usually the MAC [mean aerodynamic chord]), and
µ = viscosity (slugs/[feet sec]).
To compute viscosity, we used Sutherland’s Law, which is a relationship for µ in terms of
ambient temperature. We define an index that is a ratio of Reynolds number to the Reynolds
number at standard day, sea level at a given Mach number.
( )
2
110
398.15
T
RNI
δ
θ
] +
 `
· ⋅
]
. ,
]
(5.8)
(Note that if one were to insert standard day, sea level
values into the RNI equation you would get 1.00.)
where:
RNI = Reynolds number index. Then,
(7.101 6) RN E M l RNI · + ⋅ ⋅ ⋅ (5.9)
For a characteristic length ( l ) of 1.0, Table 5.1 gives a sense of the magnitude of RN . The
numbers used are for standard day.
Table 5.1
REYNOLDS NUMBER VARIATION WITH MACH NUMBER AND ALTITUDE
Mach
Number
Altitude
(ft)
δ
T
(deg K)
θ
RNI
/ RN l
(10
6
/ft)
C
V
(knots)
0.10 0 1.0000 288.15 1.0000 1.0000 0.7101 66.1
0.20 0 1.0000 288.15 1.0000 1.0000 1.4202 132.3
0.60 0 1.0000 288.15 1.0000 1.0000 4.2606 396.9
1.00 0 1.0000 288.15 1.0000 1.0000 7.1010 661.48
1.20 0 1.0000 288.15 1.0000 1.0000 8.5212 793.8
0.60 30,000 0.2970 228.71 0.7937 0.4010 1.7985 223.0
1.00 30,000 0.2970 228.71 0.7937 0.4010 2.8474 390.0
1.60 30,000 0.2970 228.71 0.7397 0.4010 4.5559 643.0
0.60 60,000 0.0708 216.65 0.7519 0.1027 0.4377 110.0
1.00 60,000 0.0708 216.65 0.7519 0.1027 0.7294 196.6
1.60 60,000 0.0708 216.65 0.7519 0.1027 1.1671 340.9
2.00 60,000 0.0708 216.65 0.7519 0.1027 1.4588 430.0
3.00 60,000 0.0708 216.65 0.7519 0.1027 2.1882 626.9
43
The drag coefficient due to skin friction is typically as much as 70 percent of minimum
drag coefficient and is a significant factor in the corrections to the drag polar. It is typical that
the Reynolds number correction is on the order of 1 drag count (0.0001
D
C ) per 2,000 feet of
pressure altitude. This is also a function of temperature, which cannot be ignored. For 10
degrees K off standard day, typically, a 1drag count effect can be encountered.
5.5 Skin Friction Drag Relationships
The following empirical flat plate relationships were developed by Ludwig Prandtl and
others. In Incompressible Aerodynamics (Reference 5.2), equation 5.10 is a turbulent skin
friction drag formula attributed to Schlichting.
2.58
10
0.455
(log )
f
C
RN
· (5.10)
Effect of Mach number:
( )
0.65
2
1 0.144
f compressible f
C C M
−
· ⋅ + ⋅ (5.11)
All of the sample problems in this text used equations 5.10 and 5.11.
wet
D f
S
C C
S
 `
· ⋅
. ,
(5.12)
An earlier friction drag equation is one developed by Prandtl and is shown in equation 5.13.
5
0.074
f
C
RN
· (5.13)
A laminar flow empirical formula was developed by Blasius and shown in equation 5.14.
1.328
f
C
RN
· (5.14)
A transition formula between laminar and turbulent is attributed to Prandtl and Gebers and
shown in 5.15.
5
0.074 1, 700
f
C
RN RN
· − (5.15)
Equations 5.10 and 5.13 through 5.15 are plotted versus the logarithm to the base 10 of
Reynolds number in Figure 5.2.
44
Empirical Skin Friction Drag Relationships
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
5.0 5.5 6.0 6.5 7.0 7.5
Log 10 (Reynolds Number)
S
k
i
n
F
r
i
c
t
i
o
n
D
r
a
g
C
o
e
f
f
i
c
i
e
n
t
(
C
f
)
Turbulent: Prandtl
Turbulent: Schlichting
Laminar: Blasius
Transistion: PrandtlGebers
Figure 5.2 Skin Friction Drag Relationships
5.6 Idealized Drag Due to Lift Theories
The following idealized theoretical drag due to lift models can be found in numerous
aeronautical engineering textbooks listed in the Bibliography. One of the best textbooks (in the
author’s opinion) titled, “Wing Theory” (Reference 5.3), was written by a pioneer in the wing
theory field, R.T. Jones.
a. Subsonic 1 M <<
(1) Elliptic Wing Theory
2
2
2
1
L
L
L D
C
C C
AR
AR
π
α
π
⋅
· ⋅ ·
⋅  `
+
. ,
(5.16)
Transonic 1 M ≈
(1) Slender Body Theory
2
2
2
L
L
L D
C
C AR C
AR
π
α
π
⋅
· ⋅ ⋅ ·
⋅
(5.17)
Supersonic 1 M >
45
(1) Thin Wing Theory
2
2
2
4 1
4
1
L
L D L L
M
C C C C
M
α
α
⋅ −
· · ⋅ · ⋅
−
(5.18)
All of the above are idealized and are presented only for general trends. One idealization
made is symmetry (i.e., wing is uncambered and at zero incidence angle.)
5.7 Air Force Flight Test Center Drag Model Formulation
The following equations are drag model formulations that have been proven at the AFFTC
to quite adequately curve fit actual flight test data. For a given Mach number and RN :
( ) ( )
2 2
min min
1 2
D D L L L Lb
C C K C C K C C · + ⋅ − + ⋅ − (5.19)
where:
2 0 K · when
L Lb
C C < .
The 1 K term in the drag polar model above is the pure parabola portion. The 2 K term is
zero below a ‘break’
L
C
and therefore, contributes nothing to the model until the lift
coefficient exceeds this break lift coefficient. The break lift coefficient could be thought of as
the point where flow separation begins and the drag model becomes nonlinear.
5.8 The Terminology ‘Drag Polar’
The terminology ‘drag polar’ was first used by Eiffel. That historical note is found in
Introduction to Flight, Third Edition (Reference 5.4), by John D. Anderson. However, a
second source, lists Otto Lilienthal as the ‘inventor’ of the drag polar (a.k.a., a polar plot or a
polar diagram). The term ‘polar’ is a reference to polar coordinates. A given point on a
Cartesian (xy) plot can be defined by a radius and an angle. Figure 5.3 shows two drag models
plotted. The first drag model is a pure parabola. This is the same model used in the sample
performance model section of this handbook for 0.8 M · . The second drag model represents
that parabolic model plus a deviation from the pure parabola.
46
Drag "Polar"
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000 0.1200 0.1400 0.1600 0.1800 0.2000
Drag Coefficient (CD)
L
i
f
t
C
o
e
f
f
i
c
i
e
n
t
(
C
L
)
Parabolic Model
Nonlinear Model
Tangent to Curve
Figure 5.3 Drag Polar
A secondorder parabola reasonably represents drag polar data only up to the point where
flow separation begins. A second parabola that adds to the first after the start of flow
separation has been quite successful in curve fitting AFFTC drag model formulations. The
equation for this specific parabolic model is equation 5.20 and the equation for the nonlinear
model is equation 5.21 (modified by 5.22).
( )
2
0.02 0.132 0.06
D L
C C · + ⋅ − (5.20)
( ) ( )
2 2
0.02 0.132 0.06 0.2642 0.60
D L L
C C C · + ⋅ − + ⋅ − (5.21)
( ) 0.60 0 for 0.60
L L
C C − · < (5.22)
We can plot the ratio of lift to drag, which is the same as the ratio of lift coefficient to drag
coefficient.
L
D
C
L
D C
· (5.23)
Figure 5.4 presents this lifttodrag versus lift coefficient for both the linear and the
nonlinear model. This model is a rough approximation to an actual F16A drag polar at
0.8 M · . As Figures 5.3 and 5.4 show, the drag grows substantially after the lift coefficient
increases beyond 0.6.
47
L/ D versus Lift Coeffi cient
0
1
2
3
4
5
6
7
8
9
10
11
12
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Li ft Coef fi cient (CL)
L
i
f
t
o
v
e
r
D
r
a
g
(
C
L
/
C
D
)
Par abol ic Model
Nonli near Model
Figure 5.4 LifttoDrag Ratio versus Lift Coefficient
Very roughly, maximum thrust stabilized turns occur around 0.8 lift coefficient. The
aircraft has an angleofattack limiter, which corresponds to a lift coefficient of around 1.5. At
this limit lift coefficient, this model has the following values for drag coefficient:
a. 1.50
L
C · , and
b. 0.5077
D
C · .
These are reasonable values. Let’s do a sample calculation. Assume an airplane gross
weight of 20,000 pounds, a pressure altitude of 30,000 feet, and a Mach number of 0.80.
Ignore the thrust component in lift and drag coefficient. The F16A reference wing area is 300
ft
2
. The pressure ratio (δ ) at 30,000 feet is 0.297. Solving for lift and drag from equations 5.5
and 5.6:
2 2
1.5 0.297 0.8 300.
126, 720.
0.000675 0.000675
L
C M S
L
δ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅
· · · (5.24)
2 2
0.5077 0.297 0.8 300.
42, 890.
0.000675 0.000675
D
C M S
D
δ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅
· · · (5.25)
For our 20,000pound aircraft (ignoring thrust component), the normal load factor can be
calculated as follows:
48
126, 720.
126, 720. 6.34 g's
20, 000.
z t z
L N W N · ⋅ · → · ·
Let’s say that someone told us that the aircraft could sustain 4.5 g’s in maximum
afterburner at these conditions. Since thrust equals drag in a sustained (or thrustlimited) turn,
we can calculate the drag by first calculating the lift coefficient.
2 2
0.000675 0.000675 4.5 20, 000.
1.07
0.297 0.8 300.
z t
L
N W
C
M S δ
⋅ ⋅ ⋅ ⋅
· · ·
⋅ ⋅ ⋅ ⋅
(5.26)
From the drag polar equation (5.21), the drag coefficient comes to 0.2130. Solving for drag
(which is equal to net thrust):
2 2
0.2130 0.297 0.8 300.
17, 994.
0.000675 0.000675
D
C M S
D
δ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅
· · · (5.27)
At the maximum lift point, the excess thrust is:
17, 994. 42, 890. 24, 895.
ex n
F F D · − · − · − (5.28)
That would be a longitudinal load factor of greater than a 1 g. The deceleration rate in
knots per second comes to:
0
24, 895.
1.25
20, 000.
t
x
t
V
h
N
V g
−
· · − · +
!
!
(5.29)
Assuming all the negative excess thrust is in deceleration (constant altitude slow down
turn):
( )
( )
2
ft
1.25 32.174
sec
knots
23.8
sec
ft
sec
1.6878
knot
t
V
− ⋅
· · −
 `
. ,
! (5.30)
SECTION 5.0 REFERENCES
5.1 Chanute, Octave. 1897. Progress in Flying Machines, The American Engineer and Railroad
Journal.
5.2 Twaites, Bryan, ed. 1960. Incompressible Aerodynamics: An Account of the Steady Flow of
Incompressible Fluid past Aerofoils, Wings, and other Bodies. Dover Publications.
5.3 Jones, Robert T. 1990. Wing Theory. Princeton University Press.
5.4 Anderson, John D. 1989. Introduction to Flight, Third Edition. New York, New York:
McGrawHill, Inc.
49
6.0 THRUST
6.1 Introduction
We will leave it to numerous other documents to discuss in detail the overall topic of
propulsion. In this text, we are concerned just with the measurement of thrust. We will discuss
turbine engines and propellerdriven piston engines. The term measurement is a misnomer,
since inflight thrust is a calculation based upon a number of separate measurements. Only on
the ground, either in an engine cell or during a static thrust run, do we actually measure thrust
using load cells. We will start by giving the basic principles of turbine engine thrust.
Figure 6.1 represents a turbojet engine. Other turbine engine types include low and
highbypass ratio turbofans. A turbofan engine has two separate turbine sections: a high
pressure section which drives the compressor, and a low pressure section which drives the fan.
The air flowing through the fan, referred to as bypass airflow, can be mixed with the core
airflow following the turbine, or it can be exhausted separately. Bypass ratio is the ratio of
bypass to core airflow. In addition, an afterburner (additional fuel added after the turbine
section) may be added for additional takeoff or maneuvering thrust. Engines that are more
exotic include ramjet types, as well as variable cycle engines, where the bypass ratio varies
with flight conditions and/or power level.
Air enters the engine at the face of the diffuser (Figure 6.1), the inlet. The usual station
designation for the engine face is station two. The numerical designation of the exit plane
varies with the engine complexity, so we will simply use a subscripte (e for exit).
Figure 6.1 Turbine Engine Schematic
0 t t
V V · = true airspeed (ft/sec)
2 0
t r t
P P η · ⋅ (lbs/ft
2
) total (average) pressure at station 2 (6.1)
where:
r
η = inlet recovery factor (addressed in more detail later), and
0
t
P = free stream total (average) pressure (lbs/ft
2
).
0 1 2 3 4 5 6
50
( )
0
3.5
2
1 0.2
t
P P M · ⋅ + ⋅ ( pounds/ft
2
) (6.2)
where:
P ·ambient pressure (lbs/ft
2
).
Note: All of the velocities and pressures are integrated average values.
6.2 The Thrust Equation
The net propulsive force on the vehicle is called net thrust (
n
F ). The basic thrust equation
is gross thrust (
g
F ) minus ram drag (
r
F ). The gross thrust, in layman’s terms, is thrust out the
back. Ram drag is the result of slowing the air from free stream to near zero speed at the inlet.
The term ( )
e e
A P P ⋅ − in the equation for gross thrust, 6.4 below, is the result of the pressure at
the exhaust plane being higher, in most cases, than the ambient pressure. However, this is
generally a small term compared to the
( )
a f e
W W V + ⋅
!
term.
n g r
F F F · − (6.3)
( ) ( )
g a f e e e
F W W V A P P · + ⋅ + ⋅ −
!
(6.4)
r a t
F W V · ⋅
!
(6.5)
where:
a
W
!
= airflow rate (lbs/sec) through the engine,
f
W = fuel flow (lbs/sec),
e
V = exit velocity (ft/sec) (average),
e
P = pressure (average) across exit plane (lbs/ft
2
), and
e
A = cross sectional area of the exit nozzle (ft
2
).
For turbofan engines an additional pressure times area term must be added to equation 6.4
when the fan thrust is exhausted separately. Previously defined was the fuel flow (
f
W ),
however, now we will think of it in units of pounds per second to be consistent with the airflow
rate. Note that the total mass flow into the engine is airflow, while exiting the engine mass
flow is airflow plus fuel flow. A more precise engine thrust computation would take into
account various bleed airs that extract air off the engine for cooling and other purposes.
The engine manufacturer will often provide an engine inflight thrust deck÷a computer
program with numerous inputs and outputs on engine performance and operating
characteristics. The terminology deck is left over from when this computer program was a
stack of punched computer cards.
51
6.3 InFlight Thrust Deck
The engine manufacturerprovided inflight thrust deck would vary in complexity. For the
complex augmented turbofans on the F15 and F16 engines, built by Pratt and Whitney and
General Electric, the decks are many thousands of lines of computer code plus extensive data
table lookups. These computer programs are developed using proprietary prediction methods
supplemented by engine test cell data. For the performance engineer, the deck is a black box
with numerous instrumentation measurement inputs. The inputs fall into two categories:
a. Flight conditions: Mach Number ( M ), pressure altitude (
C
H ), and ambient
temperature (T ).
b. Engine parameters: fuel flow, pressure, temperature, and fan and compressor rpm. The
engine rpm’s are the rotation rates of the rotating components. A turbojet engine may have just
a single rpm. A turbofan engine will have more than one turbine section, rotating at different
speeds. The airframe manufacturer will add options to the deck to account for installation
effects such as inlet spillage drag, airflow bleeds, and scrubbing drag.
6.4 Status Deck
The status deck, or prediction deck, predicts the performance (or status) of the engine
usually with flight conditions and throttle position (or power lever angle). In addition, fuel
flow or rotor speed may be input. This deck may contain many of the same components as the
thrust deck. The status deck will predict the pressure, temperature, rpm, and fuel flow that are
inputs to the thrust deck. Most importantly, the status deck also predicts thrust, and in the case
where fuel flow is not input, also fuel flow. In addition, in some cases the status deck could
have rpm and fuel flow as inputs and then would become an inflight thrust deck.
6.5 Inlet Recovery Factor
The inlet recovery factor ( 1.0
r
η ≤ ) is the total pressure loss factor at the engine inlet face.
Gross thrust will be degraded directly proportional to the reduction of
r
η below its maximum
value of 1.0 (100percent recovery). The terminology recovery refers to how much of the free
stream total pressure the engine inlet is able to recover. At subsonic conditions ( 1.0 M < ), the
r
η is typically quite close to 1.0. The recovery factor can be computed using the total pressure
formula below. By measuring the total pressure in the inlet, then we can compute the recovery
factor. The total pressure varies significantly over the face of the inlet. This pressure variation
is called distortion. Computing an average total pressure requires several pressure
measurements performed all across the inlet. This poses two problems. First, we would disturb
the flow in the inlet. This violates the most fundamental rule of instrumentation÷do not affect
what you are measuring by the act of measuring it. The second problem is components of these
inlet rakes may break off in the inlet, causing engine damage or failure. At supersonic speeds,
the inlet recovery factor becomes less than 1.0 due to shock waves in the inlet. In a normal
shock inlet, this recovery factor is about what one would see across an ideal normal shock. The
formula for that is the same as for the normal shock relationship for total pressure
measurement in a nose boom. From the Rayleigh supersonic Pitot equation:
52
( )
2
7
2.5
2
166.9216
7 1
t
M
P P
M
]
]
· ⋅ ⋅
]
⋅ −
]
(6.6)
The free stream total pressure is just the subsonic formula.
0
3.5
2
1 0.2
t
P P M ] · ⋅ + ⋅
]
(6.7)
Then, the recovery factor is the ratio of these two:
2
0
t
r
t
P
P
η · (6.8)
Figure 6.2 is a plot of this relationship.
Normal Shock Recovery Factor
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Mach Number
R
e
c
o
v
e
r
y
F
a
c
t
o
r

P
t
2
/
P
t
o
Figure 6.2 Normal Shock Recovery Factor
The significance of Figure 6.2 is that for Mach numbers above approximately 1.6, the
pressure losses become quite large (greater than 10 percent). The F16 has a normal shock inlet
and at speeds above 1.6; the actual inlet recovery is modeled quite accurately by the normal
shock equation. The F15, in contrast, has a series of inlet ramps, which turn the flow through
oblique shocks as shown in Figure 6.3.
53
Figure 6.3 F15 Inlet Schematic
The net effect of this oblique shock inlet is that at Mach number = 2.0, the inlet recovery
factor is about 0.92 versus only 0.72 for the normal shock inlet. The downside is the increased
complexity of the inlet producing an increase in aircraft weight. At subsonic speeds, the
recovery factor of the F15 oblique shock inlet is slightly less than that for the F16. This is
probably due to the losses in turning the flow.
6.6 Thrust Runs
Checks of installed net thrust can be performed at zero speed using a thrust stand. A thrust
stand may be as simple as a cable with a load cell. The thrust stand gives the only direct
measurement of installed thrust. In contrast, inflight thrust is a computation based upon a
large number of measurements and a computer model of the engine to predict or estimate the
thrust. From the measured thrust stand values, one can compare to values of thrust from both
the inflight thrust and status decks. This test most certainly should be performed on all
performance test programs.
The most significant test points would be the fixed throttle points (IDLE, MIL and MAX
or whatever your fixed throttle points are called). The importance of these points is that the
direct comparison to both the inflight and status decks is possible. Intermediate throttle
position data points are of less value, since the throttle positions are not distinct and repeatable.
The suggestion, since thrust stand time is costly, is to concentrate on getting a number of fixed
throttle data points and ignore the intermediate points. A good test procedure might be to start
the tests in the early morning when it is relatively cold. Get a few data points for the three
fixed power points. For instance, start the engine(s), collect data at IDLE, then go to MIL, then
to MAX, back to MIL, back to IDLE, and repeat at least once. Collect continuous data to
observe stabilization times. However, it should not be necessary to collect the excessive
amounts of data (10+ minutes at one condition would be considered excessive) that some
propulsion analysts may desire. Going up and then back down in throttle determines if there is
any thrust hysteresis (get a different value if increasing throttle versus decreasing throttle).
After collecting that data in early morning, proceed to shut the aircraft engines down and
wait. Refuel if necessary. After the temperature increases some by late morning, repeat the
whole procedure. Finally, do the process a third time in the afternoon. This will give you a
range of ambient temperatures. During the summer at Edwards AFB, that range of temperature
could be as much as 50 degrees F (see Appendix C for average surface temperatures). In 1 day
of testing, you should get IDLE, MIL and MAX data at three temperatures.
54
6.7 Thrust Dynamics
In an engine test cell, the engine manufacturer will perform throttle transients. This data
can be used to develop a thrust dynamics model for use with a takeoff simulation. The typical
aircraft is unable to stabilize at the start of a takeoff with maximum thrust. Therefore, a throttle
transient is necessary to initiate the takeoff. Figure 6.4 is an example of some actual throttle
transient data taken on the AFFTC thrust stand.
Thrust Lag versus Time
10
0
10
20
30
40
50
60
70
80
90
100
0 2 4 6 8 10 12 14
Elapsed Time (sec)
P
e
r
c
e
n
t
C
h
a
n
g
e
i
n
T
h
r
u
s
t
[
M
I
L
t
o
M
A
X
]
Figure 6.4 Thrust Dynamics from an Air Force Flight Test Center Thrust Stand
The thrust stand at the time this data was taken (late 1980s) had a 1 sample per second
sample rate. In addition, it is unknown how much of the lag is due to lag in the
instrumentation. However, using this thrust stand lag data allowed us to match the actual time
to liftoff data very accurately. As an example, for this aircraft, the time to liftoff at one
particular condition was 41.5 seconds using the simulation. For the same simulation, but
assuming 100 percent thrust at time zero, the time to liftoff was computed to be 39.1 seconds
(or over 5 percent). The change in distance to liftoff, for the same liftoff speed, was less than
1 percent. To clarify, the effect of the engine lag occurs in the early portion of the takeoff
ground roll, affecting time to takeoff much more than distance to liftoff. This becomes
significant when considering minimum interval takeoffs, for instance.
6.8 Propeller Thrust
In the examples, it was assumed that thrust was derived from a jet engine. We do not wish
to assume that is always the case. The equations of motion are just as applicable to an aircraft
powered by an engine that drives a propeller. The common unit of output power of an engine is
horsepower. In the English system, 1 horsepower was defined by James Watt in the 1700s to
55
equal 33,000 footpounds of work per minute. In aircraft applications, we will usually divide
by 60 to get 550 footpounds of work per second. As with jet engines, an engine ‘rating’ will
usually not include friction losses and transmission losses to the propeller. We start with an
indicated horsepower ( IHP), which is some fraction (up to maximum power of 100 percent)
of the rating. Then, reduce that by a factor to account for losses to the propeller ( λ ). This
factor can be 10 percent or more. That produces the shaft horsepower or brake horsepower
( BHP ).
( ) BHP IHP λ · ⋅ (6.9)
Then, there is the fact that the propeller cannot possibly convert 100 percent of the brake
horsepower to propulsive force. That factor is the propeller efficiency (η ). The result is thrust
horsepower (THP ).
( ) THP BHP η · ⋅ (6.10)
Each propeller manufacturer will usually provide propeller efficiency charts from which
one can estimate η as a function of propeller rpm, pitch, and flight conditions. If such charts
are not available, one can perhaps find similar charts for similar propellers. If all else fails,
assume a value like 0.80 as a starting point in developing a propulsion model from flight test.
From the definition of horsepower, the equation for thrust horsepower in terms of thrust
and true airspeed is as follows:
550
n t
F V
THP
⋅
· (where
t
V has units of feet/sec) (6.11)
550
n
t
THP
F
V
⋅
· (6.12)
Obviously, equation 6.12 cannot be used at zero speed. For takeoff performance, the static
thrust could be measured on a thrust stand. Then at speeds around liftoff, equation 6.13 could
be used. A thrust model might be just a linear interpolation of the thrust stand value and the
liftoff value versus speed. The AFFTC thrust stand is grossly underutilized for this purpose.
6.8.1 The Reciprocating Engine at Altitude
For the internal combustion engine, the power output for any given engine speed varies
with air density (for nonsupercharged engines). Using the density ratio (σ ) as the density
parameter, the thrust horsepower equation as a function of altitude becomes:
( ) THP BHP η σ · ⋅ ⋅ (6.13)
Richard Von Mises in Theory of Flight suggests that some experimental data indicates that
the σ factor would have an exponent ( n ) greater than 1. One particular set of data gave a
value of 1.29. Then, for that particular set of data, equation 6.13 becomes equation 6.14.
56
( ) ( )
1.29 n
THP BHP BHP η σ η σ · ⋅ ⋅ · ⋅ ⋅ (6.14)
For instance, for an engine at 20,000 feet pressure altitude on a standard day:
a. 0.4595 δ · ,
b. 0.8625 θ · ,
c. 0.5328
δ
σ
θ
· · ,
d.
1.29
0.4438 σ · , and
e.
1.29
0.833
σ
σ
· .
Hence, the altitude degradation factor for this engine is 16.7 percent greater than what
would be predicted by a straight density ratio factor.
57
7.0 FLIGHT PATH ACCELERATIONS
7.1 AirspeedAltitude Method
The classical method of determining the aircraft flight path acceleration is to differentiate
airspeed and altitude using the energy altitude relationship, as developed in the axis systems
and equations of motion section, with a temperature correction to the pressure altitude.
( )
2
0
2
t
E
V
H H
g
· +
⋅
(7.1)
0
t
E C t s
STD
V T
H H V P
T g
 `  `
· ⋅ + ⋅ ·
. , . ,
! ! !
(7.2)
s
x
t
P
N
V
· (7.3)
where:
E
H = energy altitude (feet),
H = geopotential altitude (feet),
t
V = true airspeed (feet/sec),
0
g = acceleration of gravity (32.174 feet/sec²),
x
N = longitudinal load factor in the flight path (or wind) axis, and
s
P = specific excess power (feet/sec).
Note: In this handbook,
x
N and
z
N are the symbology used to denote flight path axis
longitudinal and normal load factor, respectively. One can find other sources that use
symbology of
w
x
N and
w
z
N ( w for wind) or
f
x
N and
f
z
N ( f for flight path). In addition,
many textbooks (including those listed in the Bibliography) will use simply N for flight path
normal load factor.
Now, we can compute the excess thrust (
ex
F ). Excess thrust is the amount of the net thrust
that is more than the amount needed to achieve equilibrium between net thrust and the drag of
the aircraft.
ex x t
F N W · ⋅ (7.4)
Even if you had zero errors in measured airspeed and altitude, the airspeedaltitude method
would have a weakness. That weakness is the presence of winds. You desire to determine the
actual physical acceleration of the aircraft. By taking derivatives of airspeed, you will
invariably have some derivative of wind included. Hence, it becomes desirable to obtain the
aircraft flight path acceleration by some means other than derivatives of true airspeed and
pressure altitude. The GPS yields an alternative method.
58
7.2 GPS Method
A GPS unit will typically provide groundspeed (
g
V ), track angle (
g
σ ), and altitude ( h ).
The groundspeed is the horizontal component of the GPS speed. The parameter h
!
is the GPS
vertical velocity. One could simply use the same equations as for the airspeedaltitude method.
One catch is the track angle is not the same as the aircraft heading angle (ψ ), due again to the
wind. If one had the additional parameter of heading angle (and assuming zero sideslip)
available, then a flight path groundspeed (
gf
V ) could be computed as follows:
cos( )
gf g g
V V σ ψ · ⋅ − (7.5)
However, the above speed is the horizontal component of flight path inertial speed so a
transformation is required.
2 2
f gf
V V h · +
!
(7.6)
Then, just simply insert the appropriate GPSderived accelerations into the airspeedaltitude
equations.
An alternative to using a heading angle, which may not be an available parameter on some
projects, is to perform a cloverleaf maneuver prior to the test maneuver to derive the winds.
The cloverleaf maneuver is described in the airspeed calibration section. This would be
appropriate for constant altitude maneuvers such as accels and turns. Once the two components
of wind (north and east) are determined, one can compute the groundspeed in the wind axis.
The formula is as follows:
( ) ( )
2 2
gf gN wN gE wE
V V V V V · + + + (7.7)
7.3 Accelerometer Methods
There are three different accelerometer methods used to measure flight path acceleration.
These use either the body axis accelerometer (BAA), the flight path accelerometer (FPA), or
an INS. The BAA uses a set of accelerometers placed somewhere within the body of the
aircraft. Ideally, the accelerometers should be at the center of gravity (cg) of the aircraft.
Nevertheless, practically, the BAA is usually in an instrumentation bay away from the cg. The
accelerometers are then subjected to body axis rates and corrections need to be made to
subtract out rate effects. At the time of this writing, the INS has been the primary
accelerometer method used at the AFFTC. NASA Dryden Flight Research Center, however,
uses the BAA method as its primary method.
7.4 Flight Path Accelerometer Method
The FPA consists of a twoaxis accelerometer that is aligned with an angleofattack vane.
The angleofattack vane is connected to a nose boom. The longitudinal axis yields the local
longitudinal acceleration and the normal axis the local normal acceleration. Corrections need to
59
be made to the accelerations for not being at the cg (rate effects) and for being connected to an
angleofattack vane that is not indicating the true angle of attack.
The flight path accelerometer correction equations (ignoring roll and yaw terms) are as
follows:
2
0
cos( ) sin( ) / cos( ) sin( )
i i
x x z V t t
N N N L g q q α α α α ] · ⋅ ∆ − ⋅ ∆ + ⋅ ⋅ − ⋅
]
! (7.8)
2
0
cos( ) sin( ) / sin( ) cos( )
i i
z z x V t t
N N N L g q q α α α α ] · ⋅ ∆ + ⋅ ∆ + ⋅ ⋅ − ⋅
]
! (7.9)
t i bb
α α α α · + ∆ + ∆ (7.10)
α
i
= measured angle of attack
q u lag
α α α α ∆ · ∆ + ∆ + ∆ (7.11)
( )
1
tan
sin( )
V
q
t V t
L q
V L q
α
α
−
]
⋅
∆ ·
]
− ⋅ ⋅
]
]
= pitch rate correction (7.12)
u
α ∆ · upwash correction (7.13)
bb
α ∆ · boom bending correction (7.14)
lag
α ∆ · lag correction (7.15)
where:
q = pitch rate,
V
L = distance from accelerometer to aircraft cg (positive with the accelerometer forward
of the aircraft cg),
t
V = true airspeed,
i
x
N = indicated longitudinal load factor, and
i
z
N = indicated normal load factor.
Figure 7.1 represents an FPA unit (designated an NBIU [Nose Boom Instrumentation
Unit]) developed at the AFFTC in the late 1960s.
60
Figure 7.1 Air Force Flight Test Center Nose Boom Instrumentation Unit
This unit is installed on the AFFTC F15B Pacer (at the time of this writing). Similar units
are still being used for flight test in the late 1990s.
7.5 Accelerometer Noise
When we use an accelerometer to measure flight path accelerations, we must deal with the
noise in that data. No matter where one locates an accelerometer in the aircraft, it will be
subject to substantial quantities of noise. The noise is from structural vibration at relatively
high frequencies and lower frequency flight dynamic oscillations. Figure 7.2 is an example of
some actual data from the first flight of the B1A in December 1974. The data point was a
stabilized cruise point. Figures 7.2 and 7.3 represent indicated longitudinal load factor (
xi
N )
61
and normal load factor (
zi
N ). The accelerometers were located in an AFFTC NBIU. The data
were sampled at 64 samples per second. The analog output of the accelerometers was filtered.
This filter was a 4pole 30 Hz (cycles per second), lowpass Butterworth filter. It is called low
pass because it passes low frequencies. The 30 Hz is the cutoff frequency of the filter. In this
case, the cutoff frequency was too high. On the B1A, the lowest longitudinal vibration modes
were less than 10 Hz. This meant that our performance data had a substantial amount of
longitudinal vibration data in it. After the plots is a discussion of the characteristics of this
filter.
B1A First Flight Data: Flightpath Accelerometer: Indicated Nz
0.4
0.6
0.8
1
1.2
1.4
1.6
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Time (sec)
I
n
d
i
c
a
t
e
d
N
o
r
m
a
l
L
o
a
d
F
a
c
t
o
r
(
N
z
)
Figure 7.2 Longitudinal Load Factor – Unfiltered Data
The mean and standard deviation (sigma) of
xi
N are as follows for 58 data points.
a. Mean = 0.00831
b. Sigma = 0.01682
62
B1A First Flight Data: Flightpath Accelerometer: Indicated Nx
0.040
0.030
0.020
0.010
0.000
0.010
0.020
0.030
0.040
0.050
0.060
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Time (sec)
I
n
d
i
c
a
t
e
d
L
o
n
g
i
t
u
d
i
n
a
l
L
o
a
d
F
a
c
t
o
r
(
N
x
)
Figure 7.3 Normal Load Factor – Unfiltered Data
The mean and standard deviation for the
zi
N is as follows for the same 58 time slices:
a. Mean = 1.0047
b. Sigma = 0.2257
Ignoring pitch rate terms, the transformation equation for true flight path longitudinal load
factor (
x
N ) is as follows:
cos sin
x xi zi
N N N α α · ⋅ ∆ − ⋅ ∆ (7.16)
where:
α ∆ = upwash angle.
If
x
N was zero for this stabilized cruise point, then the above equation can be used to
solve for upwash.
1
tan
xi
zi
N
N
α
−  `
∆ ·
. ,
(7.17)
For this one data sample, the α ∆ computes to be:
( )
1
0.00831
tan 0.47 deg
1.0047
α
−
∆ · ·
63
The attenuation of a filter is expressed in terms of decibel (dB). The definition of decibel is
as follows:
0
10
20 log
i
E
dB
E
 `
· − ⋅
. ,
(7.18)
where:
0
E = output, and
i
E = input.
By definition, the cutoff frequency is at a 3.0 dB · , which is an output over input of 0.708
or an attenuation of almost 30 percent. Figure 7.4 shows the attenuation for a fourpole
Butterworth filter.
FourPole Butterworth LowPass Filter Attenuation
0
10
20
30
40
50
60
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Frequency Ratio (frequency/cutoff frequency)
T
r
a
n
s
m
i
s
s
i
o
n
L
o
s
s

N
e
g
a
t
i
v
e
D
e
c
i
b
e
l
s
Figure 7.4 FourPole Butterworth Filter Attenuation Characteristics
At the time, the solution to the noise problem with B1A flight path accelerometer data was
to change to filters with a much lower cutoff frequency. The problem with that solution was
that a filter with a low cutoff frequency also introduced substantial phase (time) lag. For this
filter, Figure 7.5 represents the time lag function versus the frequency ratio. The time delay is
defined in terms of a parameter called the group time delay (
dgroup
t ). The actual time delay
( t ∆ ) is determined as follows:
2
dgroup
c
t
t
f π
 `
∆ ·
⋅ ⋅
. ,
(7.19)
64
where:
c
f is the cutoff frequency in Hz.
FourPole Butterworth LowPass Filter Group Time Delay
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
0 1 2 3 4 5 6
Frequency ratio (frequency/cutoff frequency)
G
r
o
u
p
t
i
m
e
d
e
l
a
y
(
s
e
c
)
Figure 7.5 FourPole Butterworth Filter Group Time Delay
At maneuver frequencies less than 0.1 times the cutoff frequency, the group time delay is
2.60 seconds. A filter with a cutoff frequency of 2.0 was selected to avoid the very low
frequency firstbody bending modes of this very flexible aircraft. Since no dynamic
performance maneuvers were performed on the B1A, this was not deemed a problem.
The actual time delays for the 30 and 2.0 Hz filters compute to the following using the
above equation.
a. 0.014 sec for 30
c
t f Hz ∆ · ·
b. 0.207 sec for 2.0
c
t f Hz ∆ · ·
A time lag of 0.2 second can be a source of significant errors for highly dynamic
maneuvers such as the roller coaster. To avoid a time shift error in accelerometer data, it would
be more desirable to digitally filter the data. To illustrate this, the
xi
N was digitally filtered
with two different methods. A span of 21 data points was chosen which would include the
midpoint and 10 points on each side of the midvalue. The first was a moving secondorder
polynomial curve fit. The second was a moving average. These are shown in Figure 7.6.
65
Indicated Nx data: Digitally Filtered: 21 Point Span
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (sec)
I
n
d
i
c
a
t
e
d
L
o
n
g
i
t
u
d
i
n
a
l
L
o
a
d
F
a
c
t
o
r
(
N
i
)
Second Order Polynomial
Moving Average
Figure 7.6 Longitudinal Load Factor – Filtered Data
Figure 7.7 plots the moving secondorder polynomial fit points. A thirdorder polynomial
curve fit of the time history is also shown.
Indicated Longitudinal Load Factor
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (sec)
L
o
n
g
i
t
u
d
i
n
a
l
L
o
a
d
F
a
c
t
o
r
(
N
x
i
)
Moving Polynomial
Poly. (Moving Polynomial)
Figure 7.7 ThirdOrder Polynomial Fit of Filtered Longitudinal Load Factor Data
66
Table 7.1 summarizes the mean values and 1sigma deviations from the mean for the
different sets of data.
Table 7.1
SUMMARY OF STATISTICS FOR LONGITUDINAL LOAD FACTOR
Original
Data
Moving
Average
SecondOrder
Polynomial Moving
SecongOrder Moving
Minus ThirdOrder Fit
Mean 0.00831 0.00853 0.00848 0
1Sigma 0.01682 0.00115 0.00233 0.00140
The average value of each of the three methods was identical to three digits (1 millig). The
two digital filtering methods reduced the standard deviation by about a factor of 10. Although
(for this data set) the simple moving average produced the greatest reduction in standard
deviation, it is preferable to use the moving secondorder polynomial fit. That is because for
any maneuver where variation in acceleration is not linear, the parabola will match the
variation more accurately.
7.6 Inertial Measurement Method
The INS method involves transforming the earth axis inertial parameters of the INS into
the aircraft wind (or flight path) axis. Typically, the INS outputs will be velocities and
accelerations in the north, east, and down direction and a set of angles called Euler angles. The
Euler angles are pitch, roll, and true heading. The mathematics below will take you through the
process to compute winds. Once the winds are known, then the transformations into the wind
axis are performed.
Define:
a. θ = pitch attitude,
b. φ = roll attitude,
c. ψ = true heading angle,
d. α = angle of attack, and
e. β = sideslip angle.
7.7 Calculating Alpha, Beta and True Airspeed
The following matrices are used to transform the true airspeed from the flight path axis
(
t
V ) to the earth axis (
tN
V ,
tE
V , and
tD
V ). The transformation must be performed in the exact
order of , , , , β α φ θ ψ .
Heading (rotate about the z axis [or yaw]) (transform through ψ )
67
[ ]
cos sin 0
sin cos 0
0 0 1
ψ ψ
ψ ψ ψ
− ]
]
·
]
]
]
(7.20)
Pitch (rotate about yaxis) (transform through θ )
[ ]
cos 0 sin
0 1 0
sin 0 cos
θ θ
θ
θ θ
]
]
·
]
] −
]
(7.21)
Roll (rotate about xaxis) (transform through φ )
[ ]
1 0 0
0 cos sin
0 sin cos
φ φ φ
φ φ
]
]
· −
]
]
]
(7.22)
Angle of attack (transform through α )
[ ]
cos 0 sin
0 1 0
sin 0 cos
α α
α
α α
− ]
]
·
]
]
]
(7.23)
Sideslip angle (transform through β )
[ ]
cos sin 0
sin cos 0
0 0 1
β β
β β β
− ]
]
·
]
]
]
(7.24)
The matrix summary form of the transformation from the flight path axis true airspeed to
the true airspeed in the earth axis ( N , E , D) is as follows:
( )
( )
( )
[ ] [ ] [ ] [ ] [ ]
0
0
gN wN
t
gE wE
gD wD
V V
V
V V
V V
ψ θ φ α β
¹ ¹
+
¹ ¹
¹ ¹
¹ ¹ ¹ ¹
+ · ⋅ ⋅ ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹
+
¹ ¹
¹ ¹
(7.25)
From equation 7.25 we can solve for the winds.
68
[ ] [ ] [ ] [ ] [ ]
0
0
wN t gN
wE gE
wD gD
V V V
V V
V V
ψ θ φ α β
¹ ¹
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅ ⋅ −
' ' ' ' ' '
¹ ¹ ¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
¹ ¹
(7.26)
The equation above is the general matrix formula. During a typical wind calibration, we
will assume the vertical wind (
wD
V ), the sideslip angle ( β ), and the bank angle (φ) are equal to
zero. Equation 7.26 represents three equations with at least five unknowns. The five unknowns
are the three components of wind ( ) ,
wN wE wD
V V and V and α and β .
Then the α calculation reduces to the following:
α θ γ · − (7.27)
1
sin
t
h
V
γ
−  `
· ·
. ,
!
flight path angle (7.28)
gD
h V · − ·
!
rate of climb (7.29)
We now wish to perform the reverse transformation; that is, to transform the components
of true airspeed in the earth axis to the flight path. To transform the components, reverse the
order of the matrix multiplication and take the transpose of each individual matrix. In this case,
the transpose is the same as the inverse. To take the transpose of these unique matrices reverse
all the offdiagonal terms and keep all the diagonal terms the same. For instance, the
[ ]
T
β
matrix derives from equation 7.24 as follows:
[ ]
cos sin 0 cos sin 0
sin cos 0 sin cos 0
0 0 1 0 0 1
T
T
β β β β
β β β β β
− ] ]
] ]
· · −
] ]
] ]
] ]
(7.30)
The matrix formula is as follows:
[ ] [ ] [ ] [ ] [ ]
0
0
tN t
T T T T T
tE
tD
V V
V
V
β α φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
⋅ ⋅ ⋅ ⋅ ⋅ ·
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
(7.31)
We can calculate all the velocities in the equation 7.31 using the winds determined during
the wind calibration (equation 7.26) as follows:
tN gN wN
V V V · + (7.32)
tE gE wE
V V V · + (7.33)
69
tD gD wD
V V V · + (7.34)
( )
2 2 2
t tN tE tD
V V V V · + + (7.35)
The airspeed components in the body axis ( , , x y z ) are calculated in the following matrix
manner:
[ ] [ ] [ ]
bx tN
T T T
by tE
bz tD
V V
V V
V V
φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
(7.36)
Next, transform the body axis to the flight path axis through angle of attack and sideslip
angle as follows:
[ ] [ ]
0
0
bx t
T T
by
bz
V V
V
V
β α
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
⋅ ⋅ ·
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
(7.37)
Expanding the alpha and beta transpose matrices and writing them out:
cos sin 0 cos 0 sin
sin cos 0 0 1 0 0
0 0 1 sin 0 cos 0
bx t
by
bz
V V
V
V
β β α α
β β
α α
] ] ¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
] ]
− ⋅ ⋅ ·
' ' ' '
] ]
¹ ¹ ¹ ¹
] ] −
] ] ¹ ¹ ¹ ¹
(7.38)
cos cos sin cos sin
sin cos cos sin sin 0
sin 0 cos 0
bx t
by
bz
V V
V
V
β α β β α
β α β β α
α α
⋅ ⋅ ] ¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
]
− ⋅ − ⋅ ⋅ ·
' ' ' '
]
¹ ¹ ¹ ¹
] −
] ¹ ¹ ¹ ¹
(7.39)
Multiplying out the above matrix yields three equations from which we will derive
formulas for α and β . When complete, these formulas should be the same as presented
earlier. In the axis systems and equations of motion section, the angles were derived by
geometry without the following matrix mathematics:
cos cos sin cos sin
bx by bz t
V V V V β α β β α ⋅ ⋅ + ⋅ + ⋅ ⋅ · (7.40)
sin cos cos sin sin 0
bx by bz
V V V β α β β α − ⋅ ⋅ + ⋅ − ⋅ ⋅ · (7.41)
sin cos 0
bx bz
V V α α − ⋅ + ⋅ · (7.42)
Equation 7.42 yields a formula for angle of attack.
70
sin / cos tan
bz
bx
V
V
α α α · · (7.43)
1
tan
bz
bx
V
V
α
−  `
·
. ,
(7.44)
Inserting the result for
bx
V from equation 7.44 into equation 7.40:
cos
sin
bx bz
V V
α
α
· ⋅
2 2
cos sin
cos sin cos
sin sin
bz by bz t
V V V V
α α
β β β
α α
⋅ ⋅ + ⋅ + ⋅ ⋅ · (7.45)
Collecting terms and using the trigonometric identity
2 2
sin cos 1 α α + · :
cos sin
sin
bz
by t
V
V V β β
α
]
⋅ + ⋅ ·
]
]
(7.46)
Now, we will use equations 7.41 and 7.42 to substitute for the term in the square brackets.
Replace
bx
V in 7.41 using 7.42.
2 2
cos sin
sin cos sin 0
sin sin
bz by bz
V V V
α α
β β β
α α
− ⋅ ⋅ + ⋅ − ⋅ ⋅ ·
( )
2 2
cos sin
sin cos 0
sin
bz by
V V
α α
β β
α
]
+
] − ⋅ ⋅ + ⋅ ·
]
]
cos
sin sin
bz
by
V
V
β
α β
]
· ⋅
]
]
(7.47)
Finally, substituting equation 7.47 into equation 7.46:
2
cos sin
cos
sin sin
by by t
V V V
β β
β
β β
⋅ ⋅ + ⋅ ·
sin
by
t
V
V
β
· (7.48)
1
sin
by
t
V
V
β
−
 `
·
. ,
Compare equations 7.44 and 7.48 to equations 2.11 and 2.12.
We now wish to perform the reverse transformation; that is, to transform the components
of true airspeed in the Earth axis to the flight path. To transform the components, reverse the
71
order of the matrix multiplication and take the transpose of each individual matrix. The matrix
formula is as follows:
[ ] [ ] [ ] [ ] [ ]
0
0
tN t
T T T T T
tE
tD
V V
V
V
β α φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
⋅ ⋅ ⋅ ⋅ ⋅ ·
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
(7.49)
We can readily solve for the true airspeed components from the above.
The airspeed components in the body axis ( , , x y z ) are calculated in the following matrix
manner:
[ ] [ ] [ ]
bx tN
T T T
by tE
bz tD
V V
V V
V V
φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
(7.50)
From true airspeed and the body axis true airspeed components, angle of attack and
sideslip are computed using equations 7.44 and 7.48. The α and β are required in order to
transform the earth axis accelerations to the flight path axis.
7.8 Flight Path Accelerations
To compute the accelerations in the flight path requires first computing the accelerations in
the N, E, and D axis. Even when the accelerations are available as a direct output of an INS, it
is desirable to compute the accelerations by taking numerical derivatives of the inertial
velocities. This is because the accelerations are sensing the high frequency vibrations of the
aircraft and are usually quite noisy. The typical INS updates at 50 samples per second. If one
simply samples the velocity data at no more than about 5 samples per second and then takes a
derivative, the noise will be dramatically reduced. The acceleration formulas are as follows:
( ) ( )
( )
2
gN gN
N
V t t V t t
A t
t
+ ∆ − − ∆
·
⋅ ∆
(7.51)
( ) ( )
( )
2
gE gE
E
V t t V t t
A t
t
+ ∆ − − ∆
·
⋅ ∆
(7.52)
0
( ) ( )
( )
2
gD gD
D
V t t V t t
A t g
t
+ ∆ − −∆
· −
⋅ ∆
(7.53)
The velocities in the equations 7.51 through 7.53 are the inertial (or ground) speeds, not the
airspeeds. We are computing inertial accelerations in the N, E, and D axis. However, we will
later transform these into the wind axis. They are still inertial accelerations, but the
components in our wind axis system. Note that the down (or z ) component involves
subtracting out a gravity term. Since the vertical component of acceleration is down, we are
72
actually adding in a gravity term. For instance, at 5 samples per second, the t ∆ would be
0.20 seconds.
The transformation matrix formulation for accelerations is identical to that for velocities
and is given below. However, we will put the flight path accelerations on the left side of the
equation.
[ ] [ ] [ ] [ ] [ ]
xf N
T T T T T
yf E
zf D
A A
A A
A A
β α φ θ ψ
¹ ¹
¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹
¹ ¹
(7.54)
In performance, we normally work with load factors (acceleration over g) rather than the
accelerations. In addition, in conventional performance the standard sea level value of g (
0
g =
32.174 feet/sec
2
) is usually used. There is also a sign change on the normal load factor to
account for the positive normal load factor convention.
0
0
0
x xf
y yf
z zf
N A g
N A g
N A g
¹ ¹
¹ ¹
¹ ¹ ¹ ¹
·
' ' ' '
¹ ¹ ¹ ¹
−
¹ ¹
¹ ¹
(7.55)
Finally, note that
f
designation is dropped for the flight path axis load factors.
7.9 Accelerometer Rate Corrections
The following corrections to accelerometers are presented without derivation. Assume we
have rate gyros, which give us roll rate, pitch rate, and yaw rate in the body axis. Define these
as follows:
a. ( ) ( ) roll rate rotation about axis right wing down p x · − + ;
b. ( ) ( ) pitch rate rotation about axis pitch up q y · − + ; and
c. ( ) yaw rate (rotation about axis) nose right r z · − + .
Assume that the accelerometers are at distances ,
x y z
l l and l from the cg of the aircraft.
The x distance (
x
l ) is positive forward, y distance (
y
l ) is positive out the right wing, and the z
distance (
z
l ) is positive down. If the noncorrected body axis accelerations are designated with
a sub i designation, then the matrix correction equations are as follows:
73
( ) ( ) ( )
( ) ( ) ( )
( ) ( ) ( )
2 2
2 2
2 2
i
i
i
xb xb x
yb yb y
zb zb z
q r r p q q p r
A A l
A A r p q p r p q r l
A A l
q r p p q r q p
]
+ − ⋅ − + ⋅
¹ ¹
¹ ¹ ¹ ¹
]
¹ ¹ ¹ ¹ ¹ ¹
]
· + − + ⋅ + − ⋅ ⋅
' ' ' ' ' '
]
¹ ¹ ¹ ¹ ¹ ¹
] ¹ ¹ ¹ ¹
¹ ¹ − − ⋅ + ⋅ − +
]
! !
! !
! !
(7.56)
Note: A sign change when computing normal load factor.
a.
0
zb
zb
A
N
g
· −
This author prefers to rate correct the velocities, then take numerical derivatives to
compute accelerations. Then, one would not rate correct the resultant accelerations.
7.10 Velocity Rate Corrections
Rate corrections to the body axis velocities in the matrix format are presented in equation
7.57. These will have been accomplished by axis transformations through , and ψ θ φ , in that
order. Again, the i designation will be noncorrected velocities.
0
0
0
i
i
i
bx bx x
by by y
bz bz z
V V r q l
V V r p l
V V q p l
¹ ¹ − ¹ ¹ ] ¹ ¹
¹ ¹ ¹ ¹ ¹ ¹
]
· + − ⋅
' ' ' ' ' '
]
¹ ¹ ¹ ¹ ¹ ¹
] −
¹ ¹ ] ¹ ¹
¹ ¹
(7.57)
7.11 Calculating p, q, and r
In the case where the Euler angles ( , , ψ θ φ ) are given, we can compute the body axis rates
using the following formulas.
sin p φ ψ θ · − ⋅
!
! (7.58)
cos cos sin q θ φ ψ θ φ · ⋅ + ⋅ ⋅
!
! (7.59)
cos cos sin r ψ θ φ θ φ · ⋅ ⋅ − ⋅
!
! (7.60)
7.12 Euler Angle Diagram
Figure 7.8 illustrates the Euler angles. This Euler angle diagram pictorially illustrates the
order of transformation. Starting with the aircraft heading north, a transformation is performed
(positive east) through the heading angle (ψ ). Then, the aircraft is pitched (positive up)
through the pitch attitude (θ ). Finally, the aircraft is rotated (positive right wing down)
through the roll angle (φ ). It is critical that the order of rotation is just as described ( , , ψ θ φ ),
otherwise, one would get a different result.
74
Figure 7.8 Euler Angles
8.0 TAKEOFF
8.1 General
This section will present the theory of takeoff and landing for conventional aircraft. For this
handbook, conventional aircraft would be any aircraft with a main gear, a nose gear, and a single
source of thrust at some angle of incidence
t
i . Therefore, ‘conventional’ could include some
aircraft that are considered STOL (Short TakeOff and Landing). One could derive equations that
are more complex for a VSTOL (Vertical or Short TakeOff and Landing).
8.2 Takeoff Parameters
Let us define the following forces, distances, angles and coefficients as depicted in Figure
8.1. (Not shown on the drawing [to avoid clutter] are gross thrust [
g
F ] and the engine inlet [or
propulsive] drag [
e
F ]).
a.
bw
D = drag of the aircraft body and wing  along the aircraft flight path axis. During the
ground roll, the flight path will be parallel to the runway.
b.
t
D = drag of the aircraft tail  acts along the aircraft flight path (this term is often lumped
into the body drag for aircraft without a Ttail).
c.
1
L = lift of the wing  acts perpendicular to the flight path.
d.
2
L = lift of the tail  also acts perpendicular to the flight path.
e.
t
W = gross weight  acts through the center of gravity of the aircraft.
75
f.
n
F
= net thrust acting parallel to the flight path.
g.
1
F = load on the nose gear (perpendicular to the runway).
h.
2
F = load on the main gear (perpendicular to the runway).
i.
1
X = distance from the nose gear to the aircraft center of gravity.
j.
2
X
= distance from the main gear to the aircraft center of gravity.
k.
1
XL = distance from the center of gravity to action point of the wing lift (aerodynamic
center of the MAC [Mean Aerodynamic Chord]).
l.
2
XL
= distance from the wing lift point to the tail lift action point.
m.
1
Z
= height of the body axis of the aircraft above the ground plane.
n.
2
Z = height of the tail center of lift and drag above the aircraft body axis.
o. θ = aircraft pitch attitude (angle between Xbody axis and horizontal).
p.
rw
θ = runway slope.
Figure 8.1 Takeoff and Landing Forces and Angles
Using the above diagram, we can formulate the equations of motion for the aircraft during the
ground roll. The equations are the same for either a takeoff or a landing.
Requiring the summation of forces in the Xaxis to be zero:
cos( )
g t e rw ex
F i F D F F θ ⋅ + − · + + (8.1)
76
where:
D = total aerodynamic drag,
rw
F
= total runway resistance
= runway friction plus runway slope effect, and
ex
F = excess thrust (positive forward).
bw t
D D D · + (8.2)
1 1 2 2
sin( )
rw t rw
F F F W µ µ θ · ⋅ + ⋅ + ⋅ (8.3)
where:
1
µ = coefficient of friction associated with the nosewheels, and
2
µ
= coefficient of friction associated with the main wheels.
ex x t
F N W · ⋅ (positive forward) (8.4)
where:
x
N
= longitudinal load factor.
0
/
x x
N A g · (8.5)
x g
A V ·
!
(8.6)
where:
g
V = groundspeed.
Note that the longitudinal load factor definition on the ground includes only the velocity
derivative term. In the air, the gravity component is included. On the ground, we will account for
the gravity component in the sin( )
t rw
W θ ⋅ term.
Collecting terms:
1 1 2 2
cos( ) ( ) ( sin( ))
g t e bw t t rw ex
F i F D D F F W F θ µ µ θ ⋅ + − · + + ⋅ + ⋅ + ⋅ + (8.7)
Requiring the summation of forces in the Zaxis to be zero:
1 2 1 2
cos( )
t rw
L L F F W θ + + + · ⋅ (8.8)
Require the summation of moments about the Yaxis to be zero. Take moments about the
main wheels, since the aircraft will pitch about the main wheels during the takeoff or landing
ground roll. Ignore any pitch dynamics during the ground roll or any moment caused by the
vertical component of gross thrust.
77
( ) ( ) ( )
1 1 2 1 2 1 1 1 2 1
sin( )
bw t t rw
F X X L X XL D Z D Z Z W Z θ ⋅ + + ⋅ − + ⋅ + ⋅ + + ⋅ ⋅ ·
( ) ( )
2 1 2 1 2 2
cos( ) cos( )
t rw g t e
W X F i F Z L XL XL X θ ⋅ ⋅ + ⋅ − ⋅ + ⋅ + − (8.9)
What we now have is three equations with three unknowns for purposes of simulating a
takeoff or landing ground roll. It is assumed that one has a thrust and drag model for the lift,
drag, gross thrust, and propulsive drag terms in the above equations. However, the lift and drag
models may not be for inground effect. If no inground effect corrections are available, then
some empirical predictions can be used until flight test results are available to create an inground
effect model.
The three unknowns are the two normal forces on the wheels (
1
F and
2
F ) and the excess
thrust (
ex
F ). The primary parameter of interest is the excess thrust from which we can compute
the derivative of groundspeed. Once we have the excess thrust, we can integrate the groundspeed
derivative to obtain speed and distance versus time.
Collecting equations 8.7 through 8.9:
1 1 2 2
cos( ) sin( )
g t e bw t t rw ex
F i F D D F F W F θ µ µ θ ⋅ + − · + + ⋅ + ⋅ + ⋅ +
1 2 1 2
cos( )
t rw
L L F F W θ + + + · ⋅
( ) ( ) ( ) ( )
1 1 2 1 2 1 1 2 1
sin
t t rw
F X X L X XL D Z Z W Z θ ⋅ + + ⋅ − + ⋅ + + ⋅ ⋅ ·
( ) ( )
2 1 2 1 2 2
cos( ) cos( )
t rw g t e
W X F i F Z L XL XL X θ ⋅ ⋅ + ⋅ − ⋅ + ⋅ + −
Rearranging the equations:
1 1 2 2
cos( ) sin( )
ex g t e bw t t rw
F F F F i F D D W µ µ θ ] + ⋅ + ⋅ · ⋅ − − − − ⋅
]
(8.10)
[ ]
1 2 1 2
cos( )
t rw
F F W L L θ + · ⋅ − − (8.11)
( )
1 2 1
X X F + ⋅ ·
( ) ( )
( ) ( )
2 1 1 2 1 2 2
1 2 1 1 2
cos( ) sin( ) cos( )
t rw t rw g t e
t
W X W Z F i F Z L XL XL X
L X XL D Z Z
θ θ θ
]
⋅ ⋅ − ⋅ ⋅ + ⋅ + − ⋅ + ⋅ + −
]
] − ⋅ − − ⋅ +
]
(8.12)
We will define the terms in the square brackets in 8.10 through 8.12 as
1
A ,
2
A , and
3
A .
Then we can rewrite equations 8.10 through 8.12 in three by threematrix form as follows:
( )
1 2 1
1 2
1 2 2 3
1
0 1 1
0 0
ex
F A
F A
X X F A
µ µ ]
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹ ]
⋅ ·
' ' ' '
]
¹ ¹ ¹ ¹
]
+
¹ ¹ ¹ ¹ ]
(8.13)
78
During the course of flight test, we measure excess thrust (
ex
F ). However, the thrust and
drag may be unknown, or at least not known precisely. Therefore, we may need to iterate
between the above equation and the solution of the above equation. The
1
A term is thrust minus
drag minus the runway component of weight.
The matrix relationship in equation 8.13 can be solved by multiplying both sides by the
inverse of the square matrix.
( )
1
1 2 1
1 2
2 1 2 3
1
0 1 1
0 0
ex
F A
F A
F X X A
µ µ
−
]
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹ ]
· ⋅
' ' ' '
]
¹ ¹ ¹ ¹
]
+
¹ ¹ ¹ ¹
]
(8.14)
8.3 Developing a Takeoff Simulation
Usually, the contractor will provide an initial estimated model for lift and drag as a function
of angle of attack (α ). As mentioned before, one may need to supplement this model with
empirical ground effect estimation, such as that found in the NASA takeoff and landing
simulation program listed in the Bibliography. During the ground roll, the angle of attack is equal
to the pitch attitude (α θ · ). The thrust incidence angle is usually zero or small.
Only the most precise simulations will typically account for a separate tail and body drag, so
we can ignore
t
D in many cases. Accounting for tail lift and drag becomes more important for
modeling braking performance to determine the load distribution on the main gear and the nose
gear. For takeoff performance, a value of 0.015 is usually assumed for the rolling coefficient of
friction ( µ ). Values of µ for a dry runway up to 0.025 are also used. In addition, a point mass
model will be assumed with all the forces acting through the cg of the aircraft. Further, since
g e
F F >> at low airspeeds, we make the following approximation:
( )
cos( )
n g e t
F F F i θ ≅ − ⋅ + (8.15)
sin( )
ex n t rw
F F F D W µ θ + ⋅ · − − ⋅ (8.16)
cos( )
t rw
F W L θ · ⋅ − (8.17)
where:
F ·main gear load (assume all load on the main gear).
Combining equations 8.16 and 8.17:
( ) cos( ) sin( )
ex t rw n t rw
F W L F D W µ θ θ + ⋅ ⋅ − · − − ⋅ (8.18)
79
Equation 8.18 can be used in two ways. First, to solve excess thrust (equation 8.19). Second,
to solve thrust minus drag (equation 8.20). We know (or assume values for) the other variables:
gross weight, runway slope, rolling friction, and aerodynamic lift.
[ ] ( ) sin( ) cos( )
ex n t rw t rw
F F D W W L θ µ θ · − − ⋅ − ⋅ ⋅ − (8.19)
[ ] ( ) sin( ) cos( )
n ex t rw t rw
F D F W W L θ µ θ − · + ⋅ + ⋅ ⋅ − (8.20)
From equation 8.19, we can compute the excess thrust during the ground roll of the aircraft.
One would be provided models for net thrust drag and lift. The drag and lift models would be in
the form of drag and lift coefficients versus angle of attack. Typical model formulations are as
follows:
( ) , ,
n C
F f M H T · (8.21)
( ) ,
L AGL
C f h α · (8.22)
( ) ,
D L AGL
C f C h · (8.23)
where:
M = Mach number,
C
H = pressure altitude (subscript C denotes calibrated),
T = ambient temperature, and
AGL
h
= aircraft wing height above ground level.
The parameter
AGL
h
is needed to account for ground effect. The above are just typical model
forms. They may also include Reynolds number (or skin friction drag) terms in the drag polar. In
addition, the engine is usually not at 100percent thrust at brake release so a thrust spool up factor
needs to be supplied. One would also incorporate a fuel flow model to compute fuel used during
takeoff. This is to account for the fuel used for mission calculations.
8.4 Ground Effect
Figure 8.2 is typical of a relationship defining the decrease in drag due to lift inground
effect. The data points were taken from a curve found in two separate textbooks, neither of
which gave a source for the data. The texts are The Illustrated Guide to Aerodynamics by H.C.
Smith and Technical Aerodynamics by Karl D. Wood. The suspicion is that this is from some
early NACA work. The equation is a curve fit of the points.
80
Figure 8.2 Predicted Ground Effect Drag
A very simplified model that approximates an F16 aircraft in military thrust was created to
illustrate takeoff simulation. The model constants and equations are as follows:
a. S = 300 = reference wing area (feet
2
).
b. b = 35 = wing span (feet).
c. AR = 4.0 =
2
/ b S = aspect ratio.
d.
w
h = 5.0 = height of wing above ground while aircraft on the ground (feet).
e.
ts
W = 25,000. = start gross weight (pounds).
f.
no
F = 10,000. = thrust at zero Mach number (pounds).
g.
nslope
F = 5,000 = slope of thrust versus Mach number (pounds).
h.
no
F
K = 0.65 = thrust factor at zero time.
i. τ = 2.0 = thrust time constant (seconds).
( )
/
1
no
t
Fn F
K K e
τ −
· − ⋅ (8.24)
81
Thrust runs can be used to determine this thrust spool up factor. It may not be a simple
exponential function as we are using here. For our model, at time = zero, the thrust is 35 percent
of zero Mach number thrust and increases exponentially with a 2.0 second time constant. Then
the equation for the net thrust for this model becomes:
( )
n Fn no nslope
F K F F M · ⋅ + ⋅ (8.25)
f n
W tsfc F · ⋅ (8.26)
where:
tsfc = thrust specific fuel consumption.
A curve fit of the data points in Figure 8.2 was performed to produce an equation for ground
effect.
( )
24.12 108.29 /100
w
GE
h h
X Ln
b
]  ` +
· ⋅ +
]
. , ]
(8.27)
1.0 , 1.0
GE GE
X if X · >
Drag coefficient (
D
C ) is computed as follows:
( )
( )
2
min min
1
D D GE L L
C C X C C
AR π
 `
· + ⋅ −
⋅
. ,
(8.28)
where:
min D
C = 0.0500 = minimum drag coefficient, and
min L
C = 0.05 = lift coefficient corresponding to minimum drag.
Ambient pressure ratio (δ ) is as follows (formula derived in the altitude section):
( )
5.2559
1 6.87559 6
C
E H δ · − − ⋅ (8.29)
where:
C
H = 2,300 feet = initial pressure altitude.
SL
P
P
δ
 `
·
. ,
(8.30)
where:
P = ambient pressure, and
82
SL
P = ambient pressure at standard day sea level = 2116.22 lbs/ft
2
.
Lift coefficient (
L
C ) is as follows (from elliptic wing theory):
0
2
1
L L
AR
C C
AR
π
α
 `
⋅
· + ⋅
]
+
] . ,
(8.31)
As with the drag coefficient, an adjustment for ground effect needs to be applied to the lift
coefficient. A lift coefficient factor inground effect was determined on two separate flight test
projects÷a fighter and a transport÷at the AFFTC. In both cases, the ground effect factor at lift
off was about 30 percent. The above lift and drag models are idealizations presented to illustrate
general trends only. In a flight test project, one would initially use wind tunnel data, and later use
flight test derived models. The formula is as follows:
a.
( )
( )
1.30
L IGE
L OGE
C
C
·
In both cases, the wing height to span ( / h b ) is about 0.20. Let us assume that by the time
/ h b increased to 0.5 (half span), the lift ratio decreased to 1.05 (5 percent). Then, further assume
that the relationship is base 10 logarithmic. That yields Figure 8.3.
Lift Curve Ground Effect Factor
1.0
1.1
1.2
1.3
1.4
1.5
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6
Wing height above ground/wing span (h+hw)/b
C
L
(
I
G
E
)
/
C
L
(
O
G
E
)
Figure 8.3 Lift Ratio InGround Effect
The equation corresponding to the above curve is as follows:
83
( ) ( )
( )
10
( )
0.8609 0.6282
L IGE
w
L OGE
C
Log h h b
C
· − ⋅ + (8.32)
With the following constraint:
a.
( )
( )
1.0
L IGE
L OGE
C
C
≥
The angle of attack is held to zero during the ground roll until a rotation speed is reached.
This rotation speed (in this simulation example) is at a calibrated airspeed of 100 knots.
Calibrated airspeed is normally displayed in the cockpit and was discussed in detail in Section
4.0 Airspeed. As will be shown in the later vectored thrust takeoff section, the selection of 100
knots as the rotation speed is probably much too low for an actual F16. Upon reaching the
rotation speed, the typical takeoff will rotate to some given angle of attack. Then, that angle of
attack is held until the aircraft generates enough lift such that lift is greater than weight and the
aircraft lifts off the runway. The angleofattack profile used in this example computer simulation
is as follows:
( ) last
t
t
α
α α
∆
· + ⋅ ∆
∆
(8.33)
where:
( )
3.0
t
α ∆
·
∆
deg/sec.
The angle of attack (α ) is limited to a predetermined value. In this example simulation that
value is 13 degrees. In the numerical integration, 13 degrees α is reached at 130 knots calibrated
airspeed. The lift first exceeds weight at an airspeed of 132 knots. The aircraft (or the simulated
aircraft) will lift off the ground when lift is greater than weight.
Lift and drag (formulas in lift and drag section) are computed as follows:
2
/ 0.000675
L
L C M S δ · ⋅ ⋅ ⋅ (8.34)
2
/ 0.000675
D
D C M S δ · ⋅ ⋅ ⋅ (8.35)
Finally, the last terms in our model are the runway resistance. We will assume zero runway
slope.
µ = 0.015 rolling coefficient of friction.
Then,
( )
rw t
F W L µ · ⋅ − (8.36)
0.0
rw t
F if L W · >
Combining terms:
84
( )
ex n rw
F F D F · − + (8.37)
ex x t
F N W · ⋅ (8.38)
0
g
x
t
V
h
N
g V
· +
!
!
(8.39)
During the ground roll, the hdot ( h
!
) term is zero. During the air phase, the normal load
factor equation is used. Equation 8.40 is derived in the section on normal load factor during a
climb.
0
cos( )
t
z
V
N
g
γ
γ
⋅
· +
!
(8.40)
1
sin
t
h
V
γ
−  `
·
. ,
!
flight path angle (8.41)
From the ,
x z
N N , and γ equations (8.39 through 8.41), we can numerically integrate
groundspeed (
g
V ) and geometric height ( h ). All of the forces, however, are functions of airspeed
and pressure altitude. We have assumed a standard atmosphere for temperature. Standard
atmosphere is defined in the altitude section.
288.15 (1.9812/1000)
C
T H · − ⋅ (8.42)
t g w
V V V · + (8.43)
where:
t
V = true airspeed, and
w
V = windspeed. We will assume windspeed equals zero.
The following equations were derived in Section 4.0 Airspeed.
/
t
M V a · (8.44)
SL
a a θ · ⋅ = speed of sound (8.45)
where:
661.48
SL
a · knots.
( )
288.15
T
θ · = temperature ratio (8.46)
85
3.5
2
1 0.2 1
C
a
q
M
P
 `
] · + ⋅ −
]
. ,
(8.47)
where:
C
q = compressible dynamic pressure.
( )
( )
{ ¦
1/ 3.5
5 1 1
C SL C SL
V a q P
]
· ⋅ ⋅ + −
]
= calibrated airspeed (8.48)
where:
2116.22
SL
P · (lbs/ft
2
) = ambient pressure at standard sea level.
A plot of thrust, drag plus the runway resistance terms and excess thrust versus calibrated
airspeed, is shown in Figure 8.4.
Takeoff Forces Versus Speed
0
2,000
4,000
6,000
8,000
10,000
12,000
0 20 40 60 80 100 120 140 160 180
Calibrated Airspeed (kts)
F
o
r
c
e
(
l
b
s
)
Drag + Runway Resistance
Excess Thrust
Net Thrust
Figure 8.4 Takeoff Forces
The time history of the simulation is shown in Figure 8.5.
86
Takeoff Simulation Time History
0
20
40
60
80
100
120
140
160
180
200
0 5 10 15 20 25 30
Elapsed Time (sec)
T
a
k
e
o
f
f
P
a
r
a
m
e
t
e
r
s
Calibrated Airspeed (Knots)
Distance (Feet/50)
Angle of Attack (Degrees*10)
Altitude (Feet)
Figure 8.5 Takeoff Parameters
Table 8.1 shows the significant events during the takeoff.
Table 8.1
TAKEOFF EVENTS
Seconds
C
V
(kts)
α
(deg)
C
H
(ft)
Event
0.0 0.0 0 0.0 Brake Release/
Fn = 35 Percent
8.4 50.1 0 0.0 99Percent Thrust
15.3 100.0 0 0.0 Rotation Initiated
19.6 130.3 13 0.0 Rotation Completed
19.9 132.2 13 0.0 LiftOff
Lift>Weight
23.7 154.0 13 16.3 OutofGround Effect
26.43 167.6 13 50.0 Obstacle
Clearance
Height
The inflection points in the drag versus calibrated airspeed plot (See Figure 8.4) can easily be
correlated with the significant events in Table 8.1. For instance, from the initiation until
completion of rotation, the angle of attack is increasing (at 3 degrees per second), which shows
up in a dramatic rate of change of drag. Once angle of attack stabilizes at 13 degrees, the rate of
increase of drag is reduced.
87
8.5 Effect of Runway Slope
Using the pseudo F16 model, the values of time and distance as a function of runway slope
(in degrees) are shown in the Table 8.2. The average acceleration is computed as follows:
2
2 / a d t · ⋅ average (mean) acceleration (ft/sec
2
) (8.49)
where:
t = time at liftoff (seconds), and
d = distance to liftoff (feet).
Table 8.2
EFFECT OF RUNWAY SLOPE
Slope
(deg)
Distance
(ft)
Time
(sec)
Acceleration
(ft/sec
2
)
From Zero
(pct)
1.0 3,001 22.6 11.75 4.52
0.0 3,131 23.6 11.24 0.00
0.5 3,164 24.0 10.99 2.29
1.0 3,247 24.6 10.73 4.56
2.0 3,403 25.8 10.22 9.06
As can be seen, the effect of runway slope for this particular model is about 4.5 percent per
degree of runway slope. For a typical light aircraft the effect of runway slope is at least twice that
amount, due to the much smaller thrust to weight ratio of the typical light aircraft. The Edwards
AFB main runway has an average slope of only 0.08 degree (21 feet elevation change in 15,000
feet). The true heading for runway 22 is 238.32 degrees from true north (224.1 magnetic). The
west end of the runway is 21 feet higher than the east end. For our
F16 model, this slope would produce a 3,142foot takeoff distance compared to 3,131 feet for a
perfectly level runway.
Although the percentage change in acceleration is about the same for a positive or negative
runway slope, one must take into account the fact of having a negative absolute rate of climb at lift
off for a negative slope runway. For instance, for a liftoff at 100 knots groundspeed with a
negative 1.0degree slope runway, the absolute rate of descent is about 3 feet per second. The rate
of climb (or descent) with respect to the horizontal plane is as follows:
sin( )
g rw
h V θ · ⋅
!
(8.50)
8.6 Effect of Wind on Takeoff Distance
Again, using the same pseudo F16 model, Figure 8.6 illustrates the effect of wind. The
takeoff speed is 132 knots calibrated airspeed. A positive wind on this plot is a headwind.
88
Effect of Wind on Liftoff Distance
30%
20%
10%
0%
10%
20%
30%
20 15 10 5 0 5 10 15 20
Wind Speed (kts)
C
h
a
n
g
e
f
r
o
m
Z
e
r
o
W
i
n
d
(
p
c
t
) Simulation
Linear (Simulation)
Figure 8.6 Effect of Wind
8.7 Takeoff Using Vectored Thrust
A limiting factor in takeoff distance for a highperformance fighter may be the ability to
rotate the aircraft. Rotation would usually be achieved using the horizontal tail. The tail generates
lift from dynamic pressure. A full fuel F16 with no stores has a takeoff weight of approximately
25,000 pounds. The engine on an F16 aircraft in maximum afterburner has a static sea level
rating of about 25,000 pounds. This does not mean the engine, when installed in the aircraft,
produces that much thrust. There will be some degradation due to installation losses. For the sake
of using even numbers, however, we will assume zero losses. In addition, the simulation that will
be presented here will be for sea level. Figure 8.7 illustrates forces and dimensions for an F16
aircraft. We will presume that we have installed a nozzle with vectoring capability.
As shown, the length of the F16 is 49.25 feet. The following dimensions are approximate
values scaled from the diagram:
a. 14.5
Fn
X · feet (distance from main gear to thrust vector).
b.
1
8.7 X · feet (distance from weight vector to nosewheel).
c.
2
4.4 X · feet (distance from weight vector to main wheel).
89
1
X
Figure 8.7 F16 Dimensions
The forces are the same as for the conventional takeoff. The difference is that there will be
thrust vectoring to produce a pitching moment to rotate the aircraft.
V
θ · thrust vectoring angle (+ nozzle up, to produce a pitch up).
Requiring the summation of moments about the main gear to be equal to zero yields equation
8.51. We will assume that the lift and the weight act through the same distance (
2
X ). This is not
generally the case. We will also ignore the longitudinal forces. A more complete simulation
would not make these simplifying assumptions. The assumptions made here are deleting higher
order terms.
( ) ( )
1 1 2 2 2
0 sin
t n V Fn
M F X X L X W X F X θ · · ⋅ + + ⋅ − ⋅ + ⋅ ⋅
∑
(8.51)
Solving for the nosewheel force (
1
F ):
( )
( )
2 2
1
1 2
sin
t n V Fn
W X L X F X
F
X X
θ ⋅ − ⋅ − ⋅ ⋅
·
+
(8.52)
Rotation will begin when the nosewheel force (
1
F ) becomes zero. At zero airspeed, lift ( L )
is zero. With
1
F equal to zero, we can solve for the vector angle that would be required to pitch
the aircraft at zero airspeed.
( )
( )
2 1
sin
t
V
n Fn
W X
F X
θ
−
¹ ¹ ⋅
¹ ¹
·
' '
⋅
¹ ¹
¹ ¹
(8.53)
L
49.25 ft
t
W
1
F
2
X
Fn
X
V
θ
n
F
2
F
90
For the conditions we have chosen, the vector angle computes to:
( )
1
25, 000 4.4
sin 17.7
(25, 000 14.5)
V
θ
−
¹ ¹ ⋅
· · °
' '
⋅
¹ ¹
(8.54)
In round numbers, we would need to rotate the nozzle 18 degrees to rotate the aircraft at zero
airspeed using thrust alone. That assumes the engine is producing 100percent thrust at brake
release. At higher airspeeds, the nozzle angle required will be less due to wing lift. The engine
vectoring would only be used to initiate rotation. Once rotation begins, the vector angle can be
decreased as the wing lift increases. Ignoring any tail lift, equation 8.51 becomes:
( ) ( )
2
sin
yy t n V
M I q L W X F θ · ⋅ · − ⋅ + ⋅
∑
(8.55)
where:
yy
I = moment of inertia about the ybody axis, and
q = body axis pitch rate.
For sea level, standard day and with the aircraft model previously defined, Figures 8.8 and
8.9 illustrate liftoff performance. The simulation assumed rotation was initiated at 90 knots and
a rotation rate of 10 degrees per second was obtained. This 10degree per second rotation rate
versus the previous 3degree per second rate was used in the simulation to minimize the distance
traveled between initiation of rotation and liftoff. It was presumed that some sort of control
system function accomplishes the rotation to avoid overrotation at these high rotation rates.
Overrotation means aft airframe ground contact. The rotation was continued until liftoff attitude
(α θ · ) was attained. Then that attitude was maintained until liftoff (
t
L W > ).
91
Distance versus Vc
600
700
800
900
1,000
120 125 130 135 140 145
Calibrated Airspeed at LiftOff (kts)
D
i
s
t
a
n
c
e
t
o
L
i
f
t

O
f
f
(
f
t
)
Figure 8.8 Distance to LiftOff
Takeoff: LiftOff Alpha versus Airspeed
10
12
14
16
120 125 130 135 140 145
Calibrated Airspeed at LiftOff (kts)
A
n
g
l
e
o
f
A
t
t
a
c
k
a
t
L
i
f
t

O
f
f
(
d
e
g
)
Figure 8.9 Angle of Attack at LiftOff
92
8.8 Effect of Thrust Component
In the previous simulation, which has been the subject of this entire section so far, we have
ignored the component of thrust. Once the thrust vectoring has accomplished its task of rotating
the aircraft, the nozzle would be vectored to zero degrees with respect to the thrust axis. The
simplified formula we used to compute normal load above is shown in equation 8.56, which is
only applicable after liftoff has occurred. During the ground roll, a portion of the aircraft weight
is supported by the ground.
z
t
L
N
W
· (8.56)
The complete formula is as follows:
( ) sin
z t g t
L N W F i α · ⋅ − ⋅ + (8.57)
Hence, solving for
z
N :
( ) ( )
sin
g t
z
t
L F i
N
W
α + ⋅ +
· (8.58)
We have presumed the thrust incidence angle
t
i is zero. The effect of ignoring the sin( )
g
F α ⋅
term is quite dramatic. For instance, at the typical liftoff angle of attack for an
F16 of 13 degrees α , the term for
g
F = 25,000 pounds yields 5,624 pounds of extra equivalent
lift. A plot of liftoff speed versus angle of attack (Figure 8.10) illustrates the effect.
Effect of Ignoring Thrust Component In Lift Axis
100
110
120
130
140
150
11 12 13 14 15 16
Angle of Attack at LiftOff (deg)
L
i
f
t

O
f
f
C
a
l
i
b
r
a
t
e
d
A
i
r
s
p
e
e
d
(
k
t
s
)
Ignoring Thrust Component
With Thrust Component
Poly. (Ignoring Thrust Component)
Poly. (With Thrust Component)
Figure 8.10 Effect of Thrust Component on LiftOff Speed
93
The corresponding distances are presented in Figure 8.11. The liftoff angle of attack was
varied to produce the variation in liftoff speed.
Distance versus LiftOff Airspeed: Effect of Ignoring Thrust Component
400
500
600
700
800
900
1,000
100 105 110 115 120 125 130 135 140 145 150
Calibrated Airspeed at LiftOff (kts)
D
i
s
t
a
n
c
e
t
o
L
i
f
t

O
f
f
(
f
t
)
Ignoring Thrust Component
With Thrust Component
Figure 8.11 Effect of Thrust Component on Distance to LiftOff
At 13 degrees α , we (the simulation) are able to liftoff at 116.2 knots in only 618 feet.
Without thrust vectoring, the F16 would (for these conditions) not be able to rotate before
approximately 130 knots. We can take the nosewheel force equation and replace the thrust vector
term with a tail lift term.
( )
2 2
1
1 2
t t t
W X L X L X
F
X X
⋅ − ⋅ − ⋅
·
+
(8.59)
Now, replace the terms above with the more general terms as shown in the C141 diagram
(See Figure 8.1). However, we will ignore runway slope and vertical terms. Again, taking
moments about the main gear:
( ) ( ) { ¦ ( )
1 1 2 1 2 1 2 2 2 1 2
0
t
M F X X L X XL L XL X XL W X · · ⋅ + + ⋅ − − ⋅ − − − ⋅
∑
(8.60)
To rotate the aircraft using tail lift, the tail lift (
2
L ) must be negative. Solving for the nose
load:
{ ¦ ( ) ( )
( )
2 2 2 1 2 1 2 1
1
1 2
t
L XL X XL W X L X XL
F
X X
]
⋅ − − + ⋅ − ⋅ −
]
·
+
(8.61)
94
Rotation will occur when the nose load (
1
F ) equals zero. Solving for the required tail lift:
( )
{ ¦ ( )
1 2 1 2
2
2 2 1
t
L X XL W X
L
XL X XL
] ⋅ − − ⋅
]
·
− −
(8.62)
For our aircraft model, we have assumed
1
0 XL · and we will assume the tail force acts at
the same point where we assumed the thrust vector acted. Then:
2 2
14.5 4.4 18.9
Fn
XL X X · + · + · (8.63)
And:
( )
( )
( ) ( )
1 2
2 1 1
2 2
0.303 0.3
t
t t
L W X
L L W L W
XL X
] − ⋅
]
· · ⋅ − ≅ ⋅ −
−
(8.64)
Next, we can compute the difference between the tail lift (
2
L ) and the opposing lift from
weight (
t
W ) and wing lift (
1
L ).
( )
2 1
0.3
t
Lift L L W ∆ · − ⋅ − (8.65)
During the aircraft takeoff ground roll, the angle of attack (α ) will be zero, but the wing will
provide lift due to having flaps down configuration. A tail lift coefficient of 1.50 is assumed
along with sea level standard conditions and a gross weight of 25,000 pounds. Four values of
wing lift coefficient are chosen to be 0.10, 0.20, 0.30 and 0.40. Figure 8.12 shows the results of
plotting Lift ∆ versus calibrated airspeed (
C
V ) for a tail area of 60 ft
2
.
Figure 8.13 is for a tail area of 80 ft
2
.
DeltaTail Lift Tail CL=1.5; Tail S=60 ft^2
 4000
 3000
 2000
 1000
0
1000
2000
3000
4000
110 115 120 125 130 135 140 145 150 155 160
Cal ibr at ed Airspeed ( knot s)
D
e
l
t
a
L
i
f
t
[
L
2

0
.
3
*
(
L
1

W
e
i
g
h
t
)
]
(
p
o
u
n
d
s
)
Wing CL= 0.10
Wing CL= 0.20
Wing CL= 0.30
Wing CL= 0.40
Figure 8.12 Delta Tail Lift for Tail Area = 60 ft
2
95
Delta Tail Lift CL tail = 1.5; S tail = 80 ft^2
4000
3000
2000
1000
0
1000
2000
3000
4000
110 115 120 125 130 135 140 145 150 155 160
Calibrated Airspeed (knots)
D
e
l
t
a
L
i
f
t
[
L
2

0
.
3
*
(
L
1

W
e
i
g
h
t
)
]
(
p
o
u
n
d
s
)
Wing CL= 0.10
Wing CL= 0.20
Wing CL= 0.30
Wing CL= 0.40
Figure 8.13 Delta Tail Lift for Tail Area = 80 ft
2
The points on the plots where the Lift ∆ becomes positive is the minimum speed for rotation.
For instance, for a wing lift coefficient of 0.40 and a tail area of 80 ft
2
, the minimum rotation
speed is about 119 knots (from Figure 8.13).
For this aircraft simulation, we have assumed a constant 25,000 pounds of thrust. This is
much greater than drag at liftoff speed. By varying the rotation speed, we can generate a plot of
distance versus speed for liftoff (Figure 8.14). The rotation rate was assumed 10 degrees per
second in each case. The 10degrees per second rate is much greater than a normal rate of about 4
degress per second. The high rotation rate in the simulation was necessary to achieve reasonable
liftoff speeds. Figure 8.14 shows the results. The line is approximately a straight line and is
such, due to thrust being much greater than drag, which produces a nearly constant acceleration
versus speed.
96
Distance versus Vc at LiftOff
800
900
1,000
1,100
1,200
130 140 150 160 170
Calibrated Airspeed at LiftOff (kts)
D
i
s
t
a
n
c
e
t
o
L
i
f
t

O
f
f
(
f
t
)
Figure 8.14 Distance to LiftOff versus Airspeed
In each data point in Figure 8.14, the limiting factor in liftoff was the rotation rate. The liftoff occurred
before 13degrees α was achieved. Figure 8.15 shows rotation speed versus liftoff speed and illustrates just
how rapidly the aircraft (in this case, the aircraft model) was accelerating.
LiftOff Vc versus Rotation Vc: Thrust = 25,000 lbs
130
140
150
160
170
110 120 130 140 150
Calibrated Airspeed at Rotation (kts)
C
a
l
i
b
r
a
t
e
d
A
i
r
s
p
e
e
d
a
t
L
i
f
t

O
f
f
(
k
t
s
)
rotation=10 deg/sec
Linear (rotation=10 deg/sec)
Figure 8.15 Calibrated Airspeed at LiftOff
97
Table 8.3 shows the forces at 130 knots calibrated airspeed.
Table 8.3
FORCES AT LIFTOFF SPEED
n
F
(lbs)
α
(deg)
L
C
D
C
Lift
(lbs)
Drag
(lbs)
rw
F
(lbs)
ex
F
(lbs)
25,000 0.0 0.10 0.0501 1,716 860 345 23,795
25,000 13.0 1.420 0.1420 24,369 2,437 9 22,554
At rotation for 130 knots, for an excess thrust of 22,795 pounds, the speed is increasing at
17.2 knots per second. That is why we needed such a high rotation rate, in order to achieve a
reasonable liftoff speed. We must emphasize here that the model used was not an accurate
F16 model, but merely an approximate model used to illustrate takeoff principles. The equations
for the lift and drag models were presented earlier. Figures 8.16 and 8.17 are plots of these
equations.
Takeoff Model: CL versus Alpha
0.0
0.4
0.8
1.2
1.6
2.0
0 2 4 6 8 10 12 14 16
Angle of Attack (deg)
L
i
f
t
C
o
e
f
f
i
c
i
e
n
t
(
C
L
)
Figure 8.16 Takeoff Lift Model
98
Takeoff/Landing Drag Model: CD Vs Alpha
0.04
0.08
0.12
0.16
0.20
0 2 4 6 8 10 12 14 16
Angle of Attack (deg)
D
r
a
g
C
o
e
f
f
i
c
i
e
n
t
(
C
D
)
Figure 8.17 Takeoff Drag Model
In computing drag on the ground, you start with a given angle of attack, then compute lift
coefficient, and finally drag coefficient.
Ground:
L D
C C α → → (8.66)
Once liftoff occurs, one is able to compute lift coefficient. You can also measure angle of
attack. Then, you start with lift coefficient and compute drag coefficient. Ignoring the component
of gross thrust:
Air:
( )
2
0.000675
z t
L D
N W
C C
M S δ
⋅
· ⋅ →
⋅ ⋅
(8.67)
The lift and drag model used for this analysis is an idealized linear model. In the real world,
there will be deviations from the linear model caused by flow separation at higher angles of
attack. Experience has shown that this nonlinearity will begin at lift coefficients on the order of
0.50.
8.9 EngineInoperative Takeoff
In this section, we will discuss takeoff of a twoengine aircraft with an engine failure at some
point during the takeoff ground roll. We will use the same pseudo F16 aero model. However, we
will assume two engines instead of one. We will make simplifications, such as assuming an
instantaneous loss of thrust on the failed engine. The purpose herein is to illustrate basic
principles  not to generate an accurate simulation. Let us presume a very simple thrust model for
each engine as follows:
99
a. 5, 000
n
F
δ
· pounds.
Now, we will simulate a takeoff at high altitude where the performance would be minimal if
one engine were to fail. We will assume 10,000 feet pressure altitude ( 0.6877 δ · ). Figure 8.18
is a time history of a simulation for our 25,000pound aircraft model with both engines operating.
Takeoff Parameters versus Time
0
20
40
60
80
100
120
140
160
180
0 10 20 30 40 50
Elapsed Time (sec)
T
a
k
e
o
f
f
P
a
r
a
m
e
t
e
r
Calibrated Airspeed (kts)
Distance (ft/100)
Angle of Attack (deg*10)
Altitude (ft)
Figure 8.18 Takeoff Parameters versus Time
Takeoff forces versus calibrated airspeed up to an altitude of 100 feet are presented in Figure
8.19. The plot is for both engines operating.
100
TwoEngine Takeoff Forces
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
0 20 40 60 80 100 120 140 160 180
Calibrated Airspeed (kts)
F
o
r
c
e
(
l
b
s
)
Net Thrust
Drag + Runway Resistance
Excess Thrust
Figure 8.19 Takeoff Forces versus Airspeed
For liftoff and 50 feet, Table 8.4 presents takeoff parameters.
Table 8.4
TAKEOFF PARAMETERS AT FLIGHT EVENTS
Event*
Time
(sec)
α
(deg)
C
V
(kts)
n
F
(lbs)
rw
D F +
(lbs)
ex
F
(lbs)
h
!
(ft/sec)
V
!
(kts/sec)
1 0 0 0 6,877 375 6,502 0 4.96
2 31.800 0 130.0 6,877 1,206 5,671 0 4.32
3 33.100 13.0 134.6 6,877 2,600 4,277 0 3.26
4 39.550 13.0 150.8 6,872 2,990 3,881 3.82 2.71
5 44.725 13.0 161.6 6,864 3,423 3,441 11.41 1.94
6 47.575 13.0 165.3 6,850 3,585 3,265 24.50 1.05
*The numbered events are as follows:
1. Brake release
2. Initiate rotation
3. Liftoff
4. Outofground effect (
AGL
h = 19.7 feet)
5. 50 feet AGL (above ground level)
6. 100 feet AGL
The twoengine case in Figure 8.19 was presented primarily as a baseline of comparison for
the following engine failed case. We will now assume that one engine fails at exactly the
initiation of rotation ( 130
C
V · knots). Figure 8.20 illustrates the same parameters as shown in
Figure 8.19.
101
Engine Failure Takeoff Forces
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
0 20 40 60 80 100 120 140 160
Calibrated Airspeed (kts)
F
o
r
c
e
(
l
b
s
)
Net Thrust
Drag + Runway Resistance
Excess Thrust
Figure 8.20 Takeoff Forces versus Airspeed: Engine Inoperative
Table 8.5 duplicates Table 8.4 for the same events, except we will add an event (2.1), which
is immediately after we fail one engine in the simulation.
Table 8.5
TAKEOFF PARAMETERS AT SIGNIFICANT EVENTSENGINEINOPERATIVE
Event
Time
(sec)
α
(deg)
C
V
(kts)
n
F
(lbs)
rw
D F +
(lbs)
ex
F
(lbs)
h
!
(ft/sec)
V
!
(kts/sec)
1 0 0 0 6,877 375 6,502 0 4.96
2
2.1
31.79
31.80
0
0
130.0
130.0
6,877
3,438
1,206
1,206
5,671
2,232
0
0
4.32
1.70
3 33.70 13.0 132.0 3,438 2,503 935 0 0.71
4 68.00 13.0 147.7 3,436 2,884 552 0.63 0.38
5 100.00 13.0 154.6 3,432 3,133 299 3.49 0.01
6 109.05 13.0 153.6 3,425 3,100 325 7.04 0.20
*The numbered events are as follows:
1.0 Brake release
2.0 Initiate rotation
2.1 Engine failure
3.0 Liftoff
4.0 Outofground effect (
AGL
h = 19.7 feet)
5.0 50 feet AGL (above ground level)
6.0 100 feet AGL
As can be seen, by the time altitude equals 100 feet the aircraft is slowing. Although excess
thrust is increasing slightly, that excess thrust is being used for climb at the expense of airspeed.
102
In case of an engine failure in such a scenario, one would need to reduce the drag and pitch over
to reduce rate of climb. The drag reduction would be accomplished by raising the gear. Then,
conduct a lowg turn (to minimize drag) and return to base for landing. This is just one possible
option. The aircraft flight manual would contain the recommended emergency procedure.
8.10 Idle Thrust Decelerations
To assist in the development (or verification) of a takeoff and landing simulation, idle thrust
decelerations may be performed. One would accelerate the aircraft on the runway to some high
airspeed. Then, cut the throttle to idle and allow the aircraft to freely decelerate. We can solve for
drag (
D
) in the equation found in the Developing a Takeoff Simulation subsection and then put
D into coefficient form. Lift and drag coefficients are discussed in the lift and drag section of
this handbook.
[ ]
sin( ) cos( )
n ex t rw t rw
D F F W W L θ µ θ µ · − − ⋅ − ⋅ ⋅ + ⋅ (8.69)
103
9.0 LANDING
9.1 Braking Performance
Using the same aero model as for takeoffs, one can see the effect of braking coefficient of
friction (
µ
) upon stopping performance. The thrust has been set to a constant 600 pounds,
representing Idle thrust. Minimum drag coefficient has been increased from 0.0500 to 0.0700 to
account for additional drag devices (such as spoilers) activated during braking. In Figure 9.1, the
coefficient of friction has been set to a constant 0.35; this is a typical dry runway value. The
initial groundspeed was 130 knots for a calibrated airspeed of 124.8 knots. The gross weight has
been reduced to 20,000 pounds, more representative of landing weight. The pressure altitude is
2,300 feet with zero wind.
Braking Forces: Mu = 0.35; Cd= 0.0700; Fn = 600 lbs
8,000
6,000
4,000
2,000
0
2,000
4,000
6,000
8,000
0 20 40 60 80 100 120 140
Calibrated Airspeed (kts)
F
o
r
c
e
(
l
b
s
)
Drag + Runway
Resistance
Net Thrust
Excess Thrust
Figure 9.1 Braking Forces
For a dry runway, the µ for maximum braking is typically between about 0.35 and 0.50.
However, when one has an 8,000foot runway, you usually will not conduct a maximum
performance stop just to minimize tire and brake wear. Figure 9.2 shows the distance as a
function of µ for the 20,000pound aircraft at 2,300 feet pressure altitude with initial speed of
130 knots groundspeed.
104
Dry Runway: Distance versus Mu
1,000
1,500
2,000
2,500
3,000
3,500
0.25 0.30 0.35 0.40 0.45 0.50
Mu: Braking Coefficient of Friction
S
t
o
p
p
i
n
g
D
i
s
t
a
n
c
e
(
f
t
)
Figure 9.2 Stopping Distance versus Mu ( µ )
For the braking coefficient range of 0.25 to 0.50, Figure 9.3 illustrates the deceleration (knots
per second) versus calibrated airspeed.
Braking Deceleration
0
1
2
3
4
5
6
7
8
9
10
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Calibrated Airspeed (kts)
D
e
c
e
l
e
r
a
t
i
o
n
(
k
t
s
/
s
e
c
)
Mu = 0.50
Mu = 0.40
Mu = 0.30
Mu = 0.20
Mu = 0.10
Figure 9.3 Deceleration versus Calibrated Airspeed
For wet runway conditions, the µ is much less than for dry runway conditions. This is
especially true at high speed where hydroplaning may occur. Hydroplaning is where the tires ride
on a film of water and never contact the runway. Figure 9.4 represents actual test data. The test
105
was on a wet runway, with the water applied using water tankers. The data points are average
values of the actual data and the line is a fourthorder polynomial curve fit of the data points.
y = 3.736E09x
4
 1.381E06x
3
+ 1.811E04x
2
 1.137E02x + 4.326E01
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 20 40 60 80 100 120 140
Ground Speed (Knots)
F
r
i
c
t
i
o
n
C
o
e
f
f
i
c
i
e
n
t
(
M
U
)
Figure 9.4 Mu versus Groundspeed (Wet Runway)
Figure 9.4 shows the braking coefficient computed from braking tests. The limits that will be
used in applying the curve fit will be the curve fit values at the extreme points as follows:
a. 0.336 if 10 knots
g
V µ · < , and
b. 0.047 if 130 knots
g
V µ · > .
A warning is appropriate for using curve fits in simulations. Invariably, the data will not
extend to the full range of the desired simulation. Using the curve fit beyond the range of its data
should be avoided by use of limits. A limit would be where the curve fit value (y) would take on
some predetermined constant value if the x value exceeds the highest (or lowest) value used in
the curve fit.
Wet runway forces are shown in Figure 9.5. The forces are computed using the mu or µ from
Figure 9.4.
106
Braking Forces: Wet Runway
8,000
6,000
4,000
2,000
0
2,000
4,000
6,000
8,000
0 20 40 60 80 100 120 140
Calibrated Airspeed (kts)
F
o
r
c
e
(
l
b
s
)
Net Thrust
Drag + Runway Resistance
Excess Thrust
Figure 9.5 Braking Forces versus Calibrated Airspeed
The simulation for our wet runway model produces a total distance of 7,059 feet. This
compares to a distance of 2,236 feet for our dry runway model using a constant
µ
of 0.35. That
is a factor of more than three times longer for a wet runway. That is typical, but as the saying
goes, "your results may vary.”
9.2 Aerobraking
When one is faced with a wet or icy runway, in order to reduce the ground roll, aerobraking
may be used. Upon touching down, instead of immediately pushing over to a
3point attitude to begin braking, the aircraft is held at a high pitch angle (to produce a high
angle
of attack) to maximize the aerodynamic drag. In addition, aerobraking may be used on a dry
runway simply to reduce wear on the brakes and tires. The ability to perform aerobraking is
limited by at least two factors. First is the tail scrape angle, which limits how high of an angle of
attack may be held. Second is the control power available to hold the aircraft up at an angle of
attack. Figure 9.6 illustrates the difference in total resistance for aerobraking versus 3point
braking. For this simulation, the 3point braking has more resistance except at high airspeed.
However, in many cases, aerobraking can be more effective.
107
Drag + Runway Resistance Comparison: Aerobraking versus 3point Braking
1,400
1,600
1,800
2,000
2,200
2,400
2,600
100 105 110 115 120 125
Calibrated Airspeed (kts)
D
r
a
g
+
R
u
n
w
a
y
R
e
s
i
s
t
a
n
c
e
(
l
b
s
)
3Point Braking
Aerobraking 15 degrees alpha
Figure 9.6 Total Resistance Force Comparison
9.3 Landing Air Phase
The landing air phase will be discussed using the same aircraft model we have used for the
takeoff discussion and the landing ground roll. The simulation will be conducted by first
computing the initial conditions. We can compute the initial speed (Mach number), by assuming
that the flight path angle (
γ
) is initially constant ( 0 γ · ! ). The normal load factor equation is the
same as for takeoff (equation 8.40).
0
cos( )
t
z
V
N
g
γ
γ
⋅
· +
!
(9.1)
Then,
cos( )
z
N γ · (9.2)
Each aircraft is flown differently and different pilots may have slightly different pilot
techniques. However, a typical final approach technique is a constant angleofattack descent. For
our simulation, that angle of attack is 13 degrees. From angle of attack we can estimate the lift
coefficient (
L
C ). The simulation used an estimated
L
C of 1.05 (out of ground effect) for an
angle of attack of 13 degrees. Then, we can compute Mach number as follows when we also have
given the weight and altitude:
0.000675
Z t
L
N W
M
S C δ
⋅ ⋅
·
⋅ ⋅
(9.3)
108
Equation 9.3 is solving for Mach number from equation 5.6 in Section 5.0 (Lift and Drag).
Further, we will assume that true airspeed is constant, initially. The longitudinal load factor
equation then gives:
0
t
x
t t
V h h
N
V g V
· + ·
! ! !
(9.4)
We can then solve for the net thrust that would be required to have true airspeed constant at
the beginning of the landing descent.
n ex x t
F D F D N W · + · + ⋅ (9.5)
Having performed these computations, the initial descent rate is varied. The initial conditions
chosen÷a runway pressure altitude of 2,300 feet at a standard day and an obstacle clearance
height of 50 feet÷are what might be typical with a postmission weight of 18,000 pounds.
For this aircraft model, the simulation enters ground effect at 16 feet (AGL) and at touchdown,
the additional lift is a factor of 1.30. Figure 9.7 illustrates the dramatic impact of ground effect. A
constant angle of attack of 13.0 degrees is maintained and thrust is held constant. However, the
ground effect will increase the lift and hence, the descent rate will decrease.
Final Descent Rate versus Initial Descent Rate
0
2
4
6
8
10
12
14
16
18
10 12 14 16 18 20
Initial Descent Rate (at 50 ft) (ft/sec)
F
i
n
a
l
D
e
s
c
e
n
t
R
a
t
e
(
a
t
0
f
e
e
t
)
(
f
t
/
s
e
c
)
Figure 9.7 Final Descent Rate versus Initial Descent Rate
The aircraft simulation predicted that, for the conditions specified, the aircraft would not
touch down at any initial descent rate less than about 11.2 ft/sec. This is an ideal computer
simulation, not a real airplane. In the real world, the pilot would take action to touch down with
stick, throttle or speed brake. A pushover would decrease angle of attack, which would decrease
109
lift, thereby increasing descent rate. A pushover to about 10 degrees angle of attack would
suffice. Interestingly, a pullup would also eventually get you on the ground. By pulling up
sufficiently to dramatically increase drag, the aircraft will decelerate. With a lower airspeed, the
lift will decrease and when lift becomes less than weight, you will descend. Reducing thrust will
also cause a deceleration, however, you are already at near idle thrust and the small additional
thrust increment could be insufficient. Finally, speed brake can be used to slow down and reduce
lift.
A time history of the descent for the landing simulation is shown in Figure 9.8. The
simulation computations were begun at 50 feet AGL (above ground level), but only the last 20
feet are shown. Notice the curvature in the final phase of the altitude versus time. The total
distance from 50 feet to touchdown was computed to be 1,074 feet. When the same simulation
was performed with ground effect terms eliminated, the total distance changed to 978 feet, for a
difference of 96 feet or nearly 10 percent of the air distance.
Last 20 Feet of Landing Descent
0
2
4
6
8
10
12
14
16
18
20
2.5 3.0 3.5 4.0 4.5 5.0
Elapsed Time from 50 ft (sec)
A
l
t
i
t
u
d
e
;
R
a
t
e
o
f
D
e
s
c
e
n
t
;
D
i
s
t
a
n
c
e
/
1
0
0
Altitude (ft)
Rate of Descent (ft/min)
Distance (100 ft)
Figure 9.8 Landing Air Phase
9.4 Landing on an Aircraft Carrier
The following text is the result of information given to the author by Page Senn and Richard
Huff of the Naval Air Weapons Center, Patuxent River, Maryland. The situation we will discuss
is the landing of an F/A18 on a Nimitz class carrier. Figure 9.9 is a U.S. Navy photo of an F/A
18 with its tailhook extended. At landing attitude [ 8.1 α · ° and glideslope = 3.5 degrees (or
3.5 γ · − ° )], the vertical height from the tailhook to the pilot’s eye is 16.7 feet. The wing is
roughly half the distance between the pilots eye and the tailhook as can be seen from the photo.
110
Hence, the wing height above the tailhook is about 9 feet. We will use that height to make
estimates of ground effect.
Figure 9.9 F/A18 with Tailhook Extended
Figure 9.10 is a Navy photo of the U.S.S. Nimitz. The landings are accomplished from the aft
deck while the carrier is maintaining forward speed to give a minimum wind over the deck of 15
knots. A more normal wind is 25 knots.
Figure 9.10 The U.S.S. Nimitz
The distance from the ramp to the target hook touchdown point is 230.2 feet. For the
3.5degree glideslope, this computes to a hook to ramp clearance of 14.08 feet for no flare. For
the F/A18 at 33,000 pounds, the airspeed is 146 knots. With the minimum windspeed of 15
knots, this yields a groundspeed of 131 knots (14615) assuming standard day temperature. We
can calculate the time from passing over the ramp to tailhook touchdown as follows:
111
distance(ft) 230.2
time 1.04 sec
speed (ft/sec) 131 1.6878
∆ · · ·
⋅
(9.6)
Since 15 knots of wind is the minimum, the time will generally be longer. A wind of 25
knots, for instance, would produce a time of 1.13 seconds. The average sink rate from the ramp to
target hook touchdown computes to 13.5 fps (ft/sec). This compares to the nominal sink rate 14
fps. For the F/A18, the gear limit is 25 fps and testing at Patuxent is accomplished up to 20 fps.
Now, to estimate ground effect. The wingspan of the F/A18 is 40.4 feet. Table 9.1 shows the
height/span (h/b) of the aircraft versus distance along the deck from over the ramp to tailhook
touchdown. Also shown is an estimate of percentage reduction in drag from Figure 8.2.
Table 9.1
GROUND EFFECT PARAMETERS FOR F/A18 CARRIER LANDING
Point Over Deck
Distance
Traveled
(ft)
Wing Height
(ft)
h/b
Percentage
Drag
(pct)
0 23.1 0.57 94.8
50 20.0 0.50 91.4
100 17.0 0.42 87.4
150 13.9 0.34 82.6
Ramp
200 10.8 0.27 76.6
Hook Touchdown 230.2 9.0 0.22 72.1
Note: The percentage drag is an estimate of the drag as a percentage of the outofground effect
drag.
We can estimate the change in speed of the aircraft due to ground effect. One form of the
relationship between drag and drag coefficient is derived in the lift and drag section and is
repeated below:
( )
2
0.000675
D
C M S
D
δ ∆ ⋅ ⋅ ⋅
∆ · (9.7)
For sea level standard day, 1.0 δ · and airspeed of 141 knots yields a Mach number ( M ) of
0.2132. Airspeed and Mach number relationships are found in Section 4 (Airspeed). For an out
ofground effect drag coefficient of 0.25, we can estimate the change in speed by integrating.
From
D
C ∆ , we calculate D ∆ using equation 9.7. Then, for a weight of 33,000 pounds we
calculate longitudinal load factor and then the derivative of velocity. This assumes that all of the
drag change goes into acceleration and none into changing the rate of descent.
0 0
33, 000
t t
x
t
V V D h
N
g V g
∆
· · + ·
! ! !
0
32.174
19.06 (knots/sec)
1.6878
t x x x
V g N N N · ⋅ · ⋅ · ⋅
!
(9.8)
For a groundspeed of 126 knots (212.7 ft/sec), we will assume a constant descent rate based
upon on a 3.5degree glideslope.
112
sin 212.7 sin( 3.5 ) 12.985 ft/sec
g
h V γ · ⋅ · ⋅ − ° · −
!
(9.9)
Now, we can calculate the change in speed by integrating the speed derivative as shown in
Table 9.2.
Table 9.2
CHANGE IN TRUE AIRSPEED DURING LANDING DUE TO GROUND EFFECT
Distance
Traveled
(ft)
Percentage
Drag
(pct)
D
C ∆
Drag ∆
(lbs)
x
N
t
V
!
(kts/sec)
time ∆
(sec)
t
V ∆
(kts)
t
V
(kts)
0.0 94.8 0.0130 351 0.0106 0.20 141.00
50.0 91.4 0.0216 582 0.0176 0.34 0.24 0.06 141.06
100.0 87.4 0.0316 851 0.0258 0.49 0.47 0.10 141.16
150.0 82.6 0.0436 1,174 0.0356 0.68 0.71 0.14 141.30
200.0 76.6 0.0586 1,577 0.0478 0.91 0.94 0.19 141.48
230.2 72.1 0.0698 1,880 0.0570 1.09 1.08 0.14 141.63
Note: Above data based upon an outofground effect drag coefficient of 0.25. This was not a
Navyprovided number.
Another factor in landing on a carrier is the wind over the deck. There is a downdraft
(negative vertical wind) immediately aft of the deck. The ship is traveling at a minimum of 15
knots, the air flows downward aft of the ship. Then, when that air contacts the sea below, it is
deflected upward creating an updraft for the oncoming aircraft. So, the aircraft first encounters an
updraft, then a downdraft, and then a sudden loss of any vertical wind as it encounters the aft
deck. Navy tests did indicate a 1 to 2 knot increase in INS groundspeed during landing.
9.5 Stopping Distance Comparison
During the same series of tests that produced the braking coefficient of friction data in Figure
9.4, tests were also conducted to determine aerobraking drag and dry runway braking coefficient.
The aerodynamic drag coefficient during aerobraking at 13 degrees angle of attack was
determined to be about 0.30. The dry runway braking coefficient ( µ ) was found to be in the
vicinity of 0.35. In addition, values of lift coefficient were determined from either predicted
models or flightdetermined. For a nominal landing gross weight, the touchdown speed is 135
knots calibrated airspeed. Aerobraking can be maintained until approximately 70 knots calibrated
airspeed, limited by available horizontal tail power. Table 9.3 summarizes the data for wet
runway, dry runway, and aerobraking.
113
Table 9.3
DRY, WET, AND AEROBRAKING DATA SUMMARY
Lift Coefficient
L
C
Drag Coefficient
D
C
Braking or Rolling
Coefficient ( µ )
3Point Braking: Dry 0.20 0.095 0.350
3Point Braking: Wet 0.20 0.095 Figure 9.4
Aerobraking 0.90 0.300 0.015
In addition, an idle thrust model was provided by the engine manufacturer. Since thrust was a
small contributor to the distance integration, we will ignore thrust incidence. Plus, runway slope
and wind were assumed zero and standard day conditions at sea level were used. The equation for
excess thrust (
ex
F ) then simplifies to the following:
( )
ex n t
F F D W L µ · − − ⋅ − (9.10)
Using equation 9.8 and integrating versus time to compute distance yields Table 9.4.
Table 9.4
INTEGRATION OF BRAKING RESULTS
Airspeed
C
V
(kts)
Dry
t
V
!
(kts/sec)
Dry
Distance
(ft)
Wet
t
V
!
(kts/sec)
Wet
Distance
(ft)
Aerobraking
t
V
!
(kts/sec)
Aerobraking
Distance
(ft)
135 7.17 0 2.63 0 6.11 0
125 7.06 307 2.47 873 5.25 386
115 6.95 598 2.48 1,693 4.45 705
100 6.81 992 2.58 2,768 3.34 1,510
80 6.63 1,446 2.71 3,920 2.12 2,635
50 6.41 1,950 3.04 5,088 N/A N/A
0 6.17 2,283 5.90 5,660 N/A N/A
Note: N/A – not applicable
A few observations from Table 9.4 should be made. First, dry runway 3point braking
provides the greatest deceleration at all speeds. However, by aerobraking for the first 20 knots
(135 to 115) the difference in distance is only just over 100 feet. For this small increase in
stopping distance, a substantial reduction in energy absorption by the brakes can be achieved –
thereby increasing the service life of the brakes. Second, by using aerobraking down to 100
knots, the distance to stop on a wet runway can be reduced by more than
1,000 feet.
9.6 Takeoff and Landing Measurement
In the past (prior to this handbook), much of takeoff performance utilized external tracking. At
the AFFTC, this was from Askania cameras. Askania was the brand of the particular cameras
located in towers near each end of the main runway and about 1,500 feet from the runway. The
cameras tracked the aircraft on film at up to four frames per second. The film contained azimuth
114
and elevation data. The film was developed, read, and computerprocessed. The computer output
included time, distance, velocity, acceleration, and altitude.
Now, with the advent of INS and GPS, the onboard inertial velocity data can be integrated to
provide distance.
g
d V dt · ⋅
∫
(9.11)
where:
g
V = horizontal component of groundspeed.
Altitude would be determined by integrating the vertical velocity, beginning at the point
where liftoff occurred. The precise determination of the liftoff point would involve additional
onboard instrumentation such main gear loads or wheel speed.
v
h V dt ∆ · ⋅
∫
= altitude above the liftoff point (9.12)
where:
v
V = vertical component of groundspeed.
Since the INS is subject to small drift errors, it is necessary to subtract out any null error. For
the horizontal distance, this is obtained by simply collecting data when the aircraft was stopped.
For the height integration, the vertical velocity at the liftoff point would be subtracted out. The
GPS does not have a null error. A new device called an EGI (embedded GPS/INS) combines the
outputs of both an INS and a GPS using a filter.
To compute acceleration, it is recommended to differentiate the velocities rather than use a
direct output of the INS. That is because the INS is sensitive to body axis vibrations of the
aircraft and the acceleration data will be very noisy due to this vibration. Typically, an INS will
internally integrate the accelerations at a sample rate of at least 50 samples per second. By
sampling the INS velocities at no more than 5 samples per second, you can essentially average out
the noise in the data. The topic of noise in accelerometer data is discussed within the flight path acceleration
heading of the excess thrust section. Then, the longitudinal acceleration can be determined with something as
simple as a central difference derivative method.
( )
( ) ( ) ( )
( ) ( ) ( )
1 1
1 1
g g
x
V i V i
A i
t i t i
+ − −
·
+ − −
(9.13)
where:
i = the ' i th time sample.
Improved integration results would be produced using a moving secondorder polynomial
curve fit; a data process used by the AFFTC.
115
10.0 AIR DATA SYSTEM CALIBRATION
10.1 Historical Perspective
In Engineering Aerodynamics (Revised Edition, 1936), Walter Diehl discusses the
calibration of airspeed indicators. He references NACA Rep. T.N.135 (1923) by W.G. Brown
titled, “Measuring an Airplane’s True Speed in Flight Testing.” Diehl states, “In general,
airspeed indicators must be calibrated by runs up and downwind over a measured course.”
We later knew this as the groundspeed course method. Diehl points out that such tests should
not be done when the crosswind exceeds 15 knots as that would have resulted in an error in
airspeed of more than 1 percent. In 1923, speeds of order of 100 knots were achievable. If the
groundspeed is 100.0 knots and there is a 15knot wind exactly perpendicular to the aircraft’s
inertial speed vector, then by trigonometry we could compute that the true airspeed is 101.1
knots. This is an error greater than 1 percent and even more for speeds less than
100 knots. We rarely use the groundspeed course method at Edwards because of its lack of
accuracy at high speeds and variable surface winds. The first problem is minimized with the
advent of GPS to determine groundspeeds.
10.2 Groundspeed Course Method
The course would consist of two parallel lines connected by a line perpendicular to those two
lines. The course at Edwards, for instance, is 4 miles long. The aircraft heading (direction nose is
pointing) would be the same as the course heading in method one as shown in Figure 10.1. The
aircraft would drift from the line due to any crosswind. The way to determine true airspeed is to
simply use a stopwatch to time the aircraft between the start and end lines. These points are a
known distance apart. This requires a visual hack of when the aircraft crosses the horizontal lines
marked on the ground. Then, true airspeed is determined by the following.
Distance
Time
t
V
∆
·
∆
(10.1)
As long as wind is unchanging, it does not enter into the problem since true airspeed is
parallel to the course. Then, opposite heading passes are not needed. However, it is common
to conduct passes in opposite headings just to get an average. Note: A positive wind vector
direction is the direction from which the wind is blowing.
Figure 10.1 Groundspeed Course – Heading Method
116
With the use of GPS, one could determine the component of groundspeed parallel to the
course. Now, however, one would need to conduct opposite heading passes to average out the
wind. Then, the average true airspeed is simply the average groundspeed. You would avoid
the problem of visually determining the time passing points on the ground. In addition, GPS
groundspeed is very accurate (0.1 m/sec).
( )
1 2
2
g g
t
V V
V
+
· (10.2)
Note a distinction between conducting opposite heading (direction the nose is pointing)
and opposite direction (ground track direction) passes. The opposite direction or track angle
passes would have the aircraft fly directly down the groundspeed line with the aircraft
pointing into the wind to account for crosswind. You would need to be able to correct for
crosswind if you flew these opposite direction passes as recommended in AFFTC Standard
Airspeed Calibration Procedures (Reference 10.1). The opposite direction pass would be as
shown in Figure 10.2. The opposite heading method is preferable, due to not having to make
crosswind corrections. Note: A positive wind vector direction is the direction from which the
wind is blowing. The data reduction in Reference 10.1 ignores crosswind.
Figure 10.2 Groundspeed Method – Direction Method
10.3 General Concepts
The terminology ‘airspeed calibration’ actually involved the determination of corrections
to be added to not only airspeed, but also pressure altitude and total temperature. The basic
measurements are total pressure (
t
P), static pressure ( P ), and total temperature (
t
T ). The
static (or ambient) pressure and total pressure are used to compute calibrated airspeed (
C
V ),
pressure altitude (
C
H ), and Mach number ( M ). With Mach number and total temperature,
the true airspeed and ambient temperature can be calculated. The equations for these
parameters are included in the airspeed and altitude sections of this handbook.
On some limited evaluations, the basic measured parameters on the test aircraft are the
actual measured values of indicated airspeed, indicated pressure altitude and indicated total
temperature. The correction equations are as follows:
117
C i iC pC
V V V V · + ∆ + ∆ calibrated airspeed (10.3)
C i iC pC
H H H H · + ∆ + ∆ corrected pressure altitude (10.4)
t ti ti
T T T · + ∆ total temperature (10.5)
where:
iC
V ∆
= instrument correction to indicated airspeed,
pC
V ∆
= position error correction to instrument corrected airspeed,
iC
H ∆ = instrument correction to pressure altitude,
pC
H ∆
= position error correction to pressure altitude, and
ti
T ∆
= instrument correction to total air temperature.
The modifier ‘corrected’ on pressure altitude is often dropped in practice. However, the
modifier ‘calibrated’ on calibrated airspeed needs to be retained to distinguish it from true
airspeed. When the parameters are instrument readings that not uncorrected for instrument
and position errors then the modifier ‘indicated’ should be applied. The terminology ‘position
error’ refers to the premise that there is some location on the aircraft to locate a sensor such
that there would have been zero error in that measurement. However, there is no single
position that would yield zero error at all Mach number and angle of attack.
When dealing with the three basic measurements ( , ,
t t
P P T ) on a test aircraft the i
subscript referred to a measurement that had not been corrected for any instrumentation
errors. The total temperature probe is also subject to an error called a probe recovery factor
(η ). The relationship for total versus ambient temperature is as follows:
( )
2
1 0.2
t
T T M η · ⋅ + ⋅ ⋅ (10.6)
If, in flight test, one has an ambient temperature source (T ) and a total temperature
measurement (
t
T ) one could solve for η in the above equation and could calibrate the probe.
The value for η is typically 0.98 to 1.00 for a welldesigned system. However, in practical
application with modern probes a value of 1.0 is frequently used.
The
t
T is the test aircraft’s measured total temperature. The ambient temperature (T )
would have been from another source. The other source could have been from another
aircraft with a calibrated total temperature probe, from a weather balloon, or from a ground
temperature measurement. The ground temperature measurement would be the source during
tower flyby tests.
Weather balloon data would not be used as a primary calibration source. However, it
makes an excellent check on your data system. Too many performance engineers ignore this
valuable source of information. Appendix A contains weather balloon data from the Edwards
118
AFB weather squadron. The data illustrates average values of winds and temperatures versus
month. There is also data from a sampling of 1 month of weather soundings.
A study conducted at Edwards AFB in the 1960s indicated that balloon temperature
accuracies were on the order of t2 degrees C.
The two pressure measurements could both have ‘position’ errors as follows:
t ti ti
P P P · + ∆ (10.7)
i s
P P P · + ∆ (10.8)
Often, the symbology used here for ambient pressure ( P ) will be shown as (
s
P ). The
s
would denote static. For purposes of this handbook static and ambient are considered the
same thing.
In general, both of the pressure measurements are subject to errors. However, it is often
assumed that there is zero total pressure error. In that case, all of the Pitotstatic error is in the
ambient pressure measurement. A position error parameter called delta p over q is defined as
follows:
( )
/
i
p Cic
Cic
P P
P q
q
−
∆ · (10.9)
where:
Cic
q = indicated compressible dynamic pressure, and
p
P ∆
= error in ambient pressure (position error).
With the assumption of zero total pressure error, the correction to be added to
compressible dynamic pressure simplifies to the following:
C p
q P ∆ · −∆ (10.10)
At the AFFTC, a sign convention has been that a positive sign on
p
P ∆ would produce a
positive correction to be added to both calibrated airspeed (
C
V ∆ ) and pressure altitude
(
C
H ∆ ). (One can avoid the confusion of a sign change by thinking of
p
P ∆
as being a
positive correction to be added to the compressible dynamic pressure (
C
q
.
)
A positive
correction to be added to ambient pressure would produce a negative correction to be added
to both calibrated airspeed and to pressure altitude. So, one would need to change the sign on
the ambient pressure correction as follows:
( ) ( )
/
i i
p Cic
Cic Cic
P P P P
P q
q q
− −
∆ · − ·
(10.11)
119
10.4 Pacer Aircraft
An aircraft that is utilized in the airspeed calibration of a test aircraft is called a pacer
aircraft. The pacer will fly in formation with the test aircraft. The pacer’s computed values of
calibrated airspeed (
C
V ), pressure altitude (
C
H ), and ambient temperature (T ) are compared
to those three parameter values from the test aircraft. The test aircraft’s Pitotstatic
measurements are referred to as indicated values until a set of corrections can be determined
by simply comparing to the pacers calibrated computed parameters. Just for simplicity, the
computed ambient temperature is lumped with the pressure parameters and called Pitotstatic
parameters. The AFFTC pacer aircraft have onboard computers, which calculate
instrumentation and position errors then add these corrections to the indicated values to
present calibrated values. The position errors are the difference between the measured (or
indicated) Pitotstatic parameters and the true values.
Before pacer aircraft became the standard for Pitotstatic measurement, it needed to be
calibrated before it could be utilized in the airspeed calibration of test aircraft. One of the
methods used in calibrating a pacer aircraft is to fly against another pacer aircraft. This has
the potential of passing on errors from another pacer. To avoid that problem the new pacer is
also tested using the tower flyby, acceldecel, and cloverleaf methods.
10.5 Tower Flyby
The tower flyby method of airspeed calibration consists of flying along a flyby line on the
lakebed and passing by an observation tower perpendicular to the flyby line some 1,379 feet
away (at Edwards AFB). An observer in the flyby tower watches the aircraft pass by the tower.
With a grid on a window, the observer is able to compute the aircraft’s altitude above the tower
zero grid line as the test aircraft passes in front of the grid on the window. Figure 10.3 shows an
actual photo of an aircraft (F18) passing by the Edwards AFB flyby tower.
A pressure altitude measurement in the tower is used to determine the zero grid line
pressure altitude. Then, the pressure altitude of the aircraft is computed as follows:
/
std
C a c p tower tower
T
H H h
T
 `
· + ∆ ⋅
. ,
pressure altitude for the aircraft (10.12)
where:
p tower
H = pressure altitude measured at the zero grid line in the tower,
tower
h ∆ = geometric height of aircraft above the zero grid line measured by the tower,
std
T = standard day temperature (°K) at
p tower
H , and
T = test day ambient temperature (°K).
120
Figure 10.3 Flyby Tower Grid
Figure 10.4 (Reference 10.1) represents flyby tower data.
Altitude versus Grid Reading
y = 31.422x
0
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
320
340
360
0 1 2 3 4 5 6 7 8 9 10 11 12
Grid Reading (in)
A
l
t
i
t
u
d
e
(
f
t
)
Figure 10.4 Altitude versus Grid Reading for Flyby Tower
Extracted from Reference 10.1
121
Since h ∆ · 31.422 times grid reading and at the very best a guess to the nearest 0.1 inch
grid is possible, then the accuracy of the flyby tower data is about t3 feet. That is an
optimistic figure. Accuracies of better than 3 feet have been demonstrated with differential
GPS (DGPS) over the flyby line at Edwards.
Too often, the temperature correction is ignored. To illustrate the error that could result,
consider a 90degree F day at Edwards, which is a normal summer day. The geometric
altitude of the zero grid line of the flyby tower is 2,305 feet. Assuming the pressure altitude is
equal to the geometric altitude, then the standard day temperature computes to 283.6 degrees
K. The test day temperature of 90 degrees F equates to 305.4 degrees K. Next, assume the
aircraft flew by the tower at a geometric height of 200 feet as follows:
a.
/ C a c
H ·2,305 +
283.6
200.
305.4
 `
⋅
. ,
= 2,305 + 186. = 2,491
If one ignores the temperature effect, the error in altitude would be 14 feet. Figure 10.5
illustrates the effect of a 10foot error in pressure altitude on calibrated airspeed at a pressure
altitude of 2,500 feet. This error is computed based upon the assumption that there is zero
error in total pressure.
Effect of a 10Foot Error in Flyby Tower Altitude
0.0
0.5
1.0
1.5
2.0
2.5
0 100 200 300 400 500 600
Indicated Airspeed (kts
E
r
r
o
r
i
n
C
a
l
i
b
r
a
t
e
d
A
i
r
s
p
e
e
d
(
k
t
s
)
Figure 10.5 Effect of 10Foot Error in Flyby Tower Altitude
10.6 AccelDecel
It is difficult to obtain stabilized airspeed calibration data in the transonic regime. In
addition, at supersonic speeds, fuel consumption is very high. So, a method of accelerating
122
and decelerating starting and ending at subsonic speeds (where the airspeed calibration is
known from the tests previously described) is used. The method is as follows:
a. Perform an altitude survey over a small range of altitude (t1,000 feet, typically) from
the start condition. The start condition is some Mach number, altitude condition.
b. Acquire a few additional data points at the same indicated Mach number, but at
different altitudes.
c. Measure pressure altitude, Mach number, ambient temperature (computed from Mach
number and total temperature) and tapeline altitude (radar or GPS).
d. Compute also, the windspeed and direction, groundspeed and direction, and aircraft
true airspeed. You now have the following functions:
1. ( )
C
H f h · where h = tapeline altitude,
2. ( ) T f h · ,
3. ( )
wN
V f h · , and
4. ( )
wE
V f h · .
The four functions above are quite accurately represented by a straightline curve fit. The
altitude survey can be as few as three data points to yield a straightline fit. Then, the aircraft
is accelerated from this known calibration subsonic point through the transonic and into the
supersonic regime where the calibration is not known. The data processing involves
computing corrections to be added to airspeed, altitude, and total temperature. All of the
required equations have been presented in previous sections. Figure 10.6 is a plot of a
pressure survey taken prior to a supersonic acceldecel. The extreme data points are stabilized
points while the other points are from a subsonic acceleration. The data are corrected using a
position error curve previously determined from pacer and tower flyby data. The collection of
data points near 30,000 feet pressure altitude are from a subsonic acceleration corrected using
the pacer curve. Those data points are shown in the Figure 10.6.
In Figures 10.6 and 10.7, one supersonic acceldecel data set is shown from data
collected at the same time as AFFTC data set one. That data set is in the discussion of the
cloverleaf method. Both plots are the same data; just presented with different parameters.
Figure 10.7 is correction to be added to indicated pressure altitude. Figure 10.8 is the
position error parameter versus indicated Mach number. The assumption is made that all of
the error in the air data comes from the ambient pressure.
123
Subsonic Pressure Survey
y = 0.94734x  138.18223
28,500
29,000
29,500
30,000
30,500
31,000
31,500
30,500 31,000 31,500 32,000 32,500 33,000
GPS Altitude (ft)
P
r
e
s
s
u
r
e
A
l
t
i
t
u
d
e
(
f
t
)
Figure 10.6 Pressure Survey
Delta H versus Indicated Mach Number
200
0
200
400
600
800
1,000
1,200
1,400
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3
Indicated Mach Number
D
e
l
t
a
H
(
f
t
)
Cloverleaf Delta H data
Cloverleaf Delta V data
Accel Delta H Method
Accel Delta V Method
Decel Delta H Method
Decel Delta V Method
Figure 10.7 AccelDecel Delta H
124
Delta P/qcic versus Indicated Mach Number
0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3
Indicated Mach Number
D
e
l
t
a
P
/
q
c
i
c
Cloverleaf Delta H data
Cloverleaf Delta V data
Accel Delta H Method
Accel Delta V Method
Decel Delta H Method
Decel Delta V Method
Figure 10.8 AccelDecel Position Error Coefficient
Section 10.7 is an edited portion of a paper titled, “PitotStatic Calibration Using a GPS
MultiTrack Method” (Reference 10.2). This method is more commonly referred to as the
cloverleaf method.
10.7 The Cloverleaf Method  Introduction
In the early 1970's, the AFFTC developed a new method to calibrate airspeed, References
10.3 and 10.4. The method was originally dubbed the cloverleaf method due to the pattern
prescribed in the sky. The idea is as follows: One assumes that wind remains constant while
the aircraft performs consecutive turns to produce three passes through a common airmass.
Ideally, the passes should be equally spaced in heading (or 120 degrees apart) and at the same
indicated airspeed. Besides the two components of wind (north and east), there would be an
unknown error in true airspeed that would need to be computed. This handbook will present
the mathematics of this method and some substantiating data. They involve the solution of
three nonlinear equations in three unknowns. It does not require that each pass be executed at
the exact same airspeed or at precisely 120 degrees apart. The National Test Pilot School
(NTPS), in Mojave, California, for instance, uses a method where the passes are 90 degrees
apart, making the math much simpler (Reference 10.5).
The development that makes this method dramatically more economical for flight test is
GPS. One no longer needs to track the aircraft with radar, which reduces test time and
required test resources, and there is a reduced cost for data processing. The method has been
applied with reasonable success by the NTPS. What this handbook will contribute beyond
that which the NTPS has already contributed, is the nonlinear mathematical solution. The test
125
points do not have to be flown as precisely, since the heading angles do not have to be
exactly 90 degrees apart.
This handbook will not discuss the theory and operation the GPS system. In addition, it
will not discuss air data systems at any length. Both subjects have been written about at
length. See for instance, the U.S. Navy web site http://tycho.usno.navy.mil/gps.html. In
addition, the references and bibliography contain just a few of the numerous information
sources on these topics. For the sake of this handbook, the primary piece of information
required of GPS is the accuracy of the velocities and at what update rate they are available.
The military specification for velocity is 0.10 meters per second (0.19 knot). The data in this
handbook was available at 1 sample per second.
This handbook will attempt to explain and demonstrate the validity of a method to
calibrate true airspeed (
t
V ), which invokes the principle that the vector sum of groundspeed
plus windspeed is equal to airspeed. The terminology ‘true’ airspeed is used to avoid the
confusion with the cockpit indicator readings, which are referred to as ‘calibrated’ airspeed
(
C
V ). For those not familiar with calibrated airspeed, the cockpit airspeed indicator only
measures actual airspeed on a standard day (59 degrees F) at sea level standard pressure
(2116.22 psf). The cockpit indicator, historically, could be constructed mechanically with
only one pressure input. That input is a differential pressure between total and ambient
pressure. The true airspeed,
t
V , on the other hand, is more complex. True airspeed (
t
V )
requires computations involving total pressure (
t
P ), ambient pressure ( P ), and total
temperature (
t
T ).
By solving three equations in three unknowns, it will be shown how one can derive the
unknown error in
t
V and the north and east components of wind. Since it is easier to relate to
windspeed magnitude (
w
V ) and direction (
w
ψ ), the north and east components will be
converted to magnitude and direction.
10.8 The Flight Maneuver
Figure 10.8 illustrates a sequence of cloverleaf maneuvers. The test is performed by first
collecting stable data along a heading of
1
ψ . Only a few seconds of data are required to
acquire average airspeed and groundspeed data. Then a righthand turn to a heading of
2
ψ is
accomplished and repeats another data collection. A final righthand turn ends up at a
heading of
3
ψ and a final collection of data. The whole sequence should be performed in one
continuous sequence. Lefthand turns could also be used. In that case, the heading sequence
would be 1,3,2 instead of the 1,2,3 sequence for the right hand turns. The aircraft was flown
on heading, but the data reduction involves track angle. Heading is the direction the aircraft is
pointing while track is the angle of the aircraft groundspeed vector. Heading could also be
considered the direction of the true airspeed vector when the sideslip angle is zero.
126
Figure 10.9 Cloverleaf Flight Maneuver
On 19 August 1997, three cloverleaf runs were performed using an AFFTC F15B pacer
aircraft, USAF S/N 132 (Figure 10.10). A discussion of pacer aircraft can be found in
References 10.1 and 10.6. These runs were performed at nominal indicated conditions of
30,000 feet pressure altitude and indicated Mach numbers of 0.6, 0.7, and 0.8. Each run
consisted of three separate passes at track angles about 120 degrees apart. In round numbers,
the first pass was at a track angle of 15 degrees (NE quadrant). Then a lefthand turn was
performed bringing the aircraft around to a track angle of 255 degrees (SW quadrant). Finally,
a second righthand turn was performed to a track angle of 135 degrees (SE quadrant). Notice
that the headings are separated by the ideal value of 120 degrees. If the data were acquired at
roughly equally spaced angles, then the method should produce reasonable results. The NTPS,
in fact, has demonstrated that a separation of 90 degrees produces quite adequate results.
Figure 10.10 Air Force Flight Test Center F15 Pacer
10.9 Error Analysis
This method is a true airspeed calibration method. There are five measurements: total
pressure (
t
P ), ambient pressure ( P ), total temperature (
t
T ), ground speed (
g
V ), and track angle
(
g
σ ). The first two measurements come from pressure transducers. In many cases, the data
127
source may be altitude and airspeed. In that case, total and static pressure are computed from
altitude and airspeed. The third one is from a total temperature probe. The last two parameters
are either GPS or radar measurements. The laboratory calibration accuracy for pressure
transducers is about t 0.001 in. Hg (0.071 psf) and about t 0.10 °K for temperature probes.
Therefore, one will use these numbers and pick a typical condition near the test conditions of the
data shown in this handbook.
a. Mach number = 0.800,
b. Pressure Altitude = 30,000 feet, and
c. Ambient Temperature = 242.0 °K.
At those conditions (and carrying out computations to beyond usual resolution):
a.
t
P = 957.944 psf,
b.
a
P = 628.432 psf,
c.
t
T = 272.98 °K, and
d.
t
V
= 484.959 knots (true airspeed).
Since we are working with two different units on pressure, the conversion factor is as follows:
a. in. Hg = 70.726 psf
add 0.001 in. Hg "error" to
t
P
b. P
t
= 958.0147
computing true airspeed
c.
t
V = 484.999 knots.
The error in computed true airspeed for an error in total pressure then is:
d. (
t
V ∆ )/(
t
P ∆ ) = (484.999  484.959)/(958.0147957.944) = 0.565 (knots/psf) = 0.044
knots per 0.001 in. Hg Total Pressure.
Hence, for the laboratory accuracy of 1milliinch of mercury (0.001 in. Hg) the error in
total pressure results in a 0.044knot error in true airspeed. Keep in mind this is the error
slope at just this one set of conditions.
To examine ambient pressure errors, add the same error (0.001 in. Hg) to ambient
pressure, while keeping the other parameters the same.
128
a. P = 628.5027,
b.
t
V = 484.898, then,
c. ( /
t
V P ∆ ∆ )= (484.898  484.959)/(628.5027628.432) = 0.861 (knots/psf) = 0.067
knots per 0.001 in. Hg Ambient Pressure.
A 0.1degree error in total temperature produces a true airspeed error as follows:
a.
t
V = 485.048,
b. ( /
t t
V T ∆ ) = (485.048484.959)/(0.1) = 0.89 (knots/deg K) = 0.089 knots per 0.1 °K
Total Temperature.
For this particular flight condition, an error in the aircraft parameters equal to their
laboratory accuracies would produce errors in
t
V of less than 0.1 knot. For the AFFTC data,
some of the results will be presented to greater than 0.1knot resolution, but this does not
imply that that accuracy level has been achieved.
Errors in ground speed will produce errors in true airspeed proportional to the error in the
ground speed on each leg of the method. The ground speed error is likely to be just the
readability of the data. In the case of using a hand held GPS unit, the error in each leg might
be either to the nearest knot or to the nearest onetenth of a knot.
10.10 Air Force Flight Test Center Data Set
The results for the 19 August 1997 data are summarized in Tables 10.1 through 10.3. Note that
the numbers are displayed to at least one digit more than their accuracy level.
Table 10.1
AIRCRAFT AVERAGE MEASUREMENTS AND PARAMETERS
Run
Number
ti
P
(psf)
si
P
(psf)
ti
T
(deg K)
Ci
H
(ft)
Ci
V
(kts)
i
T
(deg K)
1 806.375 635.606 260.1 29,750 222.1 243.0
2 878.482 637.459 266.5 29,686 261.7 243.2
3 985.959 639.174 275.7 29,627 311.4 243.6
Note: The subscript i denotes indicated value.
Table 10.2
INERTIAL SPEEDS (GPS)
Run
Number
ga
V
(kts)
ga
σ
(deg)
gb
V
(kts)
gb
σ
(deg)
gc
V
(kts)
gc
σ
(deg)
1 409.65 18.39 326.41 257.76 370.26 127.14
2 471.22 16.48 390.51 258.08 431.83 127.80
3 545.07 16.74 465.88 257.20 506.79 128.23
Notes: 1. Subscripts a, b, and c denote separate passes.
2. Runs 2a and 2b used radar data.
129
Table 10.3
OUTPUTS
Run
Number
i
M
M
t
V ∆
(kts)
w
V
(kts)
w
ψ
(deg)
T
(°K)
C
H
(ft)
C
H ∆
(ft)
C
V ∆
(kts)
/
Cic
P q ∆
1 0.5947 0.6054 6.07 48.01 223.74 242.4 29,935 185 3.32 0.03098
2 0.6927 0.7088 8.94 46.93 222.54 242.2 30,004 318 4.73 0.03793
3 0.8119 0.8322 10.87 45.86 223.86 242.1 30,080 453 5.49 0.03759
The pacer corrections are known to a high degree of accuracy. These corrections are in
the form of a curve of the parameter /
Cic
P q ∆ versus indicated Mach number. This parameter
is often referred to as the position error parameter. These corrections are applied to pacer data
any time the pacer is used to calibrate another aircraft. Figure 10.11 is a plot of the three
cloverleaf data points with a comparison with the pacer curve.
F15 Pacer Position Error
0.020
0.025
0.030
0.035
0.040
0.045
0.050
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95
Indicated Mach Number
d
e
l
t
a
P
/
q
c
i
c
Pacer Curve
Cloverleaf Points
Poly. (Pacer Curve)
Figure 10.11 Position Error
Groundspeed time histories for run number one are depicted in Figures 10.12 through
10.14. Run number one consists of three separate passes (1a, 1b, and 1c). They are at the
same aim airspeed but at different groundspeeds. These compare radar data and GPS data,
both of which have been smoothed in this case with a 19point secondorder polynomial
curve fit.
130
F15: Run 1a
405
406
407
408
409
410
411
412
413
0 10 20 30 40 50 60
Elapsed time (sec)
G
r
o
u
n
d
s
p
e
e
d
(
k
t
s
)
Radar
GPS
Figure 10.12 Groundspeed – Run 1a
F15: Run 1b
323
324
325
326
327
328
329
0 10 20 30 40 50 60
Elapsed time (sec)
G
r
o
u
n
d
s
p
e
e
d
(
k
t
s
)
Radar
GPS
Figure 10.13 Groundspeed – Run 1b
131
F15: Run 1c
365
366
367
368
369
370
371
372
373
0 10 20 30 40 50 60
Elapsed time (sec)
G
r
o
u
n
d
s
p
e
e
d
(
k
t
s
)
Radar
GPS
Figure 10.14 Groundspeed – Run 1c
For the first run (number 1a), Figure 10.15 illustrates a comparison of true airspeed. The
pacer aircraft has a direct output of corrected true airspeed. This is compared to a
computation of true airspeed from GPS groundspeed plus the computed windspeed.
F15 Run1a True Airspeed
362
363
364
365
366
367
368
369
370
0 10 20 30 40 50 60
Elapsed Time (sec)
T
r
u
e
A
i
r
s
p
e
e
d
(
k
t
s
)
GPS plus Wind
Corrected Aircraft PitotStatic
Figure 10.15 True Airspeed
An interesting observation is that as long as the error in airspeed is the same on each leg, the
computed value of wind will be identical. That means one could use this technique to “measure”
winds; “measure” since one would actually compute the winds rather than measure them.
132
From the start of the first pass (1a) to the completion of the last pass (3c) was 37 minutes. This
was an excessive amount of time for these tests. It seems clear that something considerably less than
a full minute of data on each pass would be quite adequate. A 10second average would suffice.
Then, by relaxing the requirement to maintain the test airspeed exactly, an additional amount of test
time could be saved. Without the need for radar, tracking it becomes unnecessary to coordinate with
the radar tracking team and that saves even more time. It seems reasonable that a factor of two or
more savings in flight time could be achieved. Thus, not counting the time required to climb to the
test altitude, each set of three passes could be concluded in about 5 minutes or less.
10.11 Mathematics of the Cloverleaf Method
The basic vector equation that one will solve for the cloverleaf method is nothing more than true
airspeed equals the vector sum of groundspeed and windspeed.
t g w
V V V · +
! ! !
(10.13)
tN gN wN
V V V · + (10.14)
tE gE wE
V V V · + (10.15)
t ti t
V V V · + ∆ (10.16)
The north and east components of groundspeed are either direct outputs of the GPS or are
computed as follows:
cos( )
gN g g
V V σ · ⋅ (10.17)
sin( )
gE g g
V V σ · ⋅ (10.18)
The aircraft track angle (or the direction of the groundspeed vector) is
g
σ . Writing down
the relationship that true airspeed squared is equal to the sum of the squares of its
components.
2 2 2
t tN tE
V V V · + (10.20)
Substituting equations 10.14 through 10.16 into equation 10.20 yields equation 10.21.
2 2 2
( ) ( ) ( )
ti t gN wN gE wE
V V V V V V + ∆ · + + + (10.21)
Multiplying out equation 10.21 and collecting terms, one gets:
(2 ) (2 )
t ti t wN gN wN
V V V V V V ∆ ⋅ ⋅ + ∆ − ⋅ ⋅ +
2 2
(2 ) ( )
wE gE wE g ti
V V V V V − ⋅ ⋅ + · − (10.22)
133
Defining the following:
a.
t
x V · ∆
b.
wN
y V ·
c.
wE
z V ·
d.
2 2
g ti
C V V · −
1 2 2
ti t ti
A V V V x · ⋅ + ∆ · ⋅ + (10.23)
2 2 2
gN wN gN
A V V V y · ⋅ + · ⋅ + (10.24)
3 2 2
gE wE gE
A V V V z · ⋅ + · ⋅ + (10.25)
Each pass produces an equation. As show in equation 10.26, subscript 1 is the first pass, 2
is the second, and 3 is the third. The unknowns , and x y z are presumed constant for all three
runs. In matrix form, the equations are as follows:
1 1 1 1
2 2 2 2
3 3 3 3
1 2 3
1 2 3
1 2 3
A A A x C
A A A y C
A A A z C
− − ] ¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
]
− − ⋅ ·
' ' ' '
]
¹ ¹ ¹ ¹
] − −
] ¹ ¹ ¹ ¹
(10.26)
In matrix shorthand form:
[ ] { ¦ { ¦ A X C ⋅ · (10.26)
The vector of unknowns { ¦ X is solved by multiplying each side of equation 10.26 by
the inverse of the [ ]
A matrix.
{ ¦ [ ] { ¦
1
X A C
−
· ⋅ (10.27)
The unknowns , and x y z in the
{ ¦ X are also contained in [ ]
A . So an iteration is
required. The initial estimates for the X values will be zero. Then, the matrix equation is
used to compute a new set of X values. These values are inserted into [ ]
A , [ ]
A is inverted
again, and equation 10.27 is used again. Repeat the process until convergence occurs. When
the iteration is complete you have solved for the desired numbers, namely an error in true
airspeed and two components of wind.
134
SECTION 10 REFERENCES
10.1. Albert G. DeAnda, AFFTC Standard Airspeed Calibration Procedures, AFFTCTIH815,
Air Force Flight Test Center, Edwards AFB, California, June 1981.
10.2 Olson, Wayne M. 1998. “PitotStatic Calibration Using a GPS MultiTrack Method.” Paper
presented at the 29
th
Annual Symposium of the Society of Flight Test Engineers (SFTE),
Reno, September 15.
10.3. Wayne M. Olson, “True Airspeed Calibration Using Three Radar Passes,” Performance and
Flying Qualities Branch Office Memo, Air Force Flight Test Center, Edwards AFB,
California, August 1976.
10.4. J.A.Lawford and K.R.Nipress, “Calibration of Air Data Systems and Flow Direction
Sensors,” pages 1620, AGARD AG300Vol.1, September 1983.
10.5. Gregory V. Lewis, “A Flight Test Technique Using GPS For Position Error Correction
Testing,” National Test Pilot School, Mojave, California, July 1997.
10.6. William Gracey, "Measurement of Aircraft Speed and Altitude,” John Wiley and Sons, 1981.
135
11.0 CRUISE
11.1 Introduction
Cruise performance is usually considered the most important test performed during the
performance testing phase. Especially for transport and bomber aircraft since most of the fuel
consumed during a typical mission is during stabilized cruise. For accurate mission planning,
it is critical to be able to predict fuel consumption. Cruise testing was also the most time
consuming test for transport and bomber aircraft. Even for fighter aircraft, it was a significant
portion of the performance flight test program. The emphasis is on was, as efforts are being
made to reduce the amount of flight time spent collecting cruise performance data.
The primary parameters in cruise performance are specific range ( SR ) and range factor
( RF ). Specific range is nautical air miles per pound of fuel used. Range factor is specific
range multiplied by gross weight.
A typical cruise data point can take up to 10 minutes to perform. This is usually required
for engine and aircraft stabilization. The typical stabilization requirement is an airspeed
change of 1 knot per minute. This is equivalent to roughly 0.001 g in flight path acceleration,
which is roughly 1 percent in drag or fuel flow. A simple example will show this 1percent
factor. For a transport category aircraft, a typical lift to drag ratio is an even 10.
a. / 10 / 0.10 L D or D L · ·
b.
t
L W ≅ / 0.10
t
D W · 0.10
t
D W · ⋅
c.
n ex
D F F · −
d. 0.001
x
N · 0.001
t
D W ∆ · ⋅
e.
0.001
0.01 1.0%
0.10
t
t
W D
or
D W
− ⋅ ∆
· · − −
⋅
For nonafterburner operation, a 1percent change in drag will equate to about a 1percent
change in fuel flow. We strive for an accuracy of 1 percent in cruise performance. There are
many sources of error, which add up to this 1 percent. We have errors in gross weight,
pressure altitude, Mach number, ambient temperature, fuel flow, and flight path acceleration.
The main sources of error are in the last two: fuel flow and flight path acceleration. With
modern instrumentation (as of the writing of this handbook), we have been achieving at least
1percent uncertainty in fuel flow. With an INS, we have computed flight path acceleration
(
x
N ) to better than 0.001 g. By using INS data, we no longer have to spend 10 minutes to get
the aircraft perfectly stabilized because we can accurately measure any small acceleration and
make accurate corrections to the data. The other reason for 10minute speed power points
was to get the engine perfectly stabilized. During a series of cruise points, the pilot made only
small throttle changes between points and kept the throttle fixed at near constant flight
conditions for several minutes so very long stabilization periods should not be required with
modern engines.
136
11.2 Cruise Tests
Cruise tests are done to determine aircraft range and endurance and to help in the
development of drag, thrust, and fuel flow relationships. Cruise is a wings level, constant
altitude, and constant speed maneuver. Testing is often accomplished by testing a matrix of
constant aircraft gross weightpressure ratio ( /
t
W δ ) points. The altitude is varied between
points to yield an average /
t
W δ to be a specified value. It is, however, an approximation
that constant /
t
W δ generalizes the data in any way. There are altitude effects on the data.
The preferred method is to do constant altitude testing at varying gross weights to cover a
range of /
t
W δ and altitude. The data could be corrected to nominal /
t
W δ values, but by
correcting to weight and altitude it is easier to make flight manual comparisons.
Table 11.1 represents B52G data. The G model has turbojet engines that were 1950's
vintage.
Table 11.1
B52G CRUISE DATA
Altitude
(ft)
Weight
(lbs)
Specific Range
(nm/lb)
Range Factor
(nm)
35,000 400,017 0.0242 9,680
50,000 194,574 0.0437 8,503
Note: The cruise condition was 1.7 million pounds /
t
W δ
and Mach number = 0.76.
The average degradation in range factor for the B52G is 0.81 percent per 1,000 feet of
altitude increase.
In the case of the B52H model, the average degradation in range factor is 0.56 percent
per 1,000 feet of altitude increase. Another data point is early F16A data that indicated about
a 0.50 percent per thousandfoot degradation factor. The F16A is not a longrange aircraft
and as such had a much smaller fuel fraction. Fuel fraction is the ratio of total fuel weight at
engine start to empty gross weight.
Points are flown by stabilizing as nearly as possible to aim airspeed and altitude, typically ±0.01
Mach number and ±100 feet of altitude. The usual stabilization criterion is 1 knot per minute in
airspeed and 50 feet per minute in altitude. With an INS to compute aircraft acceleration, the
stabilization criterion could be relaxed somewhat. Typically, it takes up to 10 minutes to get the
aircraft stabilized followed by 30 seconds to 1 minute of recorded data. Cruise testing is very time
consuming with this method. By relaxing the stabilization criterion, considerable savings in time
could be achieved. In addition, a realtime display of computed flight path acceleration could be
useful in reducing the time required to stabilize.
11.3 Range
The computation of range ( R ) during cruise is the integration of true airspeed as follows:
t
R V dt · ⋅
∫
(11.1)
137
where:
dt = time increment (hours), and
R = range (nam [nautical air miles]), 6,076.115 feet = 1 nm (1,852 meters, exactly).
We could put the range equation in different forms by making some substitutions. First, we
want to put Mach number ( M ) into the equation by using the Mach number equation as detailed in
the airspeed section of this handbook.
a.
t
V
M
a
· , and
b. 661.48
SL
a a θ θ · ⋅ · ⋅ .
Substituting into the range equation.
( )
661.48 R M dt θ · ⋅ ⋅ ⋅
∫
(11.2)
Defining fuel flow as the negative of the rate of change of weight:
t
f
dW
W
dt
 `
· −
. ,
(11.3)
where:
f
W = fuel flow (pounds/hour), and
t
dW = incremental weight (pounds).
1
t
f
dt dW
W
 `
· − ⋅
. ,
(11.4)
Substituting for equation 11.4 into equation 11.2:
661.48
t
f
M
R dW
W
θ
 `
⋅ ⋅
· − ⋅
. ,
∫
(11.5)
Making these substitutions:
f
f
W
W δ θ
δ θ
 `
· ⋅ ⋅
⋅
. ,
(11.6)
/
t
t
W
W δ
δ
 `
·
. ,
(11.7)
138
f
t
f
t
W
W
W
W
θ
δ θ
δ
 `
 `
· ⋅ ⋅
 ` ⋅
. ,
. ,
. ,
(11.8)
The integration is from a start weight (
ts
W ) to an end weight (
te
W ).
661.48
te
ts
W
W
f
t
t
M
R dt
W
W
W
θ
θ
δ θ
δ
 `
⋅ ⋅
· − ⋅
 `
 `
⋅ ⋅
 ` ⋅
. ,
. , . , . ,
∫
(11.9)
It’s not as bad as it looks. Canceling the θ terms and putting
t
W under dt :
( )
661.48
te
ts
t
W
t W
f
W
M
dt
R
W
W
δ
δ θ
 `
⋅ ⋅
. ,
· − ⋅
 `
⋅
. ,
∫
(11.10)
If one were to fly constant Mach number and maintain constant /
t
W δ , then the
numerator term could be brought out of the integral. This would involve a slow cruise climb
and we will show how much extra thrust that requires. At constant /
t
W δ and M , the lift
coefficient would be a constant. Then, ignoring the change in skin friction drag with altitude,
the drag coefficient will be constant. Ignoring the thrust component, drag coefficient (as
derived in the lift and drag section) is as follows:
( )
2
/
0.000675
n
D
F
C
M S
δ
· ⋅
⋅
(11.11)
Then /
n
F δ will be constant, since we have assumed that Mach number and
D
C are constant.
The corrected thrust specific fuel consumption relation is as follows:
( )
/
f
f
n
n
W
W
tsfc
F
F
δ θ
θ
θ
δ
 `
⋅
. ,
· ·
 ` ⋅
. ,
(11.12)
139
We have presumed the denominator ( /
n
F δ ) to be a constant. The / tsfc θ is also
considered to be approximately a constant at constant Mach number and /
n
F δ . Now, we can
pull these (approximately) constant terms out of the integral and integrate.
( )
661.48
te
ts
t
W
t W
f
W
M
dt
R
W
W
δ
δ θ
 `
⋅ ⋅
. ,
· −
 `
⋅
. ,
∫
(11.13)
The term in front of the integral is called range factor ( RF ).
te
ts
W
t W
dt
R RF
W
· − ⋅
∫
(11.14)
You may be more used to seeing RF in the following identical form:
t
t t
f
V
RF W SR W
W
· ⋅ · ⋅ (nautical air miles) (11.15)
where:
SR · specific range (nautical air miles per pound of fuel).
From a table of integrals and natural logarithm relationships:
( ) ( )
ln ln ln ln
b
a
dx
b a
b a
a b
x
· − · · −
∫
where:
ln = natural logarithm.
ln
ts
te
W
R RF
W
 `
· ⋅
. ,
(11.16)
The above equation is convenient to get a quick estimate of range given only the average
range factor and the start and end cruise weight. Note that this is the range during the cruise
segment and does not include taxi, takeoff, climb, and descent.
11.4 Computing Range from Range Factor
Using the previous tabulated B52G data, we will compute range and show the magnitude
of the climb factor. We will assume that the two points at 35,000 and 50,000 feet are the
beginning and end of the cruise segment of a mission. The cruise is at constant 0.77 Mach
140
number and a /
t
W δ of 1,700,000 pounds. Using previously defined formulas for true
airspeed, energy altitude, and pressure ratio we construct Table 11.2. We will linearly
interpolate versus altitude for range factor.
Table 11.2
RANGE FACTOR VERSUS ALTITUDE FOR B52G
Altitude
(ft)
True
Airspeed
(kts)
Energy
Altitude
(ft)
Gross
Weight
(lbs)
Net
Thrust
(lbs)
Range
Factor
(nm)
35,000 443.84 43,721 423,547 42,355 10,843
36,089 441.65 44,724 402,052 40,205 10,777
40,000 441.65 48,635 333,155 33,316 10,539
45,000 441.65 53,635 261,986 26,199 10,234
50,000 441.65 58,635 206,020 20,602 9,930
Note: Thrust was computed by assuming a lift to drag (L/D) ratio of 10. This is typical
for a transport category aircraft.
We could get a first estimate of range by using an average range factor and the start and
end conditions.
( ) 9, 680 8, 503
400, 016
ln ln 6, 552 nam
2 194, 574
s
e
W
R RF
W
+  `  `
· ⋅ · ⋅ ·
. , . ,
(11.17)
Since we assumed a linear variation of range factor with altitude, we will get the same
result by integrating the individual segments. Range factor will not be a linear function of
altitude, usually.
The time for this mission computes to be 54,100 seconds (15.04 hours). From the table,
the delta energy altitude is 14,914 feet. The average speed is 736.5 feet per second. Now, we
can calculate the average longitudinal load factor necessary to produce enough excess thrust
to sustain this cruise climb.
( )
( )
14, 914
51, 000
0.2955
0.00040
736.5 745.6
E
x
t
H
N
V
· · · ·
!
(11.18)
At the average weight of 297,295 pounds, the average excess thrust calculates to 119 pounds.
The average thrust is 29,730 pounds, therefore the ratio of excess thrust to net thrust is:
a.
119
0.0040 0.40%
29, 730
ex
n
F
or
F
· ·
By ignoring the excess thrust, we over estimated the range by 26 nam (0.40 percent of
6,552 nam). Quite small, but not negligible. On an actual mission, the mission profile would
be step climbs. For this example, you would start the cruise segment at 35,000 feet and fly
141
constant altitude until it was decided to climb to a new altitude. This might be in increments
of 4,000 feet. When flying in civilian airspace, the altitudes are 4,000 feet apart.
11.5 Constant Altitude Method of Cruise Testing
The recommended method of doing cruise testing is the constant altitude method. The
F15 and F16 projects used constant altitude method. The B1B used constant altitude
analysis method, though the points were flown using the constant weight/pressure ratio
( /
t
W δ ) method. The constant altitude method consists of choosing a range of weight and
altitude conditions to cover the aircraft envelope and then flying each weight/altitude
combination over a range of speeds. For an aircraft with a large weight fraction, this may
mean flying up to six altitudes at up to three weights (heavy, mid, and light). This could mean
a maximum of 18 weight/altitude combinations. Nevertheless, with a reasonable amount of
thrust/drag/fuel flow analysis, this could be cut in half or more. Flying all three weights at the
predicted optimum cruise /
t
W δ is usually desirable. The altitudes are chosen by selecting
six evenly spaced /
t
W δ ’s from minimum to maximum with one at the predicted optimum.
The minimum is based upon minimum weight at a minimum altitude and the maximum is
based upon the cruise ceiling defined as a climb capability of 300 feet per minute. The
altitudes are then rounded to the nearest 5,000 feet, which allows for easy flight manual
comparisons since flight manuals typically have cruise charts at even 5,000foot increments.
For ease of flight manual comparisons, the data presented in reports are a specific range,
or range factor versus Mach number at even 5,000foot increments for standard weights,
representing rounded values of heavy, mid, and light gross weight.
11.6 Range Mission
Range missions are performed to gain confidence in the performance data collected
during climb, cruise, and descent. Rather than relying on fuel flow measurements and
thrust/drag analysis, the primary measurement during a range mission is aircraft fuel quantity
indications. The mission is performed by climbing to a given start cruise altitude,
progressively stepping up in the altitude during constant altitude/Mach number cruise
segments, and finally doing an idle power descent. Total fuel used is obtained from the fuel
quantity system. A calibration of the fuel quantity system is obtained during the aircraft
empty weight and fuel calibration. Using a performance simulation, the test day mission
performance could be estimated. The simulation thrust/drag/fuel relationships were
previously determined using data from several maneuvers including climb, cruise, and
descent. The simulation estimates of fuel used were compared with measured fuel used
during the mission.
A practical reality of the flight test programs was that it was difficult to justify devoting
an entire sortie to only a range mission. A compromise was to obtain fuelused data during
long cruise segments that often occurred during certain systems tests. During the B1B
project, fuel used data were acquired from several training sorties flown on production
aircraft at Dyess AFB, Texas. The data came from constant airspeed/altitude segments of
several hours in duration. A comparison of fuel used was made with simulation results. The
differences were well within the oftenquoted 3percent accuracy for performance data. This
provided a valuable confirmation of the flight test results.
142
11.7 Slow AccelDecel
A supplement, or perhaps even an alternative to cruise testing, is to do slow accels and
decels. The data are used to build or verify a thrust versus fuel flow model. In addition, the
data could be standardized to zero excess thrust. The maneuvers are flown sufficiently slowly
to make the maximum correction to a range factor of about 10 percent. This compared with
1percent corrections made to cruise data. We could estimate the zero excess thrust range
factor from both the accel maneuver and the decel maneuver. The average of the accel and
decel standardized range factors is a good estimate of zero excess thrust range factor since
relatively small corrections are being made.
The maneuver is done at a rate of less than 1 knot per 3 seconds to yield an accel/decel
rate of about 20 times the cruise stabilization criterion. A typical accel/decel maneuver takes
about 6 to 12 minutes. The throttle is moved in small increments during the run to keep the
accel/decel rate small, but not so small that the maneuver would take too long, thereby losing
the advantage over stabilized cruise. If the cruise tests are done with a relaxed stabilization
criterion (±100 feet and ±2 knots in 20 seconds) with only 20 seconds of recorded data, then
the dynamic cruise has an advantage over the slow acceldecel data. If it is desired to collect,
thrust and fuel flow data over a range of conditions then the slow acceldecel is a good
approach.
11.8 Effect of Wind on Range
The typical high altitude cruise for both fighter and transport aircraft is about 0.85 Mach
number. The true airspeed for standard day in the lower atmosphere (troposphere) and upper
atmosphere (stratosphere) can be computed using formulas from the airspeed section. For
standard day from 11 kilometers (36,089 feet) to 20 kilometers (65,617 feet), the temperature
is 216.65 degrees K.
a.
216.65
661.48 0.85 487.5
288.15
t
V · ⋅ ⋅ · knots
The formula for specific range (nams per pound of fuel) is just true airspeed (
t
V ) over
fuel flow (
f
W ).
t
f
V
SR
W
· (11.19)
We can compute a specific range with respect to the ground as follows:
g
g
f
V
SR
W
· (11.20)
Since groundspeed equals true airspeed minus wind and taking just the component
parallel to the direction of flight (track angle):
t g w
V V V · + (11.21)
143
( )
t w
g
f
V V
SR
W
−
· (11.22)
Finally, the ratio of specific range with respect to the ground to the specific range with
respect to the moving air mass (equation 11.22 divided by equation 11.19) is as follows:
( )
g t w
t
SR V V
SR V
−
· (11.23)
As shown in Appendix A, windspeed at an ambient pressure of 200 millibars (mb)
(38,661 feet) averages about 40 knots above Edwards AFB. The average direction is about
215 degrees (SW). Since wind direction is the direction from which the wind is blowing, an
aircraft heading of 215 degrees would have a 40knot headwind for this average Edwards
wind. A headwind is a positive wind. For this condition, the range degradation would be:
a.
( ) 487.5 40
0.918 8.2 percent
487.5
g
SR
SR
−
· · · degradation
This is for an average wind if one were heading directly into the wind. A set of data
collected for the cloverleaf paper (a portion of which is in the cloverleaf subsection of the air
data system calibration section) had winds in excess of excess of 100 knots. This data were
not included in this handbook, but was AFFTC data set number 2 in the referenced paper
(Reference 10.2). In addition, the wind data shown Appendix A indicates a standard
deviation of about 25 knots. Flying directly into a 100knot wind would produce the
following specific range degradation:
a.
( ) 487.5 100
0.795 20.5 percent
487.5
g
SR
SR
−
· · · degradation
One could just as easily be flying with that wind as a tailwind.
a.
( ) 487.5 100
1.205 20.5 percent
487.5
g
SR
SR
+
· · · improvement
In general, you would only be affected by the component of wind parallel to the flight
direction. Wind vector relationships are discussed in detail Section 10.11. This wind effect is
only relevant in computing physical (ground) nautical miles with a given wind. When
collecting cruise data, you are flying with respect to the moving air mass.
144
12.0 ACCELERATION AND CLIMB
12.1 Acceleration
Accelerations are conducted for multiple purposes. First, to determine optimum climb
schedules by observing the peak of specific excess power versus Mach number. The actual
optimum occurs to the right of the peak of specific excess power (
s
P ) versus M curves,
depending on whether it is desirable to achieve a minimum time to climb or minimum fuel
for fixed range. Second, to determine the obvious acceleration performance, i.e., fuel used,
time, and distance to accelerate. Third, to determine drag/thrust/fuel flow models. Climb data
can be used for this purpose also, however, accelerations are a more efficient method. The
accelerations are conducted over a range of altitudes.
The acceleration maneuver is performed wings level, 1g, and fixed throttle at constant
altitude. Usually a climb or turn is done at the beginning of the run to get the engine
thermally stabilized. Then the aircraft accelerates to a point where the acceleration rate is
reduced to a small value (less than 1 knot per 10 seconds). The altitude is maintained
constant during the run. Indicated altitude will jump as the aircraft passes through the
transonic speed regime. Thus, it is necessary to maintain zero flight path angle usually by
maintaining pitch attitude (θ ). Once through the transonic jump, an indicated altitude could
be used for the rest of the acceleration. Modern aircraft with a headup display (HUD) and
INS have a velocity vector displayed on the HUD. Level flight through the transonic region is
obtained by maintaining the velocity vector on the horizon.
Figure 12.1 is a sample of some actual acceleration data. The data points have been
corrected to standard conditions. Standard conditions consist of standard weight, pressure
altitude, and standard day atmospheric conditions. The fairing is the result of modeling thrust
and drag, then computing specific excess power from thrust and drag. With one relatively
short maneuver, one obtains a range of speed (Mach number) at a given altitude. By
performing accelerations at various altitudes, climb performance can be computed. However,
a few continuous climbs need to be conducted to confirm that performance (time, distance,
and fuel used) computed from accelerations yields the same result as that from climbs.
Accelerations are also performed at elevated g levels. These are discussed in the turn section.
145
Specific Excees Power (ft/min) versus Mach Number
2,000
2,250
2,500
2,750
3,000
3,250
3,500
3,750
4,000
0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90
Mach Number
S
p
e
c
i
f
i
c
E
x
c
e
s
s
P
o
w
e
r
(
f
t
/
m
i
n
)
Standardized Data
Fairing from Thrust/Drag Model
Figure 12.1 Specific Excess Power from Acceleration
12.2 Climb
The climb maneuver is performed primarily as a check of predicted climb performance
derived from acceleration data. Usually climbs are conducted at flight manualpredicted best
climb speeds. Determination of actual best climb speeds requires an analysis using data from
several sources, which include accelerations. The normal climb is a constant calibrated
airspeed climb to a break altitude above which a Mach number is maintained constant. The
climb continues to a climb ceiling (300 feet per minute rate of climb defined as the cruise
ceiling). Data are standardized to the climb schedule, standard day, standard weight, and
standard normal load factor. Thrust and drag data are obtained during the climb. The data are
reduced at constant altitude increments rather than constant time increments to yield a more
even distribution of data. A standard day rate of climb, time to climb, fuel used, gross weight,
and distance traveled are plotted versus pressure altitude. A flight manual comparison is
accomplished with this data. For high performance aircraft, there may be differences in
performance accelerating through a Mach number/pressure altitude condition versus climbing
through the same condition. This is due to an engine fuel control system lag. This effect
needs to be taken into account. Climbs are usually terminated at the “cruise ceiling.” Climb
ceiling definitions are given in Table 12.1. The definitions are from the flight manual
specification.
146
Table 12.1
CLIMB CEILING DEFINITIONS
Ceiling
Rate of Climb
(ft/min)
Combat 500
Cruise 300
Service 100
Absolute 0
12.3 Sawtooth Climbs
As seen in Appendix B, one can expect to see large changes in windspeed and direction
as a function of altitude. How this would impact climb performance was discussed in the
effect of wind gradient portion of the altitude section. A comparison was made for an average
day above Edwards AFB in January. The difference in delta energy altitude flying directly
into a headwind versus flying directly into a tailwind was 1,308 feet. This was over a
geometric altitude range from 14,605 to 23,937 feet, or a 14percent difference in rate of
climb. Before the advent of accelerometer and INS methods, climb data were attained using
the sawtooth climb method.
The sawtooth climb tests are a series of alternate heading climbs through a given altitude
at a range of speeds. For each speed, a climb would be conducted through the aim altitude
and airspeed and altitude data would be collected versus time. For instance, the aim altitude
might be 5,000 feet pressure altitude. Then test points would be chosen over a range of
speeds to bracket the expected best climb speed. Depending upon the performance level of
the aircraft, a start altitude would be determined. Then, the aircrew would establish a climb
speed and climb power at that altitude and would collect data over an established data range,
perhaps 4,500 to 5,500 feet, for instance. Then, you would descend back to the initial altitude
of 4,000 feet and repeat the same airspeed point, but this time at an opposite heading angle
(based upon magnetic compass). The idea here is that the average of these two points would
be a zero wind gradient condition. Using the acceleration factor, you would correct the data
to zero acceleration. A zero acceleration rate of climb is the rate of change of energy altitude.
A sample of some actual flight test sawtooth climb data from an AC119G (Figure 12.2)
is shown Figure 12.3. Data were obtained from FTCTR694, AC119G Aircraft Limited
Performance and Stability and Control Test (Reference 12.1). This was one of the last
AFFTC projects where sawtooth climbs were flown. The thrust designation METO on
Figure 12.3 denotes Maximum Except for TakeOff.
147
Figure 12.2 AC119G Aircraft
Sawtooth Climbs: AC119G Cruise Configuration METO Power
200
300
400
500
600
700
800
900
1,000
1,100
1,200
1,300
90 100 110 120 130 140 150 160 170
Calibrated Airspeed (kts)
Z
e
r
o
A
c
c
e
l
e
r
a
t
i
o
n
R
a
t
e
o
f
C
l
i
m
b
(
f
t
/
m
i
n
)
10,000 ft; 58,000 lbs
opposite heading
5,000 ft; 58,000 lbs
opposite heading
10,000 ft; 66,000 lbs
opposite heading
5,000 ft; 66,000 lbs
opposite heading
Figure 12.3 AC119G Sawtooth Climb Data
We can take these data points, without distinguishing opposite headings, and present
them in a different manner. Since we had two altitudes and two weights, let us attempt to
minimize the weight effect in the data by computing the excess thrust. Then, take the excess
thrust and divide by the pressure ratio (δ ) to minimize the altitude effect. The data are
presented in Figure 12.4.
ex x t t
t
h
F N W W
V
 `
· ⋅ · ⋅
. ,
!
(12.1)
The h
!
is the zero acceleration rate of climb in Figure 12.3. The specific algorithms used
to standardize that data can be found in AF TR No. 6273, Flight Test Engineering Handbook
(Reference 12.2).
148
Sawtooth Climbs: AC119G: Fex/delta versus Mach Number
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30
Mach Number
E
x
c
e
s
s
T
h
r
u
s
t
/
d
e
l
t
a
(
l
b
s
)
58,000 lbs: 10,000 ft
58,000 lbs: 5,000 ft
66,000 lbs: 10,000 ft
66,000 lbs: 5,000 ft
Figure 12.4 AC119G Excess Thrust Data
12.4 Continuous Climbs
A climb could be done with any number of different climb schedules. A climb schedule is
a speed or attitude variation with altitude. The most common type of climb is one that keeps
calibrated airspeed (
C
V ) constant until a given Mach number ( M ) is reached at which time
Mach number is kept constant. A variation on that schedule is one in which calibrated
airspeed is a function of altitude. Usually, both calibrated airspeed and Mach number may
have been a function of gross weight (
t
W ), but they do not vary during the climb. For high
performance fighters (with installed thrusttoweight ratios greater than 1) the initial part of
the climb may be done at a constant pitch attitude (θ ) transitioning to a Mach number at a
given altitude. Alternatively, the early part of the climb may be performed at less than
maximum thrust. These types of climbs are required for high performance fighters when the
aircraft has a longitudinal acceleration load factor greater than 1.00 and can accelerate flying
straight up. The flight path angle for the constant θ climb is as follows:
γ α θ · − + (12.2)
Other types of climbs are variable climb schedules such as a varying airspeed schedule, a
constant true airspeed climb, or a varying Mach number climb. The C130H climb schedule
is an example of a varying calibrated airspeed climb. At 150,000 pounds gross weight at sea
level the recommended schedule is 181knots calibrated airspeed while at 20,000 feet the
climb speed is down to 166 knots. In contrast, most aircraft use a constant calibrated
airspeed/Mach number climb schedule.
149
Accelerations and climbs are both fixed throttle maneuvers. They are usually done with
power settings like MIL or MAX. Decelerations and descents are usually done in power
settings such as IDLE, though there could have been a MIL power deceleration under certain
conditions such as supersonic.
12.5 Climb Parameters
/ R C H ·
!
0
1
t t
V dV
AF
g dH
 `
 `
· + ⋅
. ,
. ,
(12.3)
where:
/ R C = rate of climb (ft/sec), and
AF = acceleration factor.
12.6 Acceleration Factor (AF)
The acceleration factor ( AF ) is used in climb performance as a simple conversion
between a rate of change of tapeline or geopotential altitude and rate of change of energy
altitude.
a.
E
H
AF
H
·
!
!
Most aircraft climbs are conducted by either holding calibrated airspeed (
C
V ) or Mach
( M ) number constant. In reality, the calibrated airspeed or Mach number is not exactly
constant but let us make some calculations assuming that they are held exactly constant and
that there is zero wind so that true airspeed (
t
V ) and inertial speeds (
g
V ) are identical. The
true airspeed vector defines the flight path (or wind) axis. The component of aircraft
acceleration parallel to the flight path is the longitudinal acceleration (
x
A ). The longitudinal
load factor (
x
N ) is simply the
x
A divided by the acceleration of gravity (
0
g ). In
conventional aircraft performance, g is assumed a constant at the reference gravity and given
the value of 32.174 ft/sec². Figure 12.5 is a representation of acceleration factor for climb at
constant calibrated airspeed.
150
Constant Calibrated Airspeed Acceleration Factor
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000
Pressure Altitude (ft)
A
c
c
e
l
e
r
a
t
i
o
n
F
a
c
t
o
r
Vc = 100 kts
Vc = 200 kts
Vc = 300 kts
Vc = 400 kts
Vc = 500 kts
Vc = 600 kts
Vc = 661.48 kts
Vc = 700 kts
Figure 12.5 Acceleration Factor – Constant Calibrated Airspeed
The discontinuity in Figure 12.5 at 36,089 feet is due to the transition from a temperature
decreasing with altitude to a constant temperature. The above chart is for a standard
atmosphere.
12.6.1 Two Numerical Examples for AF
To illustrate the importance of the concept of AF , let us illustrate AF by two numerical
sample cases. The two cases will cover the range from a highspeed, highaltitude fighter to a
lowspeed, lowaltitude aircraft.
12.6.1.1 Case 1
High speed, high altitude, high performance typical of a fighter type aircraft:
a. For case 1, assume the following flight conditions:
1. H = 30,000 feet, and
2. M = 0.900.
For standard conditions, we could compute the values for calibrated and true airspeed,
using the equations found in the airspeed section of this text. Please note that we are listing
the numbers to at least one more significant figure than our limits of flight test data accuracy.
The following additional significant figures are necessary to make the computations
accurately:
151
1.
C
V = 346.24 knots, and
2.
t
V = 530.39 knots = 895.19 feet/sec.
Then,
b. At 31,000 feet and 0.900 Mach number:
1.
C
V = 338.90 knots, and
2.
t
V = 528.09 knots = 891.31 feet/sec (Note that the aircraft is decelerating while
climbing at a constant Mach number.).
Now we could numerically calculate the AF :
/
t
t
V
dV dH
H
∆
·
∆
( )
( )
( )
0
891.31 895.19
2
891.31 895.19
1 1 0.8923
32.174 31, 000 30, 000
t t
V V
AF
g H
 ` ] +
]
−  ` ∆  `
]
· + ⋅ · + ⋅ ·
∆ −
. ,
. ,
. ,
For a
s
P of 200 feet per second, the / R C would be 224.1 feet second.
200
/ 224.1
0.8923
E
H
R C
AF
· · ·
!
For a climb through 30,000 feet holding a constant calibrated airspeed of 340 knots, the
AF computes to 1.3576 for a / R C of 147.3 feet per second. The difference in rate of climb
between holding constant Mach number versus constant calibrated airspeed is 52 percent.
This illustrates how large an effect the acceleration factor could be and that it certainly needs
to be taken into account. The percentage difference gets proportionately smaller at lower
airspeeds.
12.6.1.2 Case 2
The second case is what is a typical climb for a light aircraft. Assume a 100knot calibrated
airspeed climb through 5,000 feet. The difference in rate of climb between a constant calibrated
airspeed and a constant Mach number climb is now down to just 1.9 percent. At a
s
P of 1,000
fpm, the rate of climb at a constant Mach number is 1,003.7 fpm and the rate of climb at constant
calibrated airspeed is 984.8. This is small, but not small enough to ignore. Below 36,089 feet in
the standard atmosphere, a constant calibrated airspeed climb would be accelerating in true
airspeed and hence, rate of climb would be less than the specific excess power. Conversely,
below 36,089 feet in the standard atmosphere in a constant Mach number climb, the true airspeed
would decrease with increasing altitude (Figure 12.6). Above 36,089 feet, when temperature is a
152
constant with altitude for the standard atmosphere, the true airspeed is a constant for a
constant Mach number. Hence, the acceleration factor would be 1.00 at all Mach numbers.
Keep in mind that Figure 12.6 is for standard day.
Accelertion Factor: Constant Mach Number: H<36,089
0.84
0.86
0.88
0.90
0.92
0.94
0.96
0.98
1.00
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
A
c
c
e
l
e
r
a
t
i
o
n
F
a
c
t
o
r
Figure 12.6 Acceleration Factor – Constant Mach Number
12.7 Normal Load Factor During A Climb
To derive the formula for the normal load factor in a climb, consider the aircraft flying in
a pullup maneuver. Figure 12.7 illustrates the vectors during a pullup. The first velocity
vector (
t
V ) is at a flight path angle of
1
γ . The second
t
V is at
2
γ . The magnitude of the
change is exaggerated, but consider the change infinitesimal. The aircraft rotates about a
point C , with a radius R . The acceleration perpendicular to the flight path (ignoring gravity)
is a centripetal acceleration.
153
Figure 12.7 Centripetal Acceleration Diagram
The centripetal acceleration is as follows:
2
t
V
a
R
· (12.4)
The radius is related to the linear velocity through the angular velocity (ω ).
t
V R ω · ⋅ (12.5)
The angular velocity ω is just the derivative of the flight path angle.
( )
( )
2 1
2 1
d
dt t
t t
γ γ
γ γ
ω γ
−
∆
· · · ·
∆ −
! (12.6)
Solving for the radius R in equation 12.5 and substituting into the acceleration equation 12.4:
2
t
t
t
V
a V
V
γ
γ
· · ⋅
 `
. ,
!
!
(12.7)
Adding in the component of gravity yields:
0
cos
t
a g V γ γ · ⋅ + ⋅ !
(12.8)
Finally, dividing by
0
g yields the load factor in the normal axis.
0
cos
t
z
V
N
g
γ
γ
⋅
· +
!
(12.9)
C
154
The above equations are valid for constant winds. Usually, the load factors are computed
from INS velocities and angles plus true airspeed to enable a transformation from the inertial
axis to the flight path axis. What is desired are inertial accelerations in the wind (or flight
path) axis. Therefore, if the aircraft has an INS, and the appropriate software to do the axis
transformations, then there is no need to be concerned about horizontal winds and wind
gradients. In addition, the difference between a tapeline rate of climb and pressure altitude
rate of climb is taken into account, since the INS yields geometric rate of climb. The INS
data is, however, sensitive to the presence of any vertical winds, so efforts are made to fly in
areas where no vertical winds are expected. For Edwards AFB, the best place to conduct
performance tests is over the ocean. Both the B1B and C17A aircraft conducted their entire
cruise testing over the ocean.
12.8 Descent
A typical descent schedule is a constant Mach number intersecting a constant calibrated
airspeed. The data are used to generate descent performance, an idle thrust map, and drag
polar information to complete the performance model. The performance model is used to
check mission performance. The idle power descent could be accomplished with speed
brakes extended.
12.9 Deceleration
Decelerations are conducted to provide data to compute descent performance. A
deceleration is performed by accelerating to the Mach number limit then moving the throttle
to idle and conducting a wings level, constant altitude deceleration. This maneuver gives us
idle thrust versus speed. Due to inaccuracies in the inflight thrust deck, there could be a drag
difference at idle thrust versus drag polar data acquired at higher power settings. The same
maneuver could be accomplished with the speed brakes extended.
SECTION 12.0 REFERENCES
12.1 Pape, James K. and McDowell, Edward D., AC119G Aircraft Limited Performance
and Stability and Control Tests, FTCTR694, AFFTC, Edwards AFB, California,
March 1969.
12.2 Herrington, Russel M., et al, Flight Test Engineering Handbook, AF TR 6273,
AFFTC, Edwards AFB, California, revised January 1966.
155
13.0 TURNING
13.1 Introduction
Turning performance is defined as flight at other than 1 g, usually in the horizontal plane.
There are four different types of turns: accelerating or decelerating, thrustlimited, stabilized,
and liftlimited.
13.2 Accelerating or Decelerating Turns
Accelerating or decelerating turns are performed at a fixed throttle, constant g, and
constant altitude. For accelerating turns, the maneuver is done by starting fast, applying
specified throttle, and pulling into a turn to decelerate the aircraft. Next, reduce g level to the
specified value and accelerate to either the specified Mach number or the maximum speed.
The data acquired could be used to generate energy maneuverability charts or to contribute to
the aircraft drag, thrust, and fuel flow model.
Turns at fixed g, constant altitude, and fixed throttle are referred to as accelerating or
decelerating turns. Turns, in general, are used to quantify the turning performance capability
of the aircraft and to help in the development of the drag and lift curves. With the advent of
dynamic performance, fewer turns are conducted in flight test. Turns are used primarily to
check the performance model created from 1g acceleration and dynamic performance
maneuvers. Nevertheless, some turns are still necessary as confidence builders in the model
and to demonstrate specification performance.
13.3 ThrustLimited Turns
A thrustlimited turn is a turn where the pilot attempts to maintain throttle setting, Mach
number, and pressure altitude while varying normal load factor. Usually about 30 seconds or
180 degrees of turn data are recorded at stabilized conditions; however, maintaining
stabilized conditions is often difficult. The data are used to verify the thrust/drag model for
sustained g and to assist in the development of the drag and lift curves. The data are collected
at a stabilized g and as such, may be of higher quality than data from dynamic maneuvers.
Nevertheless, keep in mind that the thrustlimited turn is dynamic since it is at elevated g
values (and large pitch rates) and may be at different power settings than the dynamic
performance data. There may have been throttle effects on the drag polar due to inaccuracies
in the inflight thrust computation. One value of thrustlimited turns is it produces thrust data
that is stabilized while accelerations and decelerations are dynamic in thrust. So, the lag time
constant for thrust could be estimated. With fuel controls scheduling on total temperature in
the inlet, there may be a different lag constant depending on whether the aircraft is climbing
or accelerating through a point. The thrustlimited turn is stabilized at a given Mach number
and pressure altitude
condition. As with accelerating or decelerating turns, only a limited
number of sustained or thrustlimited turns are performed because they are very fuel and
timeconsuming tests compared with the more efficient dynamic maneuvers. It is still
necessary to perform a limited number of turns as checks on the model. It has been necessary
on past projects to do significant numbers of turns because of disagreements between turn
data and dynamic data on the drag polar. Developing correlation factors to adjust the drag
polars to match the measured turn performance may be necessary. Not relying completely
upon data obtained from dynamic performance maneuvers is important.
156
Using an INS for flight path accelerations requires a 1g level run be accomplished before
the turn to get a wind calibration. This applies to all turning maneuvers. Winds are
computed from the wind calibration maneuver assuming zero sideslip. These winds are
assumed to remain constant during the turn. The thrust and fuel flow data obtained in climbs
and acceleration is dynamic and subject to engine and instrumentation lag. It is possible to
attain lag time constants by comparing thrustlimited turn data to climb and acceleration data.
13.4 Stabilized Turns
Stabilized turns are turns where Mach number, pressure altitude, and normal load factor
are specified and throttle is varied to obtain a stabilized condition. These maneuvers are
useful to obtain lift and drag data at specific points along the drag and lift curves and to check
for specification compliance. The flight test objective is to determine if such conditions can
be achieved in stabilized flight at something less than or equal to maximum throttle. Another
way to evaluate that spec point would be to do a thrustlimited turn at MAX thrust at the
specified flight conditions and then determine whether the desired normal load factor in
stabilized flight is achieved. Specs are usually written for standard day at a standard weight,
center of gravity, etc. Therefore, you must correct the data to standard conditions to
determine spec compliance since the spec may have been missed on the test day but the
aircraft would have achieved the spec on a more favorable standard day. For the stabilized
turn, you would have needed some specialized software to perform the standardization or the
turn could have been standardized assuming it is an accelerating turn at a given pressure
altitude, Mach number, and normal load factor, then determine the flight path acceleration for
standard conditions. If the longitudinal flight path load factor (
x
N ) was positive for the given
spec conditions, then the spec condition was met.
13.5 LiftLimited Turns
When it is desired to determine limit performance at the angleofattack (α ) limit or the
normal load factor (
z
N ) limit then a liftlimited turn is performed. If the aircraft has an α/g
limiter, as is the case on the F16, then the turn is a full aft stick maneuver. Otherwise, the
pilot must observe the flight manual limits, which makes this maneuver very difficult to fly
without exceeding aircraft limits. The angleofattack limited portion of the maneuver is used
to quantify the lift coefficient at the limit angle of attack and to check the angleofattack
calibration at the limit. The check of angle of attack is performed with INS data. This
maneuver produces data at the highest limits of the drag polar and the lift curve. You also
obtain limited angleofattack data from a splits. The splits maneuver is discussed in the
dynamic performance section.
Liftlimit and glimit turns are accomplished by accelerating to limit speed then pulling
into a maximum allowable g turn and allowing the aircraft to decelerate to the lift limit. This
defines the lift limit and g limit performance. The throttle setting is usually MIL or MAX, but
the maneuver may be done at any power setting. Besides getting limit performance, drag
polar data at or near maximum lift coefficient are obtained.
157
13.6 Turn Equations
13.6.1 Normal Load Factor
The transformation equations for load factors from the body axis system to the flight path
axis are as follows (ignoring sideslip):
cos sin
sin cos
xb x
zb z
N N
N N
α α
α α
¹ ¹ ¹ ¹ ]
· ⋅
' ' ' '
]
−
] ¹ ¹ ¹ ¹
(13.1)
The additional sideslip transformation matrix is given in the Accelerometer Methods
subsection of the Flight Path Accelerations section. The inverse transformation from the
flight path axis to the body axis is as follows:
cos sin
sin cos
xb x
zb z
N N
N N
α α
α α
− ¹ ¹ ¹ ¹ ]
· ⋅
' ' ' '
]
] ¹ ¹ ¹ ¹
(13.2)
where:
x
N = flight path axis longitudinal load factor,
z
N = flight path axis normal load factor,
xb
N = body axis longitudinal load factor, and
zb
N = body axis normal load factor.
For a constant altitude, constant speed turn, the normal load factor in the wind (flight
path) axis system in terms of the turn rate can be derived in a similar manner as the formula
for normal load factor in a climb. There are two components. One, the vertical component is
exactly 1.0, for the ideal case of exactly constant altitude. Two, the horizontal component is a
centripetal acceleration. Figure 13.1 shows these vectors.
z
N
zv
N
zh
N
Figure 13.1 Normal Load Factor Vectors In a Turn
0
t
zh g
V
N
g
σ · ⋅ ! (13.3)
158
( )
2
2 2 2
0
1
t
z zv zh g
V
N N N
g
σ
]
 `
] · + · + ⋅
]
. ,
]
! (13.4)
Where
g
σ is the ground track angle and the assumption of zero wind is made. With the
same idealized assumptions of constant altitude, constant speed, and zero wind, the normal
load factor in terms of the bank angle can be determined as shown in Figure 13.2.
Figure 13.2 Banked Turn Diagram
Where:
1.0
zv
N · , and
1
cos
zv
z z
N
N N
φ · · .
Hence,
1
cos
z
N
φ
· (13.5)
What both of the
z
N equations have in common is that they rely upon unrealistic
idealizations of zero wind and exact constant altitude and speed. In flight test, either
accelerometer methods or INS methods are used to compute the actual flight path axis load
factors.
159
13.6.2 Turn Radius
In a steady, level turn the centripetal acceleration is the horizontal component of normal
acceleration. The vertical component is 1g; just the right amount to maintain exactly
constant altitude for this idealized relationship.
2
t
zh
V
A
R
· (feet/sec
2
) (13.6)
where:
R = turn radius (ft),
t
V = true airspeed (ft/sec), and
zh
A = horizontal component of normal acceleration (ft/sec
2
).
From trigonometry:
( )
2
1
zh z
N N · − (13.7)
and,
0
zh
zh
A
N
g
· (13.8)
Substituting equations 13.7 and equations 13.8 into equations 13.6 and solving for R :
a.
( ) ( )
2 2
2 2
0
1 32.174 1
t t
z z
V V
R
g N N
· ·
⋅ − ⋅ −
For R in feet and
t
V in knots:
( ) ( )
2
2
2 2
1.6878
32.174 1 91.653 1
t
t
z z
V
V
R
N N
 `
. ,
· ·
⋅ − ⋅ −
(13.9)
13.7 Turn Rate
Once the turn radius is determined (equation 13.9), we can compute the turn rate. The
relationship derives from the kinematics of constant speed rotation about a point.
t
V R ω · ⋅ (13.10)
where:
160
R = radius of turn, and
ω = turn rate.
The symbology we previously used for turn rate was σ! ; the rate of change of ground
track angle. Then, solving for turn rate:
a.
t
g
V
R
σ · !
The above equation is valid for units of R in feet,
t
V in feet per second and
g
σ! in
radians per second. For R in feet,
t
V in knots and
g
σ! in degrees per second we get:
( )
1.6878
57.2958 33.947
t
t
g
V
V
R
R
σ
 `
. ,  `
· ⋅ · ⋅
. ,
! (13.11)
13.8 Winds Aloft
Since the advent of the INS in the 1970s, it has been possible to compute accurate values
of air data parameters in dynamic maneuvers such as turns. However, this required the use of
wind calibration runs conducted in wingslevel 1g flight where the air data system errors
were known from conventional tests. In addition, INS data had small drift errors in the
groundspeeds. With the availability of the GPS in the 1990s, an accurate value of
groundspeed was available. The mathematics and illustrating data for one such technique
used in turning flight (that does not require the use of a wind calibration) will be presented.
The INS gives you six parameters of interest for performance and flying qualities. These
are three angles called Euler angles and three velocities in the north ( N ), east ( E ) and down
( D) directions. The Euler angles are the heading from true north designated psi (ψ ), the roll
(or bank) angle designated phi (φ ), and the pitch attitude designated theta (θ ). The
groundspeed components from an INS are
gN
V ,
gE
V , and
gD
V . The problem is that we
assumed we knew the groundspeeds accurately. We didn’t! The typical drift rate of an INS
was on the order of 1 nautical mile per hour. Therefore, we had typical errors of about 1 knot
in the horizontal groundspeeds at any one time. Now (late 1990s) we have a new device
designated as embedded GPS/INS (EGI). This combines the outputs of an INS with the
velocities and position data from the GPS using a filter. The GPS specification accuracies for
the horizontal speeds are 0.1 m/sec (0.19 knot). This small error does not drift with time.
Therefore, we have introduced a new level of accuracy into our data. Now, we will proceed
to develop the equations starting with the basic vector relationship of true airspeed,
groundspeed, and wind.
t g w
V V V · +
" " "
(13.12)
Solving for the magnitude of the true airspeed vector:
161
( ) ( ) ( )
2 2 2
ti t gN wN gE wE gD wD
V V V V V V V V
]
+ ∆ · + + + + +
]
]
(13.13)
We will assume the vertical wind is zero. Taking the square of both sides:
( ) ( ) ( )
2 2 2
2
ti t gN wN gE wE gD
V V V V V V V
]
+ ∆ · + + + +
]
]
(13.14)
From here on in the derivation, we will simply strive to minimize the sum of the
difference between the left and right side of the above equation. Defining a parameter we
shall call F
*
(F – star), we want to minimize the sum of this parameter simultaneously with
respect to each of the three unknowns (
wN
V
,
wE
V
,
t
V ∆ ). The iteration is the method of
Taylor’s series in three dimensions:
( )
2 2 2 2 *
0.5
tx ty tz t
F V V V V · ⋅ + + − (13.15)
The 0.5 factor is just to eliminate ½factors in the final formulation.
tx gN wN
V V V · + (13.16)
ty gE wE
V V V · + (13.17)
tz gD
V V · (13.18)
t ti t
V V V · + ∆ (13.19)
Defining three more parameters: , f g and h :
*
1
N
i tx
i
f F V
·
· ⋅
∑
(13.20)
*
1
N
i ty
i
g F V
·
· ⋅
∑
(13.21)
*
1
N
i t
i
h F V
·
· ⋅
∑
(13.22)
There are N data points and N must be at least three. The , , x y z unknowns are as
follows:
a.
wN
x V · ,
b.
wE
y V · , and
162
c.
t
z V · ∆ .
We will assume zero initial estimates for the unknowns.
a. 0 x y z · · ·
In addition, initialize , , f g h and the partial derivatives to zero as follows:
a. 0 f g h · · · ,
b. / / / 0 f x f y f z ∂ ∂ · ∂ ∂ · ∂ ∂ · ,
c. / / / 0 g x g y g z ∂ ∂ · ∂ ∂ · ∂ ∂ · , and
d. / / / 0 h x h y h z ∂ ∂ · ∂ ∂ · ∂ ∂ · ,
Next we will generate a matrix of partial derivatives of , f g and h . Summing from one
to N :
( )
2
*
1
/
N
tx
i
f x V F
·
]
∂ ∂ · +
]
∑
(13.23)
( ) ( )
1
/ ( ) ( )
N
ty tx
i
f y V i V i
·
]
∂ ∂ · ⋅
]
∑
(13.24)
( ) ( )
1
/ ( ) ( )
N
t tx
i
f z V i V i
·
] ∂ ∂ · − ⋅
]
∑
(13.25)
( ) ( )
1
/ ( ) ( )
N
tx ty
i
g x V i V i
·
]
∂ ∂ · ⋅
]
∑
(13.26)
( )
2
*
1
/ ( )
N
ty
i
g y V i F
·
]
∂ ∂ · +
]
]
∑
(13.27)
( ) ( )
1
/ ( ) ( )
N
t ty
i
g z V i V i
·
]
∂ ∂ · − ⋅
]
∑
(13.28)
( ) ( )
1
/ ( ) ( )
N
tx t
i
h x V i V i
·
] ∂ ∂ · ⋅
]
∑
(13.29)
( ) ( )
1
/ ( ) ( )
N
ty t
i
h y V i V i
·
]
∂ ∂ · ⋅
]
∑
(13.30)
163
( )
2
*
1
/ ( )
N
t
i
h z V i F
·
]
∂ ∂ · − +
]
∑
(13.31)
The following matrix formulation will solve for improved values for the unknowns:
1
1
/ / /
/ / /
/ / /
wN wN
wE wE
t t
j j
V V f x g x h x f
V V f y g y h y g
V V f z g z h z h
−
+
∂ ∂ ∂ ∂ ∂ ∂ ¹ ¹ ¹ ¹ ] ¹ ¹
¹ ¹ ¹ ¹ ¹ ¹
]
· − ∂ ∂ ∂ ∂ ∂ ∂ ⋅
' ' ' ' ' '
]
¹ ¹ ¹ ¹ ¹ ¹
] ∆ ∆ ∂ ∂ ∂ ∂ ∂ ∂ −
¹ ¹ ¹ ¹ ] ¹ ¹
(13.32)
With improved values for the unknowns, simply return to the beginning of the algorithm
and repeat the process until convergence occurs. This will usually occur after just a few steps.
The parameter j is the iteration number. We now have the north and east components of
wind and the previously unknown error in true airspeed.
14.0 DYNAMIC PERFORMANCE
14.1 Introduction
Dynamic performance typically involves the collection of lift and drag data at near
constant Mach number with maneuvers that last less than 15 seconds. This is accomplished
by varying normal load factor (
z
N ) in a short time period. There are three dynamic
performance maneuvers: roller coaster, splits, and windup turn.
14.2 Roller Coaster
The roller coaster is a smooth sinusoidal variation of load factor versus time. The
maneuver begins with a stabilized trimmed point at an aim Mach number, altitude (
C
H ), and
z
N = 1.0. The throttle is kept constant during the maneuver. The maneuver is also called a
pushoverpullup because that is what is done. The maneuver begins with a pushover to a g
level less than 1.0. On fighter aircraft that is usually to an
z
N of 0.0 and on transport aircraft
that is usually to an
z
N of 0.5. Then a pullup is performed back through
z
N of 1.0 to an
z
N
of 1.5 on transport aircraft, or 2.0 or more on fighter aircraft. Some fighter projects used a
maximum
z
N of more than 2.0 and some have used an aim angle of attack (α ) instead of a
maximum load factor as the maximum point in the roller coaster. This maximum α is
usually (but not always) something less than the limit α . This is because a large maximum
α would produce large Mach number losses during the maneuver because the aircraft is at a
high drag condition at a positive flight path angle (γ ) and is decelerating very rapidly. After
attaining maximum
z
N then a pushover is performed back to
z
N = 1.0.
The rate of change of
z
N is between 0.25 and 0.50 g per second. The slower rate would
produce larger Mach number variations but would also produce smaller rate effects on the
data. Both Mach number and rate corrections are made to the data; therefore, the maneuver
will take an average of 8 seconds to perform. Generally, there is a net altitude loss during the
maneuver and a net Mach number loss, but both are quite small. The Mach number loss is
usually no more than 0.01 and the altitude loss is less than 1,000 feet. If
z
N is more than 2.0
during the pullup, then the Mach number loss could be more than 0.01, but corrections are
made to the data to nominal Mach numbers. Nominal Mach numbers would typically be 0.70,
0.80, 0.85, 0.90, etc.
A simulation of a roller coaster maneuver was conducted. The aircraft drag model was
the same as for the takeoff simulation presented in the takeoff section. This was for a pseudo
F16 aircraft. For a lift coefficient less than 0.6 and low Mach numbers where compressibility
is not substantial, Figure 14.1 represents the drag polar used.
Lift Coefficient versus Drag Coefficient
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.018 0.020 0.022 0.024 0.026 0.028 0.030 0.032
Drag Coefficient  CD
L
i
f
t
C
o
e
f
f
i
c
i
e
n
t

C
L
Figure 14.1 Drag Model
The initial condition chosen to illustrate the roller coaster is 0.6 Mach number at 30,000
feet pressure altitude, standard day. The first data point was at
z
N = 1.0 and then thrust was
set equal to the drag at that point and kept constant during the remainder of the maneuver.
The
x
N and
z
N formulas used are those derived in earlier sections for nonbanked flight as
follows:
0
t
x
t
V H
N
g V
· +
! !
(14.1)
0
cos
t
z
V
N
g
γ
γ
⋅
· +
!
(14.2)
A sinusoidal variation of normal load factor was chosen to produce a period of 4 seconds
with amplitude of 1.0 g. The time histories of normal load factor, Mach number, and
pressure altitude are shown in Figures 14.2, 14.3 and 14.4. As shown, there is a relatively
small loss in altitude (80 feet) and gain in Mach number (0.004). However, for a fighter type
aircraft, the range of ,
L
C α is small. On the positive side, due to the slow
z
N variation, the
noise in the data is usually quite low.
Roller Coaster Simulation: Normal Load Factor versus Time
0.0
0.5
1.0
1.5
2.0
0.0 1.0 2.0 3.0 4.0
Elapsed Time (sec)
N
o
r
m
a
l
L
o
a
d
F
a
c
t
o
r
(
g
'
s
)
Figure 14.2 Roller Coaster Normal Load Factor
Roller Coaster Simulation: Altitude versus Time
29,900
29,920
29,940
29,960
29,980
30,000
0 1 2 3 4
Elapsed Time (sec)
A
l
t
i
t
u
d
e
(
f
t
)
Figure 14.3 Roller Coaster Altitude Time History
Roller Coaster Simulation: Mach Number versus Time
0.600
0.601
0.602
0.603
0.604
0.605
0 1 2 3 4
Elapsed Time (sec)
M
a
c
h
N
u
m
b
e
r
Figure 14.4 Roller Coaster Mach Number Time History
14.3 Windup Turn
The windup turn begins at wings level trimmed at an aim Mach number and altitude. The
throttle is kept constant during the maneuver because most inflight thrust computer
programs are ineffective at computing thrust accurately during throttle transients. Then, the
aircraft is gradually pulled into a turn, at a rate of up to 1.0 g per second, until a limit
condition on
z
N or α is reached. This usually takes no more than 8 seconds and is often as
little as 3 seconds. The aircraft is pointed downhill during the maneuver to minimize the
Mach number loss during the highg maneuver as drag gets very high and the aircraft
decelerates rapidly. The aircraft is trading altitude for airspeed. Since the maneuver only
lasted a few seconds, even large deceleration rates would not vary the Mach number more
than about 0.02. There is also an altitude loss during the maneuver of up to 2,000 feet. The
total maneuver, including the recovery, could produce an altitude loss of up to 10,000 feet as
the aircraft ends up pointed nearly straight down at the conclusion of the maneuver. A better
maneuver to perform is a pure inverted pullup, which is a portion of a splits.
14.4 SplitS
The splits is a fighter tactics maneuver used to change direction and altitude very
rapidly. A portion of the maneuver is an inverted pullup during which
z
N is varied from
near 1.0 to the limit g of the aircraft. This is ideal to collect dynamic performance data. The
aircraft is trimmed at an aim Mach number and altitude. The throttle is kept constant during
the maneuver to give an accurate thrust computation. The aircraft is rolled inverted
(180 degrees roll angle) and an inverted pullup is performed at a rate of up to 1.0 g per
second to the limit
z
N or α . This takes approximately 3 to 8 seconds. No attempt is made
to minimize the Mach number variation, but the Mach number usually decreases no more
than 0.02 during the data portion of the maneuver, which is less than 8 seconds. As with the
windup turn, an altitude loss of up to 2,000 feet during the data acquisition portion of the
maneuver is typical, but the total maneuver including recovery could produce an altitude loss
of up to 10,000 feet. We attempt to collect data from pitch attitudes (θ ) of 0 to about
70 degrees to avoid getting data during the INS transition through 90 degrees of θ at which
the heading (ψ ) changes by 180 degrees. This would often dictate the g onset rate since it is
desired to achieve maximum g or α before the aircraft reaches about a negative 70 degrees
pitch angle. This maneuver is better than the windup turn for data processing with an INS
since there are only small bank angle (φ ) variations from 180 degrees and terms in the INS
equations involving φ are negligible. We also did not have any significant roll rate effects.
To illustrate the splits, a simulation is shown. The drag model was modified, from that
used for the roller coaster, with the addition of a separation drag term as follows:
( )
2
0.5 0.6
D L
C C ∆ · ⋅ − (14.3)
0 if 0.6
D L
C C ∆ · <
The
x
N formula is identical to the one used for the roller coaster; however, the
z
N
formula is the negative of the roller coaster formula. This can be seen from the axis
transformations in the excess thrust section. The transformation for
z
N involves sinφ and
cosφ terms. For the pure inverted case ( 180 φ · degrees):
a. sin 0 φ · , and
b. cos 1 φ · − .
Then,
0
cos
t
z
V
N
g
γ
γ
] ⋅
· − +
]
]
!
(14.4)
Figure 14.5 plots the drag model used. The simulation was performed at a rate of 1.0 g
per second. The simulation was ceased at a lift coefficient of 1.60. The initial conditions
chosen were 30,000 feet and a Mach number of 0.85.
Splits and Pullup Drag Model: CL versus CD
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Drag Coefficient  CD
L
i
f
t
C
o
e
f
f
i
c
i
e
n
t

C
L
Figure 14.5 SplitS Drag Model
The timehistory parameters of normal load factor, Mach number, and pressure altitude
follow in Figures 14.6 through 14.8.
SplitS Simulation: Nz versus Time
1
2
3
4
5
6
7
0 1 2 3
Elapsed Time (sec)
N
o
r
m
a
l
L
o
a
d
F
a
c
t
o
r
(
N
z
)
Figure 14.6 SplitS Normal Load Factor
Splits Simulation: Mach Number versus Time
0.80
0.81
0.82
0.83
0.84
0.85
0.86
0 1 2 3
Elapsed Time (sec)
M
a
c
h
N
u
m
b
e
r
Figure 14.7 SplitS Mach Number Time History
SplitS Simulation: Pressure Altitude versus Time
29,400
29,600
29,800
30,000
0 1 2 3
Elapsed Time (sec)
P
r
e
s
s
u
r
e
A
l
t
i
t
u
d
e
(
f
t
)
Figure 14.8 SplitS Altitude Time History
14.5 Pullup
On the F15 projects, a pullup maneuver has been used in lieu of the splits to obtain
highα data. They have found that the pullup maneuver has one big advantage over the
splits. That is, there is no need to recover back to the original altitude. A simulation for the
pullup was conducted using the same drag model and initial conditions as for the splits. The
pullup simulation was conducted at the same g onset rate of 1.0 g per second. In addition, the
end condition of 1.60
L
C · was the same. The Mach number and pressure altitude time
histories are in Figures 14.9 and 14.10.
Pullup Simulation: Mach Vs Time
0.78
0.79
0.80
0.81
0.82
0.83
0.84
0.85
0 1 2 3
Elapsed Time (sec)
M
a
c
h
N
u
m
b
e
r
Figure 14.9 Pullup Mach Number Time History
Pullup Simulation: Altitude Vs Time
30,000
30,100
30,200
30,300
0 1 2 3
Elapsed Time (sec)
A
l
t
i
t
u
d
e
(
f
t
)
Figure 14.10 Pullup Altitude Time History
Table 14.1 compares the initial conditions and end conditions of the pullup and the
splits.
Table 14.1
PULLUP AND SPLITS INITIAL AND END CONDITIONS
z
N
M
t
V
(kts)
C
H
(ft)
H
!
(ft/sec)
t
V
!
(kts/sec)
Initial 1.00 0.850 500.9 30,000 0.0 0.0
Pullup 6.450 0.785 462.3 30,219 +226.0 58.7
SplitS 6.936 0.800 472.8 29,452 428.2 45.1
As can be seen, the splits has the advantage of not losing as much Mach number.
However, the pullup does not end up with a very large vertical velocity.
14.6 Angle of Attack
During the roller coaster, pullup, and splits maneuvers the computation of angle of
attack from the INS is quite simple for bank angles near 0 or 180 degrees. In practice, the full
transformation equations are used.
( 0) α θ γ φ · − · roller coaster and pullup (14.5)
( 180) α θ γ φ · − + · splits (14.6)
The roller coaster maneuver, particularly, could be used to calibrate production angleof
attack probes or vanes. Only for very high angle of attack would you want to use the splits
for calibration of production systems. The above equations are simplified for illustration
purposes only. The full equations involved bending and rate corrections and allowance for
being off exactly φ = 0 or 180 degrees. As discussed in the flight path acceleration section,
the one shortfall of the INS method is that vertical wind is assumed zero. You can detect
vertical wind by comparing data on the lift curve.
a. ( ) ,
L
f C M α ·
In addition, one can use an INS method to calibrate angle of attack during turns. The turn,
especially a highg (high bank angle) turn, will be less sensitive to vertical wind since the
vertical component of velocities in the angleofattack formula is proportional to the cosine
of the bank angle.
14.7 Vertical Wind
If there is an unexplained bias in your data, then it could be that there is a vertical wind.
One way to minimize the effect of vertical wind is to do a varying g maneuver during a
stabilized highg turn, keeping the bank angle (φ ) near 90 degrees. Since you are not trying
to get drag data, the throttle could be varied to maintain speed. The vertical wind would not
affect the turn data as much, since the vertical wind is nearly perpendicular to the axis of the
angle of attack.
15.0 SPECIAL PERFORMANCE TOPICS
15.1 Effect of Gravity on Performance
Below is the international gravity formula as adopted by the International Union of
Geodesy and Geophysics as presented in Britannica¹ Online.
2 4
0
978.03185 1 0.005278892 sin 0.000023462 sin γ ϕ ϕ ] · ⋅ + ⋅ + ⋅
]
cm/sec
2
(15.1)
Where the symbology used by the International Union is as follows:
a.
0
γ = sea level gravity (cm/sec
2
), and
b. ϕ = latitude (degrees).
In this text, we have used a rather simplified gravity model of g = constant = 32.174
ft/sec
2
. As of the writing of this text, that simplification is widely used in the conventional
aircraft flight testing community. This topic will address the magnitude of error that this
simplification produces. As will be seen, the error is quite small (<1 percent), but not zero.
First, we will take the liberty of changing the International Union’s sea level gravity
symbology from
0
γ to
0
g .
Consider only a 1g flight where the aircraft is unbanked and has zero vertical velocity
and zero rate of change of vertical velocity. Under these conditions, the normal load factor
(
z
N ) would not be precisely 1.00. There are four variables: latitude, altitude, speed, and
heading. We will consider them individually.
The internationally agreed upon exact conversion factor between meters (or metres in
Great Britain) is 0.3048 (divide meters by 0.3048 to yield feet) and the number of centimeters
(cm) in a meter is 100. Given that and using equation 15.1, some typical values of sea level
gravity are shown in Table 15.1.
Table 15.1
EFFECT OF LATITUDE ON GRAVITY AT SEA LEVEL
Place
Latitude
(deg)
g
9.80665
(m/sec
2
)
g
32.17405
(ft/sec
2
)
Variation
from the
Standard
(pct)
Reference North Pole 90.00 9.8322 32.2578 0.26
Northern Greenland 80.00 9.8306 32.2526 0.24
Pt. Barrow, Alaska 71.00 9.8267 32.2397 0.20
Arctic Circle 66.50 9.8239 32.2306 0.18
Anchorage, Alaska 62.00 9.8207 32.2202 0.14
St. Petersburg, Russia 60.00 9.8192 32.2151 0.13
Copenhagen 55.50 9.8155 32.2031 0.09
London, England 51.30 9.8118 32.1911 0.05
Lake of the Woods, Minn. 49.33 9.8101 32.1854 0.04
45 deg latitude 45.00 9.8062 32.1725 0.00
Bldg. 2750, AFFTC 34.92 9.7973 32.1432 0.10
Baghdad 33.00 9.7957 32.1380 0.11
Florida Keys, Florida 24.58 9.7893 32.1170 0.18
Mexico City 20.00 9.7864 32.1075 0.21
Costa Rica 10.00 9.7819 32.0928 0.25
Equador (Equator) 0.00 9.7803 32.0877 0.27
Note: The local gravity at Edwards of 32.136 ft/sec
2
has been measured and agrees with the
model.
The above local g values are computed for sea level. Edwards is at 2,300 feet geometric
altitude and the gravity at that altitude is 32.136 ft/sec
2
. The gravity varies with altitude.
Using latitude of 35 degrees, Table 15.2 illustrates this effect using the inverse square gravity
law. The places in Table 15.1 were chosen to represent either even latitudes or interesting
places. For instance, Point Barrow, Alaska, and Florida Keys, Florida, represent the extreme
latitudes of the continental United States. Lake of the Woods, Minnesota, is the highest
latitude in the lower 48 states.
The earth’s radius (20,925,643 feet) is also from the International Union of Geodesy and
Geophysics and is a value for the equator. This compares to 20,855,553 feet from the 1976
U.S. Standard Atmosphere.
Table 15.2
EFFECT OF ALTITUDE ON GRAVITY
Altitude
(ft)
g
(ft/sec
2
)
Percent from
Surface
Percent from
Standard
0 32.143 0.02 0.10
2,300 32.136 0.00 0.12
10,000 32.113 0.07 0.19
20,000 32.082 0.17 0.29
30,000 32.051 0.26 0.38
40,000 32.021 0.36 0.48
50,000 31.990 0.45 0.57
60,000 31.960 0.55 0.67
70,000 31.929 0.64 0.76
80,000 31.899 0.74 0.86
90,000 31.869 0.83 0.95
100,000 31.838 0.93 1.04
The last two variables are speed and heading which need to be considered together.
Speed has an effect upon normal load factor due to coriolis terms in the gravity equations that
are functions of the true heading. Using 40,000 feet and latitude of 35 degrees,
Table 15.3 illustrates the speed and heading effect.
Table 15.3
EFFECT OF HEADING AND SPEED ON NORMAL LOAD FACTOR
Heading
(deg)
Mach
Number
Normal Load Factor
(g)
0 0.0 0.9952
0 0.8 0.9943
0 2.0 0.9896
90 0.0 0.9952
90 0.8 0.9914
90 2.0 0.9824
180 0.0 0.9952
180 0.8 0.9943
180 2.0 0.9896
270 0.0 0.9952
270 0.8 0.9972
270 2.0 0.9968
So, what is the significance of this? The normal load factor experienced by an aircraft
varies with latitude over the earth, how high and how fast the aircraft is flying and in what
direction. For a given mass of aircraft, we needed to generate 0.23 percent more lift over St.
Petersburg, Russia, than over Edwards AFB. We needed 0.36 percent less lift at 40,000 feet
than at 2,300 feet over Edwards AFB. At 0.8 Mach number, 40,000 feet, 0.59 percent more
lift is required heading west than heading east. Generally, for conventional aircraft
performance, we have been ignoring these factors.
How did these variations in
z
N translate to performance? As
z
N increased, it was
necessary to generate more lift and therefore, more drag due to lift was created. In cruise
performance, a 1percent increase in drag is about a 1percent increase in fuel flow required
to sustain stabilized flight. Using a B52G drag polar at 0.8 Mach number, corresponding to
an optimum cruise at 40,000 feet, Table 15.4 was generated.
Table 15.4
EFFECT OF HEADING ON DRAG COEFFICIENT
Heading
z
N
D
C
Percent from
Reference
Reference 1.0000 0.02641 0.00
270 (west) 0.9972 0.02634 0.26
0 or 180 0.9943 0.02628 0.49
90 (east) 0.9914 0.02622 0.72
Very similar percentage differences were obtained using an F15 drag polar. At Mach
number 2.0 for the F15 aircraft, the variations in drag are less than 0.1 percentage. This is
due to the much smaller amounts of drag due to lift at the higher speeds. Although
z
N varied
more at M=2.0 than at M=0.8, the effect on performance was actually much less.
The significant comparison is between west and east being nearly ½of 1 percent apart.
The bias between the reference and the other data tended to fall out in flight test data as the
drag polars generated are biased to compensate for this effect and there is not a ½percent
error in range data. Nevertheless, the data collected heading west would have shown about ½
of 1 percent more drag and fuel flow than the data collected heading east, if the data were
accurate enough to detect that small difference.
What we are talking about is roughly up to a ½of 1percent factor we had been ignoring.
This does not produced a bias in our data (unless all our cruise data is collected heading east)
but is rather a source of the scatter. With an INS as a data source, we can account for the
variation in gravity.
15.2 Performance Degradation during Aerial Refueling
A common misconception is that the drag of the receiver aircraft during aerial refueling is
increased. The drag of the receiver aircraft is unchanged. The thrust required of the receiver
is increased due to the receiver climbing in the tanker downwash. The tanker downwash
creates a negative vertical wind that the receiver aircraft encounters. Relative to the wind
axis, the receiver is climbing at a flight path angle exactly equal to the tanker downwash
angle to maintain a constant altitude. To sustain this climb, the receiver aircraft requires
additional thrust and a resultant increase in fuel flow.
During tests of the KC10 aircraft with 10 different types of receiver aircraft, the average
increase in fuel flow for the receiver aircraft was 25 percent. The B1B behind a KC135
aircraft showed a 15percent increase. The YC141B increase in fuel flow behind a KC135
was 20 percent.
To estimate the increase in thrust required for a receiver aircraft, you only need to know
the theoretical downwash angle behind the tanker and then apply a downwash factor. The
downwash factor (K) is simply a multiplicative factor to account for the fact that the receiver
aircraft is in a flow field that is a combination of the tanker flow field and the free stream. For
both the KE3A and the B1B aircraft, this K factor is about 0.5. The theoretical downwash
angle (
0
ε ) is exactly twice the ideal angle of attack.
( )
( )
0
2
Lt
t
C
AR
ε
π
⋅
·
⋅
(15.2)
where:
Lt
C = lift coefficient of the tanker aircraft, and
t
AR = aspect ratio of the tanker aircraft.
The actual downwash angle is found (with a K of 0.5) to be approximately equal to the
ideal angle of attack of the tanker.
( )
Lt
t
t
C
AR
ε
π
·
⋅
(15.3)
Then the increase in thrust of the receiver could be computed by the component of weight
through the downwash angle. With respect to the wind axis, the receiver aircraft is climbing
while behind a tanker in level flight.
sin( )
n t
F W ε ∆ · ⋅ (15.4)
15.3 Performance Degradation during Terrain Following
Flight while performing terrain following results in an increase in average fuel flow when
compared to flight at the same average Mach number and altitude level. While in the terrain
following mode, the aircraft is constantly either pulling up or pushing over. In a pullup
(
z
N >1) the drag is increased over that for an
z
N =1 due to an increase in drag due to lift (or
induced drag). In a pushover, (
z
N <1) the drag is reduced due to a decrease in the drag due to
lift. Because of the parabolic nature of the drag polar, the magnitude of the drag increase in
the pullup is greater than the magnitude of the drag decrease in the pushover. The net effect is
there is a net increase in average thrust required and a resultant increase in average fuel flow.
For the case of an aircraft with automatic terrain following and afterburner, the average
increase in fuel flow can be substantial. Every time afterburner is used, the fuel flow
increases dramatically. The thrust specific fuel consumption ( ) tsfc will typically be less than
1.0 in nonafterburner and >2.0 in afterburner.
15.4 Uncertainty in Performance Measurements
There is no precise answer to the question, “how accurately do we measure certain
performance flight test parameters,” as each instrumentation system is different.
Nevertheless, our experience has given us some approximate uncertainties that we feel are
obtainable and had been achieved. Some typical parameter uncertainties are shown in Table
15.5. In some cases, these parameters are not direct instrumentation measurements, but rather
the result of computations involving several measurements.
Table 15.5
PARAMETER UNCERTAINTIES
Parameter Units Symbol Uncertainty
Fuel Flow lbs/hr
f
W t1%
Calibrated Airspeed kts
C
V t0.5 knots
Gross Weight lbs
t
W t0.5%
Longitudinal Load Factor g
x
N t0.001 g
Normal Load Factor g
z
N t0.01 g
Ambient Temperature °K
T
t0.5 °K
Pressure Altitude ft
C
H t25 feet
15.5 Sample Uncertainty Analysis
For a transport category aircraft, a performance figure of merit might be the specific
range at optimum speed and altitude. Let us choose a typical high altitude cruise condition:
a.
C
V
= 280 knots (calibrated airspeed), and
b.
C
H
= 35,000 feet (pressure altitude).
On a standard day the ambient temperature is:
c. T = 218.81 °K.
Calculating the Mach number:
d. M = 0.8213.
True airspeed is:
e.
t
V = 473.44 knots.
If the computed ambient temperature is in error on the high side by 0.5 degree K then the
true airspeed would be
t
V
= 473.98 knots for a 0.11percent error. In addition, an altitude
error of 25 feet produces a 0.04percent error, and a calibrated airspeed error of 0.5 knot
produces a 0.26percent error.
At an / 10.0 L D · , an error of 0.001 g in longitudinal load factor yields a 1.0percent
error in drag. We shall assume error in drag produces a 1.0percent error in range factor.
Then, for range factor ( RF ), we have the following errors:
a.
t
V 0.11 percent due to T error,
b.
t
V 0.04 percent due to
C
H error,
c.
t
V 0.26 percent due to
C
V error,
d.
x
N 1.00 percent,
e.
t
W 0.50 percent, and
f.
f
W 1.00 percent.
The root mean square (rms) of the three
t
V uncertainties computes to be 0.285 percent.
The RMS of the four uncertainties computes to be 1.53 percent. Please note that carrying out
the speeds to five significant figures did not imply that we could measure speeds to that level
of accuracy. At the time of this handbook, with the advent of EGI even greater accuracies
than those presented above may be achieved for airspeeds, altitudes, and flight path
accelerations.
15.6 Wind Direction Definition
What may seem to be an improper definition of wind direction (from which the wind is
blowing) may derive from ancient Greece. Improper in the sense that defining the wind
direction as from which it is blowing is opposite from the vector direction of wind. In
Britannica¹ Online, a structure called the Tower of the Winds is discussed briefly. In about
100 BC an octagonal (eightsided) marble structure, 42 feet high and 26 feet in diameter, was
constructed. The eight sides face points of the compass (N, NE, E, etc). It would seem
logical that a wind blowing on the structure would be considered a positive wind. The wind
would always be positive, since it would be blowing on some side of the structure – never
away from the structure, so to speak. Therefore, if the wind were blowing directly on the
north side of the Tower of the Winds, this positive wind would have a direction of north
(0 degrees). This direction is the direction from which the wind is blowing, the same as the
compass heading of the Tower. One could think of this Tower as either an aircraft control
tower or an aircraft.
180
16.0 STANDARDIZATION
16.1 Introduction
For presentation and comparison purposes, performance data are usually corrected to
standard conditions. The standard conditions are specified values of gross weight, pressure
altitude, cg (center of gravity), and Mach number. Standard ambient temperature is usually
based on the 1976 U.S. Standard Atmosphere. Standardization relies upon a predicted model
of drag, thrust, and fuel flow. Usually, small corrections to standard day conditions are made,
but these could be large when temperature is substantially off standard day. If there is a 10
percent error in the predicted model and we made 10percent corrections to the data, we
incurred only a 1 percent error in the standardized results. At the AFFTC in midsummer, the
temperature at 30,000 feet is, on average, 10 degrees C hotter than standard day, which
produces, typically, about a 10percent decrease in thrust at MIL or MAX. The
standardization is performed using an additive increment method.
16.2 Increment Method
The general principle of standardization is an additive increment method. The formulas
used to standardize net thrust (
n
F ), fuel flow (
f
W ), and drag ( D) are as follows:
( )
ns nt ns nt
F F F F ′ ′ · + − (16.1)
where:
ns
F = standardized net thrust (pounds),
nt
F = test day net thrust (pounds),
ns
F′ = standard day predicted net thrust (pounds), and
nt
F′ = test day predicted net thrust (pounds).
( )
fs ft fs ft
W W W W ′ ′ · + − (16.2)
where:
fs
W = standardized fuel flow (pounds/hour),
ft
W = test day fuel flow (pounds/hour),
fs
W′ = standard day predicted fuel flow (pounds/hour), and
ft
W′ = test day predicted fuel flow (pounds/hour).
Fuel flow is first standardized to a minimum fuel lower heating value (LHV), usually
18,400 Btu/pound.
18, 400
test
ft ft
LHV
W W
 `
· ⋅
. ,
(16.3)
181
Typical test values of LHV are in the vicinity of 18,550 Btu/pound, which amounts to a
½percent correction. The correction will generally increase fuel flow, since the spec is a
minimum. That is, almost all actual fuel will have an LHV greater than the spec.
( )
s t s t
D D D D
′ ′
· + − (16.4)
where:
s
D = standardized drag (pounds),
t
D = test day drag (pounds),
s
D
′
= predicted standard day drag (pounds), and
t
D
′
= predicted test day drag (pounds).
t nt ext
D F F · − (16.5)
t
ex x t
F N W · ⋅ = test day measured excess thrust (16.6)
Then,
( ) ( )
s t
ex ex ns s nt t
F F F D F D
′ ′ ′ ′
· + − − − (16.7)
The above equations illustrate the general principle. The test net thrust is determined,
usually, from an inflight thrust deck. The predicted thrust and fuel flows are determined
from a prediction (or status) deck. These are described briefly in the thrust section. The
predicted drags are obtained from a contractorprovided predicted drag model subroutine.
The contractor drag model should include an accounting for skin friction drag. In lieu of that,
formulas presented in the lift and drag section could be used.
Each maneuver involves a different parameter being adjusted to standard conditions but
the basic method is the same incremental difference method. The standardization parameters
for various maneuvers are discussed in the following text.
16.2.1 Climb/Descent
Excess thrust and fuel flow are standardized:
a.
z
N is computed.
16.2.2 Acceleration/Deceleration
Excess thrust and fuel flow are standardized:
a.
z
N = 1.0.
182
16.2.3 Accelerating/Decelerating Turn
Excess thrust and fuel flow are standardized:
a.
z
N is specified.
16.2.4 Cruise
Fuel flow is standardized:
a.
z
N = 1.0 (usually) (Note: a rare exception to the 1.0g would be for standardizing
data in an endurance turn.), and
b. Excess thrust = 0.0.
16.2.5 ThrustLimited Turn
z
N and fuel flow are standardized:
a. Excess thrust = 0.0.
16.3 Ratio Method
An alternative to the increment method of standardization is a method based upon ratios.
The formulas for standard day net thrust, fuel flow, and drag would be as follows:
ns
ns nt
nt
F
F F
F
] ′
] · ⋅
′
]
]
(16.8)
fs
fs ft
ft
W
W W
W
]
′
]
· ⋅
′ ]
]
(16.9)
s
s t
t
D
D D
D
] ′
] · ⋅
′
]
]
(16.10)
Then, standard day excess thrust (
s
ex
F ) would be:
s
ex ns s
F F D · − (16.11)
For fixed throttle maneuvers (climb, turn, and accel), the above equation would suffice.
For cruise, where standard excess thrust should be zero, an iteration is required.
The question that needs to be answered is “what is the difference in the magnitude of
difference between the ratio and difference methods?” Take the case of the standardized
183
excess thrust in acceleration. If there was zero error in both test day measured net thrust and
in the thrust model, then there would be zero error in the standardization for both ratio and
increment methods. From the above equations, let us write out the full
s
ex
F formula for both
increment and ratio methods.
s
ns s
ex nt t
nt t
F D
F F D
F D
] ] ′ ′
] ] · ⋅ − ⋅
′ ′
] ]
] ]
ratio method (16.12)
However,
t
t nt ex
D F F · − for both methods (16.13)
Then, the ratio method becomes:
s t
s ns s
ex ex nt nt
t nt t
D F D
F F F F
D F D
] ] ] ′ ′ ′
] ] ] · ⋅ + ⋅ − ⋅
′ ′ ′
] ] ]
] ] ]
ratio method (16.14)
( ) ( )
s t
ex ex ns nt s t
F F F F D D
′ ′ ′ ′
· + − − − increment method (16.15)
Then, whichever method introduces the most error into the standardized excess thrust is a
function of the errors in the prediction models. If the prediction models are in error by
approximately a constant percentage, then the ratio method will introduce the least error. This
is because the errors would cancel out when doing the division. Conversely, if the models are
in error by approximately a constant magnitude, then the increment method will introduce the
least error. This is due to the errors canceling out when doing the subtraction.
Either way, one is invariably introducing some errors (hopefully small) into your data by
the very process of standardization. Standardization is performed as a means of convenient
data presentation. One should recognize that a data point on a plot presented as standard
conditions is a data point that was not flown. It represents an extrapolation of an actual test
point. The following are two sources of error in standardization.
a. For cruise at high altitude, the standard day conditions may be unachievable. That is
due to having sufficient thrust on a test day, but not on a standard day. The test day
temperature may have been substantially colder than standard day giving the engine much
more thrust than would be available on the warmer standard day. Your cruise standardization
algorithm should check to assure that standard day drag is less than the maximum available
thrust.
b. The engine may be in some manner limited (turbine temperature or rpm limit) on the
test day. If this limiter is not accurately modeled in the status deck, then the correction to
standard day will have errors. For instance, the engine may not be on this limit on the
standard day, yielding additional thrust. Conversely, it may not be on the limit on the test
day, but would be on the standard day.
184
17.0 A SAMPLE PERFORMANCE MODEL
17.1 Introduction
In this section, we will construct a performance model. The model will be highly
idealized. The purpose of this section is to illustrate some general concepts. One should not
assume that their drag, thrust, or fuel flow models would be the same as, or as simple as,
those presented here.
17.2 Drag Model
17.2.1 Minimum Drag Coefficient
In order to illustrate the shape of performance parameters, such as specific excess power
as a function of Mach number or altitude, we will construct a drag model. That drag model is
fiction, but approximates that of an F16 aircraft. Drag has three components. These are skin
friction, profile drag, and drag due to lift. We could think of drag as having only two
components: minimum drag and drag due to lift. Minimum drag is then the sum of profile
drag and skin friction drag. Drag due to lift is also called induced drag. Profile drag is
sometimes called form drag. For the purposes of our model, we will make up numbers for
standard day at 30,000 feet pressure altitude. Then, our predicted skin friction drag formulas
will be used to compute minimum drag at conditions other than standard day at 30,000 feet.
Our basic formula for drag coefficient is the AFFTC drag model formulation from the
previous section. We will start by assuming that
min D
C = 0.0200 (200 drag counts) for Mach
number < 0.80. That is a typical minimum drag coefficient for a wide range of aircraft. From
the subsonic condition to Mach number = 1.0, the drag coefficient approximately doubles.
Some data points were assumed and a curve fit was applied. Figure 17.1 is delta drag
coefficient for the subsonic condition. The equation for minimum drag coefficient at any
given Mach number is as follows:
min
0.0200
D D
C C · + ∆ (17.1)
185
delta Cd versus Mach Number  Subsonic
y = 2.9003x
3
 7.1998x
2
+ 5.9828x  1.6633
0.0050
0.0000
0.0050
0.0100
0.0150
0.0200
0.0250
0.70 0.80 0.90 1.00 1.10
Mach Number
d
e
l
t
a
C
d

d
r
a
g
r
i
s
e
Figure 17.1 Subsonic Drag Increment
The drag coefficient in the transonic regime will peak out somewhere just past Mach
number = 1.0 and then will sometimes decrease slightly with increasing Mach number. Each
aircraft will have different characteristics, of course. Data values for minimum drag were
assumed at various Mach numbers and curve fits were applied. Figures 17.2 and 17.3 are for
transonic and supersonic speeds.
Delta Cdmin  Transonic
y = 25.5066x
4
+ 113.4193x
3
 188.9433x
2
+ 139.7543x  38.7038
0.005
0.007
0.009
0.011
0.013
0.015
0.017
0.019
0.021
0.023
0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25
Mach Number
d
e
l
t
a
C
d
Figure 17.2 Transonic Drag Increment
186
delta Cd  Supersonic
y = 0.011534x
3
+ 0.061267x
2
 0.109113x + 0.083435
0.0175
0.018
0.0185
0.019
0.0195
0.02
0.0205
0.021
0.0215
0.022
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
Mach Number
d
e
l
t
a
C
d
Figure 17.3 Supersonic Drag Increment
Notice that there were overlapping data points in each of the plots. For instance, 0.95 and
1.0 Mach number appeared in both the subsonic and transonic plots.
Summarizing the following curve fit formulas (where X = Mach number and Y = delta
D
C ):
a. Subsonic
1. Y = 2.9003⋅X
3
 7.1998⋅X
2
+ 5.9828⋅X 1.6633
b. Transonic
2. Y = 25.5066⋅X
4
+ 113.4193⋅X
3
188.9433⋅X
2
+ 139.7543⋅X 38.7038
c. Supersonic
3. Y = 0.01153⋅X
3
+ 0.06127⋅X
2
0.10911⋅X +0.08343
Table 17.1 contains the data points, the corresponding curve fits values, and the errors in
the curve fits.
187
Table 17.1
TABULATED DRAG RISE DATA
Mach Number
D
C ∆ Data
D
C ∆ Fit Error = Data – Fit
0.7993 0.00000
0.8000 0.0000 0.00002 0.00002
0.8750 0.0020 0.0023 0.00028
0.9000 0.0040 0.0037 0.00030
0.9500 0.0090 0.0092 0.00019
0.9995 0.01984
1.0000 0.0200 0.0199 0.00010
1.0500 0.0215 0.0218 0.00031
1.0750 0.0216 0.0216 0.00004
1.1000 0.0216 0.0214 0.00019
1.1467 0.0214 0.02148
1.1500 0.0213 0.02144 0.00021
1.2000 0.0210 0.0208 0.00021
1.4000 0.0190 0.0191 0.00011
1.6000 0.0185 0.0184 0.00005
2.0000 0.0180 0.0180 0.00000
Notes: 1. Bold numbers are at Mach numbers where the curve fits equate.
2. The error numbers are carried to one extra digit.
The model for minimum drag is then the three equations (1, 2, and 3 on page 187) with
transition points at the following Mach numbers:
a. 0 for 0.7993
D
C M ∆ · < ,
b. subsonic for 0.7993 0.9995
D
C M ∆ · < < ,
c. transonic for 0.9995 1.1467
D
C M ∆ · ≤ ≤ ,
d. supersonic for 1.1467 2.000
D
C M ∆ · < ≤ , and
e. 0.0180 for 2.0
D
C M ∆ · > .
The Mach number ranges for the above are not meant to imply any general definition of the
terms subsonic, transonic, or supersonic. They are simply where the curve fits for this particular
arbitrary data set intersected.
The first and last conditions are constraints applied to the model. The lowend constraint
( 0.7993 M < ) is to keep the minimum drag at 0.0200 for all Mach numbers less than 0.7993.
The highend constraint ( 2.0 M > ) is to keep the polynomial from giving very unreasonable
results in event the model is used beyond the last Mach number. If this were actual flight test
data, we could not be certain what the behavior of the minimum drag might be beyond where
actual test data were acquired. However, wind tunnel data could perhaps be utilized to
extrapolate beyond where flight test data were obtained. Figure 17.4 puts all three pieces of
the minimum drag model together on a single plot.
188
delta Cd versus Mach Number
0.000
0.005
0.010
0.015
0.020
0.025
0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0
Mach Number
d
e
l
t
a
C
d
Data Points
Subsonic Fit
Transonic Fit
Supersonic Fit
Figure 17.4 Summary of Delta Drag Coefficient
17.3 Skin Friction Drag Coefficient
Skin friction drag coefficient varies with Reynolds number and Mach number. We will use the
empirical skin friction flat plate turbulent boundary layer equations presented in the lift and drag
section, and presume a characteristic length of 10 feet. Figure 17.5 is for standard day conditions.
Skin Friction Drag Coefficient versus Mach Number
0.0010
0.0015
0.0020
0.0025
0.0030
0.0035
0.0040
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Mach Number
S
k
i
n
F
r
i
c
t
i
o
n
D
r
a
g
C
o
e
f
f
i
c
i
e
n
t
Sea Level
10,000 ft
20,000 ft
30,000 ft
40,000 ft
50,000 ft
Figure 17.5 Skin Friction Drag Coefficient
189
At 30,000 feet and 0.8 Mach number, on a standard day, the slope of the
f
C curve is
0.000014 per 1,000 feet. This is positive with increasing altitude; that is, the higher altitude
has the higher skin friction drag. Again, at the same condition, the slope of the
f
C curve
versus temperature is 0.0000018 per 1 degree K. The temperature slope is positive with
increasing temperature; that is, the
f
C
is higher on a day that is hotter than standard. Those
f
C ∆ might appear small until one considers that the typical ratio of wetted area to wing area
is about 4 and the altitude range of a fighter aircraft is 50,000 feet. Therefore, at 0.8 Mach
number, for instance, the total variation in drag coefficient due to skin friction (at the same
lift coefficient) can be calculated as follows:
4 0.000014 50 0.0028
f
wet
D
C
S
C h
S h
∆
∆ · ⋅ ⋅ ∆ · ⋅ ⋅ ·
∆
(28 drag counts) (17.2)
That is a 28drag count number over the range of sea level to 50,000 feet. Compare that
to the typical number of 200 for the minimum drag coefficient. Quite significant!
For our fictional aircraft (modeled after an F16 aircraft), we will presume the following
dimensional data:
a. S = 300 ft
2
 wing area,
b. l = 10 feet  MAC (characteristic length),
c. b = 35 feet  wing span,
d.
2
/ AR b S · = 4.083,
e.
wet
S = 4.0 S ⋅ = 1,200 ft
2
,
f.
Zf
W = 18,000 pounds  zero fuel weight, and
g. Fuel = 6,000 pounds  fuel capacity.
These numbers will be used to illustrate performance parameters in other sections of this
handbook.
17.4 Drag Due to Lift
A drag due to lift (induced drag) model will be derived based upon the formulas
presented in the lift and drag section of this handbook. This model (as with the minimum
drag and skin friction drag) is developed only as a rough approximation of an actual airplane.
Figure 17.6 presents idealized drag due to lift slope data points and a secondorder
polynomial curve fit of those points. With actual flight test data, one will be able to develop a
much more detailed and accurate model. As you can see, we have mostly ignored the
variation in the transonic Mach number range.
190
Theoretical Drag Due to Lift Slope
y = 0.0182x
2
+ 0.0294x + 0.0990
0.05
0.07
0.09
0.11
0.13
0.15
0.17
0.19
0.21
0.23
0.25
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Mach Number
C
d
l
/
C
l
^
2
Figure 17.6 Drag Due to Lift Slope
The above drag due to lift model is for the linear (or pure parabola) portion of the drag
polar. The curve is a parabolic fit of the data and ignores the variations in the transonic speed
range. In general, there will be a deviation from the linear model as flow separation develops.
We will call this the nonlinear portion of the model. As shown in the lift and drag section, a
general formula for drag coefficient that seems to match most flight test data quite well for a
given Mach number, pressure altitude, and longitudinal center of gravity position condition is
as follows:
( ) ( )
2 2
min min
1 2
D D L L L Lb
C C K C C K C C · + ⋅ − + ⋅ − (17.3)
where:
2 0
L Lb
K if C C · < .
The y parameter in the theoretical drag due to lift plot is equal to 1 K . In most textbooks,
the
min L
C
is ignored. The
min L
C (lift coefficient at minimum drag coefficient) is usually some
small positive value due to positive camber on most wings and positive wing incidence. In
our model, we will assume the following for a
min L
C
.
min
0.100 0.05
L
C M · − ⋅ (17.4)
191
Hence, for this model the
min L
C is 0.10 at M = 0.0, 0.05 at M = 1.00, and 0.00 at M =
2.00. We need to emphasize that this model is pure fiction, but the trends do roughly
approximate that of a real aircraft such as the F16.
For the break lift coefficient
Lb
C
,
we will assume a constant value of 0.60. To get a rough
number for 2 K , consider that the drag coefficient will double over that predicted by the
linear model by the time a
L
C of 1.50 is attained. Both 2 K and
Lb
C
will, in general, be
functions of Mach number, but for simplicity, we will give them constant values. From our
models at M = 0.0 and
L
C <0.60.
2
0.0200 0.099 ( 0.10)
D L
C C · + ⋅ − (17.5)
At
L
C = 1.50;
D
C = 0.2140.
Solving for 2 K from equation 17.5:
a.
( )
( )
2
min min
2
1 (
2
D D L L
L Lb
C C K C C
K
C C
]
− + ⋅ −
]
·
−
, and
b.
[ ]
( )
2
2 0.2140 0.2140
2 0.2642
1.5 0.6
K
⋅ −
· ·
−
.
Figure 17.7 is for this model at M = 0.80.
Drag Coefficient versus Lift Coefficient (Mach Number = 0.80)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Lift coefficient  Cl
D
r
a
g
c
o
e
f
f
i
c
i
e
n
t

C
d
Linear Model
NonLinear Model
Figure 17.7 Drag Model at 0.8 Mach Number
192
Figure 17.7 illustrates how dramatically the drag polar can deviate from the pure
parabola. The vast majority of 1g flight occurs at lift coefficients below the point where
significant flow separation begins. To illustrate the general shape of the polar for
L Lb
C C <
,
we will plot drag coefficient versus lift coefficient as a function of Mach number. Figure
17.8 represents only the subsonic Mach numbers, and Figure 17.9 includes all Mach
numbers. Note to those who are accustomed to seeing drag coefficient on the xaxis: the plot
axes are opposite of the usual convention.
Drag Coefficient versus Lift Coefficient {f(Mach Number)}
0.018
0.022
0.026
0.030
0.034
0.038
0.042
0.046
0.050
0.054
0.058
0.062
0.066
0.00 0.10 0.20 0.30 0.40 0.50 0.60
Lift coefficient  Cl
D
r
a
g
c
o
e
f
f
i
c
i
e
n
t

C
d
Mach=0.60
Mach=0.80
Mach=0.90
Figure 17.8 Subsonic Drag Model
Drag Coefficient Versus Lift Coefficient {f(Mach)}
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Lift coefficient  CL
D
r
a
g
c
o
e
f
f
i
c
i
e
n
t

C
D
Mach=0.60
Mach=0.80
Mach=0.90
Mach=1.10
Mach=1.40
Mach=1.60
Figure 17.9 Drag Model – All Mach Numbers
193
We now have all of the required components for a sample drag model. This will be used
in combination with a thrustfuel flow model to compute performance parameters. We will
use this to compute performance during cruise, climb, and turn.
17.5 Thrust and Fuel Flow Model
As with the drag model, we will construct a set of equations to represent net thrust and
fuel flow. There will be two separate models. One will be for nonafterburner engine
operation and the other will be for maximum afterburner. We will begin with a fuel flow
model for nonafterburner.
17.6 Thrust Specific Fuel Consumption
Thrust specific fuel consumption ( tsfc ) is simply the ratio of fuel flow to net thrust.
f
n
W
tsfc
F
· (17.6)
The parameter will sometimes generalize by dividing by the square root of the total
temperature ratio.
2 t
tsfc
tsfcr
θ
· (17.7)
2
2
288.15
t
t
T
θ · (17.8)
2
2
(1 0.2 )
t
T T M · ⋅ + ⋅ (17.9)
Ideally, the total temperature would be measured in the engine inlet. However, that
parameter is difficult to measure and even more difficult to model so one usually (but not
always) will use a ram air temperature measurement. Ram air temperature is total temperature.
Figure 17.10 is a sample representation of thrust specific fuel consumption referred
( tsfcr ) versus referred net thrust (
2
/
n t
F δ ). The parameter referred net thrust is net thrust
divided by total pressure ratio at the inlet. In this case, we will use a Pitotstatic derived total
pressure ratio. That means we have assumed zero inlet losses.
2
n
nr
t
F
F
δ
· (17.10)
For M < 1.0:
2 3.5
2
(1 0.2 )
t
M δ δ · ⋅ + ⋅ (17.11)
194
For M ≥ 1:
( )
2.5
7 2
2
166.9216 7 1
t
M M δ δ
]
· ⋅ ⋅ ⋅ −
]
]
(17.12)
TSFC/sqrt(thet2) Vs. Fn/delt2
y = 1.606E16x
4
 2.265E12x
3
+ 1.046E08x
2
 7.792E05x + 1.324E+00
0.8
0.9
1.0
1.1
1.2
1.3
1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 11,000
Fn/delt2  lbs
T
S
F
C
/
s
q
r
t
(
t
h
e
t
2
)
Figure 17.10 Thrust Specific Fuel Consumption
To better illustrate real effects, an additional term will be added to our tsfcr model.
There is, generally, degradation in the parameter with increasing altitude (or decreasing
Reynolds number). We will assume the above curve is valid up to a Reynolds number
corresponding to a standard day at 30,000 feet. The parameter Reynolds number index
( RNI ) is introduced in the lift and drag section. This is the ratio of Reynolds at the test
condition to the Reynolds number at sea level, standard day, for the same test day Mach
number. For standard day, we have the following values for RNI :
a. 30,000 feet RNI = 0.4010, and
b. 50,000 feet RNI = 0.1661.
A typical degradation in tsfcr is on the order of ¼ percent per 1,000 feet of altitude.
Therefore, for 20,000 feet we would have a 5percent degradation. Hence, a formula for a
multiplicative factor on tsfcr would be as follows:
(0.4010 )
1 0.05
(0.4010 0.1661)
tsfcr
RNI
F
−
· + ⋅
−
(17.13)
195
or:
1 (0.4010 ) 0.2129
tsfcr
F RNI · + − ⋅ (17.14)
1.0 0.4010
tsfcr
F if RNI · >
The above multiplicative factor is a number greater than one for Reynolds number
indices less than 0.4010. With that term, we have a simplified model for fuel flow for
nonafterburning. We must emphasize again, that the models presented here are very
simplified and are presented to illustrate general trends only.
17.7 Military Thrust
For maximum thrust without afterburner, usually designated MIL power, we will
construct a generalized form. First, we have already introduced the parameter called referred
net thrust. For our model, we will assume a relationship of referred net thrust versus inlet
total temperature (
2 t
T ).
2 t r t
T T η · ⋅ (17.15)
where:
r
η = inlet temperature recovery factor.
For this model, we will presume that
r
η = 1.0. Usually, the recovery factor is difficult to
measure and even more difficult to model anyhow. Therefore, typically, the
r
η = 1.0
assumption is made with actual data analysis. A turbine engine is often said to be flat rated.
That means that the thrust is constant to some value of inlet total temperature. We will
presume that value to be standard day sea level temperature (288.15 degrees K). After that
point, the thrust will decrease at some lapse rate. We shall presume the lapse rate to be
1 percent per 1.0 degree K. We will take a value of 9,000 pounds as the flat rated value of
referred net thrust. Then, the equation for our model is as follows:
2
9, 000 if 288.15
nr t
F T · ≤ (17.16)
( )
2 2
9, 000 1 0.01 288.15 if 288.15
nr t t
F T T ] · ⋅ − ⋅ − >
]
(17.17)
Figure 17.11 is a graphical representation of the above equations. It should be noted that
this model is highly idealized. An actual model will have altitude and Mach number effects.
For standard day, the model presented in Figure 17.12 is for thrust versus Mach number
as a function of altitude.
196
Referred Net Thrust versus Total Temperature: MIL Thrust
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
200 220 240 260 280 300 320 340 360 380
Total Temperature  deg K
R
e
f
e
r
r
e
d
N
e
t
T
h
r
u
s
t
F
n
/
d
e
l
t
2

l
b
s
Figure 17.11 Military Referred Net Thrust
Net Thrust versus Mach Number (Nonafterburning) Standard Day
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
N
e
t
T
h
r
u
s
t
(
l
b
s
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
35,000 ft
40,000 ft
42,500 ft
Figure 17.12 Military Thrust
197
17.8 Maximum Thrust
For maximum (MAX) thrust, we will construct a similar model. First, the formulas for the
pressure ratio are presented for an assumption of a normal shock inlet. A normal shock inlet is
one where the recovery is across a normal shock. This is just what you have in a Pitot probe.
For the maximum thrust with afterburner model, we were going to use the same lapse rate
(1.0 percent per 1.0 degree K) but ran into the effect of thrust going to zero within the range
of achievable total temperatures. So, a lapse rate of ½percent is used instead. We took a flat
rated value for referred thrust of an even 20,000 pounds. By comparison, the static sea level
uninstalled thrust ratings in the F16 engines are (as of this writing) on the order of in excess
of 25,000 pounds. The equations for referred thrust are as follows:
2
20, 000 if 288.15
nr t
F T · ≤ (17.18)
( )
2 2
20, 000 1 0.005 288.15 if 288.15
nr t t
F T T ] · ⋅ − ⋅ − >
]
(17.19)
A graphical representation of the model is shown in Figure 17.13. This model is also
highly idealized, ignoring Mach number and altitude effects.
Referred Net Thrust versus Total Temperature Maximum Afterburner
Th t
6,00
8,000
10,000
12,000
14,000
16,000
18,000
20,000
22,000
200 220 240 260 280 300 320 340 360 380 400
Total Temperature  deg K
R
e
f
e
r
r
e
d
N
e
t
T
h
r
u
s
t
(
F
n
/
d
e
l
t
2
)
l
b
s
Figure 17.13 Referred Net Thrust for Maximum Thrust
The maximum thrust model is presented as net thrust versus Mach number as a function
of altitude for standard day in Figure 17.14.
198
Net Thrust (with Afterburning) versus Mach Number (Standard Day)
0
5,000
10,000
15,000
20,000
25,000
30,000
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Mach Number
N
e
t
T
h
r
u
s
t
(
l
b
s
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
40,000 ft
50,000 ft
Figure 17.14 Maximum Thrust
For fuel flow during maximum thrust operation, we will assume a very simple model.
Experience has shown that thrust specific fuel consumption during maximum afterburner
operation is at least 2.0. Let us, arbitrarily, assume a value of 2.5:
a. 2.50 tsfcr · .
17.9 Cruise
Using the previously developed drag and fuel flow models, we can compute cruise
parameters. The parameter range factor was developed in the cruise section and is repeated
here.
t
t
f
V
RF W
W
· ⋅ (nam) (17.20)
An equivalent form of the equation is as follows:
661.48
t
f
W
M
RF
W
δ
δ θ
]
 `
⋅ ⋅
]
. ,
] ·
 `
]
]
⋅
. , ]
(17.21)
199
The term in the denominator is called corrected fuel flow and can be expressed in another
form.
f
n
W
F tsfc
δ
δ θ θ
 `  `  `
· ⋅
⋅ . , . , . ,
(17.22)
In order to differentiate between dividing by total or ambient conditions, we will use the
convention of ‘corrected’ for ambient conditions and ‘referred’ for total conditions. Hence,
corrected tsfc
tsfc
tsfcc
θ
 `
·
. ,
(17.23)
2
referred tsfc
t
tsfc
tsfcr
θ
 `
·
. ,
(17.24)
This may not be a universal convention, but will be used in this text.
Combining the range factor in equation 17.21 and corrected fuel flow in equation 17.22
yields:
661.48
t
n
W
M
RF
F
tsfcc
δ
δ
]  `
⋅ ⋅
]
. ,
·
]
 `
]
⋅
] . , ]
(17.25)
The concept behind the old constant weightoverdelta ( /
t
W δ ) method of testing was
that if one kept M and /
t
W δ constant, then drag would be constant. That derived from the
simplified forms of lift and drag coefficient for 1g flight and thrust equals drag.
2
0.000675
t
L
W
C
M S
δ
 `
⋅
. ,
·
⋅
(17.26)
( )
2
0.000675
D
D
C
M S
δ
⋅
·
⋅
(17.27)
( )
n
F
D
δ δ
 `
·
. ,
(17.28)
However, we know that both drag and engine thrust specifics vary with Reynolds
number.
200
17.10 Range
For our model aircraft on a standard day, at 22,500 pounds gross weight, we can compute
the parameter range factor. Figure 17.15 is a plot of range factor for a series of altitudes.
Either the minimum Mach number is dictated by the left scale of the plot, attaining a
maximum lift coefficient or thrust required exceeding the thrust available. The thrust
available is deemed to be that determined from our military thrust model. The maximum lift
coefficient is simply:
a.
max
1.50
L
C · .
We will use the same 1.50 value for maximum lift coefficient for all the problems in this
section.
Range Factor versus Mach Number (Weight=22,500 lbs)
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
5,500
6,000
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
R
a
n
g
e
F
a
c
t
o
r
(
n
m
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
35,000 ft
40,000 ft
42,500 ft
Figure 17.15 Range Factor
By picking off the peaks of the curves we can plot (Figure 17.16) peak range factor
versus weightoverdelta. The topic of optimum flight profiles is a topic that will not be
covered in this section, but suffice it to say that in a sense the closest distance between two
points is not necessarily a straight line.
201
Constant Altitude Cruise: Weight=22,500 lbs: Range Factor versus WeightOverDelta
2,000
2,500
3,000
3,500
4,000
4,500
5,000
5,500
6,000
20,000 40,000 60,000 80,000 100,000 120,000 140,000
Weight/delta  lbs
R
a
n
g
e
F
a
c
t
o
r

N
M
Figure 17.16 Maximum Range Factor
Figure 17.17 illustrates the effect of Reynolds number on cruise performance and demonstrates
that you do not get the same range factor at a given /
t
W δ and Mach number regardless of altitude
(or temperature). This is due to skin friction effects on both aircraft drag and on the engine. The
engine blades are experiencing the same skin friction drag effects as the aircraft wing and other
surfaces. The weightpressure ratio ( /
t
W δ ) is 125,000 pounds for all the data in the next two plots.
Altitude Effect (W/delta=125,000 lbs)
4,800
5,000
5,200
5,400
5,600
5,800
6,000
0.70 0.72 0.74 0.76 0.78 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94
Mach Number
R
a
n
g
e
F
a
c
t
o
r
(
n
m
)
Weight=20,000 lbs
Weight=22,500 lbs
Weight=25,000 lbs
Figure 17.17 Range Factor – Altitude Effect
202
At 0.85 Mach number, Table 17.2 summarizes the numbers off the above plot.
Table 17.2
RANGE FACTOR VARIATION WITH ALTITUDE
Altitude
(ft)
Weight
(lbs)
RNI
Range Factor
(nm)
43,030 20,000 0.2322 5736.7
40,580 22,500 0.2612 5794.3
38,388 25,000 0.2903 5849.7
The percentage change per 1,000foot change in altitude calculates to 0.39 percent. This
number is comparable to the actual flight test derived values shown in the cruise section for
three different aircraft.
Taking the midweight as the baseline, we can also vary temperature and keep altitude and
weight constant. This will achieve a variation in Reynolds number, as shown in Figure 17.18.
Weight=22,500 lbs:Altitude=40,580 ft
4,800
5,000
5,200
5,400
5,600
5,800
6,000
0.72 0.74 0.76 0.78 0.8 0.82 0.84 0.86 0.88 0.9 0.92 0.94
Mach Number
R
a
n
g
e
F
a
c
t
o
r
(
n
m
)
T=+20 deg K
T=Std Day
T=20 deg K
Figure 17.18 Range Factor – Variation with Temperature
At the same 0.85 Mach number and weightpressure ratio, the effect of temperature is
shown in Table 17.3.
203
Table 17.3
RANGE FACTOR VARIATION WITH TEMPERATURE
Temperature Above Standard
(deg K)
20
(196.65)
Std
(216.65)
+20
(236.65)
Reynolds Number Index 0.2977 0.2612 0.2312
Range Factor (nm) 5,836.6 5,794.3 5,736.8
By comparing the numbers Tables 17.2 and 17.3, it can be seen that the slope of range
factor versus Reynolds number index is essentially identical between varying altitude and
weight at constant weightpressure ratio and varying ambient temperature. Both will achieve
a variation in Reynolds number index.
17.11 Endurance
For the case where it is desired to maximize endurance, we would need to find the Mach
number for minimum fuel flow. Figure 17.19 is a plot of fuel flow versus Mach number for
the same weight and altitudes considered for range.
Fuel Flow (Wt=22,500 lbs)
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
5,500
6,000
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
F
u
e
l
F
l
o
w
(
l
b
s
/
h
r
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
35,000 ft
40,000 ft
42,500 ft
Figure 17.19 Fuel Flow  Endurance
17.12 Acceleration Performance
Acceleration performance will be computed using our model. The parameterspecific
excess power (
s
P ) was defined in the axis systems and equations of motion section. To
compute
s
P from our model the following computations are performed. The drag and thrust
models are defined in previous parts of this section.
204
( , , )
D L
C f C M RNI ·
( )
2
0.000675
D
M S C
D
δ ⋅ ⋅ ⋅
· (17.29)
2
2
(1 0.2 )
t
T T M · ⋅ + ⋅ (17.30)
2
( )
nr t
F f T ·
2 n nr t
F F δ · ⋅ (17.31)
288.15
T
θ · (17.32)
1116.45
t
V M θ · ⋅ ⋅ (ft/sec) (17.33)
ex n
F F D · − (17.34)
ex
x
t
F
N
W
· (17.35)
s x t
P N V · ⋅ (17.36)
17.13 Military Thrust Acceleration
For military thrust (maximum without afterburner), our model and the above calculations
produce Figure 17.20 for standard day.
The above altitudes and weights were chosen to be the same as for the cruise. At 42,500
feet, the model computes a just barely positive
s
P , where
s
P could be considered the rate of
climb achievable for constant true airspeed.
To illustrate the effect of temperature on acceleration performance, an altitude of 10,000
feet was chosen for Figure 17.21.
205
MIL Thrust Specific Excess Power (Wt=22,500 lbs)
0
20
40
60
80
100
120
140
160
180
200
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
S
p
e
c
i
f
i
c
E
x
c
e
s
s
P
o
w
e
r

P
s
(
f
t
/
s
e
c
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
35,000 ft
40,000 ft
42,500 ft
Figure 17.20 Military Thrust Specific Excess Power
Ps versus Mach Number (Weight=22,500 lbs)
0
50
100
150
200
250
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
S
p
e
c
i
f
i
c
E
x
c
e
s
s
P
o
w
e
r

P
s
(
f
t
/
s
e
c
)
T=20 deg K
T= Std Day
T=+20 deg K
Figure 17.21 Military Thrust – Specific Excess Power, Temperature Effect
206
The above difference in acceleration (and hence, climb) performance as a function of
temperature is due primarily to thrust. There is, however, a small increase in drag at the
higher temperatures due to skin friction. To repeat the thrust model presented in equations
17.16 and 17.17:
a.
2
9, 000 for 288.15
nr t
F T · < , and
b.
[ ] ( )
2 2
9, 000 1 0.01 288.15 for 288.15
nr t t
F T T · ⋅ − ⋅ − ≥ .
This produces net thrust versus Mach number for 10,000 feet pressure altitude as shown
in Figure 17.22. Drag is also plotted for standard day.
There is a small drag difference due to skin friction as illustrated in Figure 17.23.
At the point of minimum drag, we have the following points from the model. Mach
number is 0.42 in Table 17.4.
Thrust and Drag (10,000 ft; Wt=22,500 lbs)
0
2,000
4,000
6,000
8,000
10,000
12,000
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mach Number
T
h
r
u
s
t
a
n
d
D
r
a
g
(
l
b
s
)
Thrust: T=20 deg K
Thrust: T= Std Day
Thrust: T=+20 deg K
Drag (Std Day)
Figure 17.22 Military Thrust – Thrust and Drag at 10,000 Feet
207
Drag versus Mach Number (Weight = 22,500 lbs; Altitude=10,000 ft)
1,800
1,850
1,900
1,950
2,000
2,050
2,100
2,150
2,200
0.35 0.40 0.45 0.50 0.55
Mach Number
D
r
a
g
(
l
b
s
)
T= 20 deg K
T= Std Day
T= +20 deg K
Figure 17.23 Drag at 10,000 Feet – Temperature Variation
Table 17.4
DRAG VARIATION WITH TEMPERATURE
Temperature
(deg K)
20
(248.3)
Std
(268.3)
+20
(288.3)
Drag (lbs) 1,825.0 1,833.5 1,841.5
Now, this 16.5pound difference in drag, between t20 degrees K of standard day at
10,000 feet, is quite small for purposes of acceleration performance. However, if the aircraft
were doing endurance tests, those 16.5 pounds would be almost a full 1 percent.
17.14 Maximum Thrust Acceleration
The analysis of data for maximum thrust is identical to that for military thrust. It’s just
that the numbers are larger. In addition, we get to travel through the transonic region where
some interesting drag effects may occur. First, we present the standard day
s
P plot in
Figure 17.24.
The thrust model presented earlier had a referred net thrust of 20,000 pounds for total
temperature below 288.15 (standard day sea level). The sea level rating for F16 engines are
somewhat larger than that number. Be aware, however, that a rating is uninstalled. By
installing an engine in the aircraft, you will incur substantial inlet and other losses.
208
Ps versus Mach Number (Weight=22,500 lbs)
0
100
200
300
400
500
600
700
800
900
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Mach Number
S
p
e
c
i
f
i
c
E
x
c
e
s
s
P
o
w
e
r

P
s

(
f
t
/
s
e
c
)
Sea Level
10,000 ft
20,000 ft
30,000 ft
40,000 ft
50,000 ft
Figure 17.24 Maximum Thrust Specific Excess Power
As we did with military thrust, we shall examine the effect of temperature on acceleration
performance. This time we will choose 30,000 feet to conduct a comparison. Note that the
temperature deltas this time are only 10 degrees K, versus 20 degrees K for the military thrust case.
In addition, the thrust model chosen had only a ½percent per degree K slope. This
s
P comparison
is shown in Figure 17.25.
Maximum Thrust Effect of Temperature (Wt=22,500 lbs; 30,000 ft)
0
50
100
150
200
250
300
350
400
0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Mach Number
S
p
e
c
i
f
i
c
E
x
c
e
s
s
P
o
w
e
r

P
s

(
f
t
/
s
e
c
)
T= 20 deg K
T= Std Day
T= + 20 deg K
Figure 17.25 Maximum Thrust Specific Excess Power Temperature Effect at 30,000 Feet
209
We chose to plot only between 0.9 and 1.60 Mach number for a specific reason. The
prototype F16 (YF16) was involved in a flying competition with an aircraft designated the
YF17 (later evolved into the Navy F18) in 1974. One of the performance specification
points was the time to accelerate from 0.9 to 1.6 Mach number at 30,000 feet. There were
other rules: the time would be computed for a standard day and with the weight held constant
at a midcombat weight. To compute time is a simple numerical integration.
( )
0
n
t
s
x
t
t t
F D
V h
P
N
V
W g V
−
· · + ·
! !
(17.37)
We also had zero wind, because the above equation is only valid for zero wind. In
addition, since we are accelerating at constant altitude, the h
!
term is zero.
0
32.174
t x x
V g N N · ⋅ · ⋅
!
(17.38)
32.174
t
x
V
N
t
∆  `
· ⋅
∆
. ,
(17.39)
32.174
t
x
V
t
N
∆
∆ ·
⋅
(17.40)
At 30,000 feet, standard day ambient temperature is –44.44 degrees C (easy number to
remember) = 228.71 degrees K. A little historical footnote here to illustrate the criticality of
getting data at as cold a test day ambient air temperature as possible at 30,000 feet. The
YF17 performance tests were conducted in late summer and early autumn. A specification
compliance condition was the time to accelerate from 0.90 to 1.60 Mach number at 30,000
feet on a standard day. In Appendix A note that the average temperatures at 30,000 feet
above Edwards AFB are all greater than standard day. We were never able to accelerate the
YF17 aircraft to 1.60 Mach number on a test day. The competition (YF16) had no problem
getting to 1.60 Mach number even on days hotter than standard.
228.71
1116.45 994.65
288.15
t
V M M · ⋅ ⋅ · ⋅ (17.41)
994.65
30.915
32.174
x x
M M
t
N N
⋅ ∆ ∆
∆ · · ⋅
⋅
(17.42)
Finally,
1.60
0.9
1
30.915
M
M x
t M
N
·
·
 `
· ⋅ ⋅ ∆
. ,
∑
(17.43)
The results of the time integration as a function of ambient temperature are shown in
Figure 17.26. Also shown is a second thrust model, which is a 25,000pound model with the
same ½percent lapse rate beginning at 288.15 degrees K.
210
Time to Accelerate: 30,000 ft: Weight=22,500 lbs
0
50
100
150
200
250
20 15 10 5 0 5 10 15 20
Delta temp above standard (deg K)
T
i
m
e
t
o
a
c
c
e
l
e
r
a
t
e
0
.
9
t
o
1
.
6
M
a
c
h
N
u
m
b
e
r
(
s
e
c
)
Thrust= 20,000 lbs
Thrust= 25,000 lbs
Figure 17.26 Acceleration Time – Variation with Thrust
17.15 Sustained Turn
A sustained (or stabilized) turn is a constant altitude, constant speed turn. In order to
achieve that condition, thrust must equal drag.
cos( )
n g t e
F F i F D α · ⋅ + − · (17.44)
For this example, we will ignore the angleofattack component and simplify to:
n
F D · (17.45)
We will make a similar simplification in the normal axis (perpendicular to the velocity
vector).
z t
L N W · ⋅ (17.46)
Knowing thrust, compute drag, then drag coefficient. From drag coefficient, find lift
coefficient, then lift, then solve for
z
N . Since we do not usually have lift coefficient as a
function of drag coefficient, an iteration scheme is required. Here are the basics of what was
used in this example.
We know drag coefficient from the following:
211
2
0.000675
n
D
F
C
M S δ
⋅
·
⋅ ⋅
(17.47)
Begin at 1g, but use some positive drag polar slope for the first iteration, such as 0.10.
This is necessary since the slope of the drag polar at 1g may be zero or even negative.
( )
2
2 2
0.1
( )
Dnew Dold
D
L
L Lold new
C C
C
C
C C
−
∆
· ·
∆
−
(17.48)
For the first iteration, the old values of
L
C
and
D
C
are the 1g values. We always know
the new
D
C . It is the one above, computed from the available net thrust. Solve for
Lnew
C
from
the above equation. After the first iteration, compute values for the slope numerically by
choosing some small change in lift coefficient and computing the slope. For instance, we
used 0.01.
2
2 2
( ( 0.01)) ( ( ))
( 0.01)
D L D L
D
L
L L
C f C C f C
C
C
C C
+ −
∆
·
∆
+ −
(17.49)
Then, just simply repeat the process a few times until the change in
L
C is sufficiently
small (say < 0.001) between steps. Now that you know lift coefficient, then just compute
z
N .
The results for maximum thrust are shown in Figure 17.27.
Nz versus Mach Number (Wt= 22,500 lbs)
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
Mach Number
N
o
r
m
a
l
L
o
a
d
F
a
c
t
o
r

N
z
Sea Level
10,000 ft
20,000 ft
30,000 ft
40,000 ft
Figure 17.27 Maximum Thrust – Sustained Turn Normal Load Factor
212
The constraints imposed on this turn problem were the following.
a.
max L L
C C < ,
b.
max
1.50
L
C · ,
c.
max z z
N N < , and
d.
max
9.0
z
N · .
213
18.0 CRUISE FUEL FLOW MODELING
This section had contained a regression analysis model of fuel flow and thrust extracted
from the AFFTC C17A (Figure 18.1) testing report titled, “C17 Cruise Configuration
Performance Evaluation” (Reference 18.1), but since this handbook is intended for public
viewing, it was necessary to delete the scales on the data plots shown in this section.
Figure 18.1 C17A Aircraft
( )
661.48
t
f
W
RF M
W
δ
δ θ
· ⋅ ⋅
 `
. ,
(18.1)
Solving for corrected fuel flow.
( )
661.48
t
f
fC
W
W
W M
RF
δ
δ θ
 `
· · ⋅ ⋅
. ,
(18.2)
The lift coefficient was computed using the curve fits for angle of attack (α ) and gross
thrust (
g
F ) provided in the report (Reference 18.1). Pressure ratio (δ ) formulas used are
found in the altitude section.
0.000675 sin( )
g
t
L
F
W
C α
δ δ
]
· ⋅ − ⋅
]
]
(18.3)
Since the data presented in the report (Reference 18.1) were corrected to a reference
Reynolds number, an estimate of drag at test and reference conditions was computed. Instead
214
of the usual ‘standardization’ we are essentially ‘unstandardizing’ the drag data. We are
going from a reference condition to a standard condition. The formulas used are those
presented in the lift and drag section.
The reference wing area ( S ) and the wetted area (
wet
S ) are as follows:
a. S ·3,800. ft
2
, and
b.
wet
S ·19,075. ft
2
.
Skin friction drag relationships are as follows:
2.58
10
0.455/ log ( )
f
C RN · (18.4)
2 0.65
/(1 0.144 )
fC f
C C M · + ⋅ (18.5)
wet
D fC f
S
C C
S
· ⋅ (18.6)
The assumption was made that the characteristic length used was the mean aerodynamic
chord ( MAC ). That value is as follows:
l MAC · · 25.794 feet.
To perform a curve fit of the fuel flow data, we will remove the skin friction drag
correction from the thrust data. The standard day drag coefficient (
Ds
C ) was computed from
the drag polar curve fit formulas in the report. The drag coefficient formula in the report was
referenced to a Reynolds number of 1,800,000 per foot. The test day drag coefficient (
Dt
C )
was computed as follows:
( )
t
D Ds Dft Dfs
C C C C · + − (18.7)
The standard (or reference) skin friction drag coefficient is based upon the standard Reynolds
number per foot and the characteristic length. Inserting these numbers into equation 18.4:
2.58
10
0.455/ log (1, 800, 000 25.794)
fs
C · ⋅ = 0.00238 (18.8)
From a formula defined in the lift and drag section,
( )
2
110
398.15
T
RNI
δ
θ
] +
 `
· ⋅
]
. ,
]
(18.9)
6
7.101 10 RN M RNI l · ⋅ ⋅ ⋅ ⋅
215
Finally, the test values of corrected thrust are computed. Note a distinction between test
values and test day, since the data points are still at standard day temperatures. We will take
out the correction to a reference Reynolds number.
[ ]
2
/
0.000675
Dt
n
t
C M S
F δ
⋅ ⋅
· (18.10)
18.1 Thrust Specific Fuel Consumption
Next, we compute the thrust specific fuel consumption corrected as follows:
( )
[ ]
/
/
/
f
t
C
n
t
W
TSFC TSFC
F
δ θ
θ
δ
]
⋅
]
· · (18.11)
The following (Figure 18.2) is a plot of the 141 data points being analyzed. Even though
the plot has no scales, it will however give you some interesting information. The maximum
value of the dependent variable ( / tsfc θ ) is 11.2 percent greater than the mean and the
minimum value if 17.9 percent less than the mean. The 1sigma about the mean is
7.0 percent. This is a large variation, however, it should be noted that range factor had a 14.3
percent variation about its mean (more than twice as much – percentage wise). The use of
these ‘generalizing’ parameters is a good first step in modeling your data. That is analogous
to drag where we use lift and drag coefficients to aid in modeling. We still wish to reduce this
variation, so we proceed to curve fit the data using multiple regression.
TSFC/sqrt(theta) versus Fn/delta
[H1<10,000; H2 20,000 to 30,000; H3> 30,000 feet]
0.52
0.56
0.60
0.64
0.68
0.72
0.76
20000 40000 60000 80000 100000 120000 140000
Fn/delta (lbs)
T
S
F
C
/
s
q
r
t
(
t
h
e
t
a
)
(
1
/
h
r
)
H1: M <= 0.46
H1: M 0.50 to 0.60
H2: M 0.50 to 0.60
H2: M=0.65
H2: M 0.70 to 0.76
H2: M 0.77 to 0.81
H3: M 0.70 to 0.76
H3: M 0.77 to 0.825
Figure 18.2 Thrust Specific Fuel Consumption
216
18.2 Multiple Regression
Now, we will strive to develop an equation that fits the data presented in Figure 18.2. The
simplest possible equation is a constant. We will use Reynolds number index ( RNI ) as an
altitude parameter. In general, the formula will be as follows:
( ) ( )
/ / , ,
n
TSFC f F M RNI θ δ · (18.12)
For ease of representation, we will make the following variable name changes:
a. / Y TSFC θ · ,
b. 1 /
n
X F δ · ,
c. 2 X M · , and
d. 3 X RNI · .
Then, equivalently:
( ) 1, 2, 3 Y f X X X · (18.13)
The author used MS Excel¹ to evaluate the data. Excel has matrix operators, however it
was necessary to develop a multiple regression method for use with Excel. For those who do
not have a multiple regression program available, the following is the formulation for
multiple regression.
The general case for linear multiple regression:
0 1 1 2 2 m m
Y a a X a X a X · + ⋅ + ⋅ + + ⋅ " (18.14)
The coefficients are solved by the following:
1
1, 2, ,
0
2
1, 1, 2, 1, 1, , 1,
1
2
2
2, 2, 2, 1, 2, 2, ,
2
, , , 1, , 2, ,
i i m i i
i i i i i m i i i
i i i i i i i m i
m
m i i m i m i i m i i m i
N X X X Y
a
X X X X X X X Y
a
a
X Y X X X X X X
a
X Y X X X X X X
−
]
]
]
]
] ⋅ ⋅ ⋅
]
]
] · ⋅
] ⋅ ⋅ ⋅
]
]
]
]
]
]
]
⋅ ⋅ ⋅
] ]
∑ ∑ ∑ ∑
∑ ∑ ∑ ∑ ∑
∑ ∑ ∑ ∑ ∑
∑ ∑ ∑ ∑ ∑
"
"
"
#
# # # # #
"
]
]
]
]
]
]
]
]
]
]
(18.15)
where:
N = number of data points.
The above general curve fit formula was developed by minimizing the sum of the squares
of the residual errors ( SS ). The formula for SS is as follows:
217
( )
2
ˆ
i i
SS Y Y · −
∑
(18.16)
where:
ˆ
Y = the curve fit equation.
There are a number of ways to evaluate the quality of a curve fit. We will look at the
standard deviation.
/( 1) SS N σ · − (18.17)
A percentage standard deviation will be calculated,
% ( / ) 100 Y σ σ · ⋅ (18.18)
where:
Y = the mean value of the independent variable.
Here are the results of the curve fits:
a.
0
ˆ
Y a · % 7.00% σ · ,
b.
0 1
ˆ
1 Y a a X · + ⋅ % 5.30% σ · , and
c.
2
0 1 2
ˆ
1 1 Y a a X a X · + ⋅ + ⋅ % 5.16% σ · .
At this point, we should pause to examine the residual errors rather than just blindly
adding additional terms to the equation. From Figure 18.3, we can see some apparent
additional Mach number and Reynolds number effects. So far, we have only reduced the
1sigma about the mean from 7.0 percent to 5.16 percent. This is a disappointing result;
however, we suspect there may be a substantial altitude and Mach number effect. The
parameter we will plot is the percentage error as follows:
( )
ˆ
%
100
Y
Error Y Y
]
· − ⋅
]
]
(18.19)
The
ˆ
Y used will be from the last curve fit (equation 18.18).
218
% Error in TSFC/sqrt(theta) Versus Mach Number
20
15
10
5
0
5
10
15
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85
Mach Number
%
E
r
r
o
r
i
n
T
S
F
C
/
s
q
r
t
(
t
h
e
t
a
)
Altitude < 20,000 Feet
Altitude 20,000 to 30,000 Feet
Altitude > 30,000 Feet
Figure 18.3 Percentage Error in Thrust Specific Fuel Consumption
We can now proceed to add additional terms to our model.
a.
2
0 1 2 3
1 1 2 Y a a X a X a X · + ⋅ + ⋅ + ⋅ % 1.237% σ · ,
b.
2
0 1 2 3 4
ˆ
1 1 2 3 Y a a X a X a X a X · + ⋅ + ⋅ + ⋅ + ⋅ % 1.230% σ · ,
c.
2 2
0 1 2 3 4 5
ˆ
1 1 2 3 2 Y a a X a X a X a X a X · + ⋅ + ⋅ + ⋅ + ⋅ + ⋅ % 1.229% σ · , and
d.
2 2 2
0 1 2 3 4 5 6
ˆ
1 1 2 3 2 3 Y a a X a X a X a X a X a X · + ⋅ + ⋅ + ⋅ + ⋅ + ⋅ + ⋅ % 1.224% σ · .
At this point, no significant additional gains are evident. Actually, we did not make
significant gains past equation (a) but proceeded just to illustrate what additional gains were
made. This particular data set was not a very good one to develop a complete fuel flow
model. There were no data collected below 6,000 feet pressure altitude, for instance. Only
stabilized cruise data points were used. Throttle settings above and below that required for
stabilized cruise should be included in any fuel flow model.
The C17A project (Reference 18.1) illustrates that too much time was expended
collecting cruise data. Enormous quantities of flight time were expended to collect these
relatively few cruise data points. The stabilization criterion was much too stringent. To quote
from the report (Reference 18.1), “it was not uncommon for a single cruise point to take 20
minutes to complete.” They required “not less than 2.5 minutes of stabilized data” on each
data point. There is no reason for that with the advent of INS and GPS measurements to give
219
instantaneous acceleration data. Once some reasonable stabilization is achieved, a few
seconds of data is all that is required. With the addition of a series of accelerations and
decelerations at partial thrust, a much more complete fuel flow model could have been
obtained at a much lower cost in terms of flight time.
To present just a few of the data points we choose to present those that illustrate an
altitude effect. The data points are all from the aforementioned C17 Cruise Performance
report (Reference 18.1). Range factor variation with altitude is shown in Figure 18.4.
Range Factor versus Altitude
9000
10000
11000
12000
10,000 15,000 20,000 25,000 30,000 35,000 40,000
Pressure Altitude (FT)
R
a
n
g
e
F
a
c
t
o
r
(
n
m
)
Model:M=0.60;W/delta=1,100,000
Data: M=0.6;W/delta=1,100,000
Model:M=0.77;W/delta=1,800,000
Data:M=0.77;W/delta=1,800,000
1,000
Figure 18.4 Range Factor Variation with Altitude
The degradation factor of range factor with altitude was 0.20 percent per 1,000 feet at
1,100,000 pounds /
t
W δ and 0.26 percent per 1,000 feet at 1,800,000 pounds /
t
W δ . This is
more than a factor of two less than the degradation factor of older generation aircraft such as
B52 aircraft.
SECTION 18.0 REFERENCE
18.1 Weisenseel, Charles W. and Chester Gong, C17 Cruise Configuration Performance
Evaluation, AFFTCTR9323, AFFTC, Edwards AFB, California, December 1993.
220
19.0 EQUATIONS AND CONSTANTS
This section is a summary of the primary equations and constants that were derived and
used in this handbook.
19.1 Equations
Acceleration factor
0
1
t t E
V dV H
AF
g dH H
 `  `
 `
· + ⋅ ·
. ,
. , . ,
!
!
Aircraft geometric height (Edwards flyby tower) 31.422 (grid reading)
tower
h ∆ · ⋅
Aircraft pressure altitude (flyby tower data)
/
std
C a c p tower tower
T
H H h
T
 `
· + ∆ ⋅
. ,
Alpha transformation body to flight path
[ ]
cos 0 sin
0 1 0
sin 0 cos
α α
α
α α
]
]
·
]
] −
]
Angle of attack ( )
1
tan
bz bx
V V α
−
·
Angle of attack (zero bank) α θ γ · −
Angle of sideslip
( )
1
sin
by t
V V β
−
·
Aspect ratio
2
b
AR
S
·
Beta transformation body to flight path
[ ]
cos sin 0
sin cos 0
0 0 1
β β
β β β
]
]
· −
]
]
]
Body axis airspeeds
[ ] [ ] [ ]
bx tN
T T T
by tE
bz tD
V V
V V
V V
φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
Body axis pitch rate cos cos sin q θ φ ψ θ φ · ⋅ + ⋅ ⋅
!
!
Body axis roll rate sin p φ ψ θ · − ⋅
!
!
Body axis yaw rate cos cos sin r ψ θ φ θ φ · ⋅ ⋅ − ⋅
!
!
221
Calibrated airspeed ( )
C SL
V a <
3.5
2
1 0.2 1
C C
SL SL
q V
P a
]
 `
· + ⋅ −
]
. ,
]
Calibrated airspeed ( )
C SL
V a <
(1 3.5)
5 1 1
C
C SL
SL
q
V a
P
¹ ¹
]
¹ ¹
 `
· ⋅ ⋅ + −
' ' ]
. ,
¹ ¹ ] ¹ ¹
Calibrated airspeed ( )
C SL
V a ≥
( )
( )
7
2.5
2
166.9216
1
7 1
C SL
C
SL
C SL
V a
q
P
V a
⋅
· −
]
⋅ −
]
Calibrated airspeed ( )
C SL
V a ≥
2.5
2
1
0.881285 1 1
7
C
C SL
SL
C
SL
q
V a
P
V
a
¹ ¹
]
]
¹ ¹
]
]
¹ ¹
 ` ] ¹ ¹ ]
· ⋅ ⋅ + ⋅ −
' '
] ]
 ` . ,
¹ ¹
] ]
⋅
¹ ¹
] ]
. , ]
]
¹ ¹
¹ ¹
Cloverleaf method solves this equation
2 2 2
( ) ( ) ( )
ti t gN wN gE wE
V V V V V V + ∆ · + + +
Compressible dynamic pressure( ) 1 M <
( )
3.5
2
1 0.2 1
c
q
M
P
· + ⋅ −
Compressible dynamic pressure ( 1) M ≥
( )
7
2.5
2
166.9216 1
7 1
C
q
M
P
M
]
]
· ⋅ −
]
⋅ −
]
Corrected net thrust /
n
F δ
Corrected thrust specific fuel consumption
( )
/
f
f
n
n
W
W
tsfc
F
F
δ θ
θ
θ
δ
 `
⋅
. ,
· ·
 ` ⋅
. ,
Density altitude
( )
1
4.2559
1 / 6.87559 6
d
H E
δ
θ
]
 `
]
· − −
] . ,
]
Density ratio
δ
σ
θ
·
Drag (test day)
t nt ext
D F F · −
222
Drag coefficient ( ) /
D
C D q S · ⋅
Drag coefficient
( )
2
0.00067506
D
C D M S δ · ⋅ ⋅ ⋅
Drag coefficient due to skin friction
wet
D f
S
C C
S
 `
· ⋅
. ,
Drag Model (given M ) ( ) ( )
2 2
min min
1 2
D D L L L Lb
C C K C C K C C · + ⋅ − + ⋅ −
2 0 K · when
L Lb
C C <
Earth axis winds
[ ] [ ] [ ] [ ] [ ]
0
0
wN t gN
wE gE
wD gD
V V V
V V
V V
ψ θ φ α β
¹ ¹
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅ ⋅ −
' ' ' ' ' '
¹ ¹ ¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
¹ ¹
Elliptic Wing Theory (M <<1)
2
2
2
1
L
L
L D
C
C C
AR
AR
π
α
π
⋅
· ⋅ ·
⋅  `
+
. ,
Energy altitude
( )
2
0
2
t
E
V
H H
g
· +
⋅
Energy per unit weight
( )
2
0
/
2
t
t
t t
PE KE
V
E W H
g
W W
]
· + · +
]
⋅
]
Equivalent airspeed
e t
V V σ · ⋅
Excess thrust
ex x t
F N W · ⋅
Excess thrust [ cos( ) ]
ex g t e
F F i F D α · ⋅ + − −
Excess thrust test
t
ex x t
F N W · ⋅
Flight path accelerations
cos sin 0 cos 0 sin
sin cos 0 0 1 0
0 0 1 sin 0 cos
x bx
y by
z bz
A A
A A
A A
β β α α
β β
α α
¹ ¹ ] ] ¹ ¹
¹ ¹ ¹ ¹
] ]
· − ⋅ ⋅
' ' ' '
] ]
¹ ¹ ¹ ¹
] ] −
¹ ¹ ] ] ¹ ¹
223
Flight path accelerations
[ ] [ ] [ ] [ ] [ ]
x N
T T T T T
y E
z D
A A
A A
A A
β α φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
Flight path angle
1
sin
t
h
V
γ
−  `
·
. ,
!
Flight path load factors
0
0
0
x xf
y yf
z zf
N A g
N A g
N A g
¹ ¹
¹ ¹
¹ ¹ ¹ ¹
·
' ' ' '
¹ ¹ ¹ ¹
−
¹ ¹
¹ ¹
Flight path to earth axis transform
( )
( )
( )
[ ] [ ] [ ] [ ] [ ]
0
0
gN wN
t
gE wE
gD wD
V V
V
V V
V V
ψ θ φ α β
¹ ¹
+
¹ ¹
¹ ¹
¹ ¹ ¹ ¹
+ · ⋅ ⋅ ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹
+
¹ ¹
¹ ¹
Fuel flow
t
f
dW
W
dt
 `
· −
. ,
Geopotential altitude
0
g dh g dH ⋅ · ⋅
Geopotential vs. geometric altitude
( )
0
0
r
H h
r h
]
· ⋅
]
+
]
]
Gross thrust
( ) ( )
g a f exit exit exit
F W W V A P P · + ⋅ + ⋅ −
!
Groundspeed east sin( )
gE g g
V V σ · ⋅
Groundspeed north cos( )
gN g g
V V σ · ⋅
Heading matrix (rotate about the z axis (or yaw))
[ ]
cos sin 0
sin cos 0
0 0 1
ψ ψ
ψ ψ ψ
− ]
]
·
]
]
]
Heating value corrected fuel flow
18, 400
test
ft ft
LHV
W W
 `
· ⋅
. ,
Ideal gas equation of state P R T ρ · ⋅ ⋅
224
Incompressible dynamic pressure
2 2
0.5 0.5
t SL e
q V V ρ ρ · ⋅ ⋅ · ⋅ ⋅
Inverse square gravity law
( )
2
0
0
0
r
g g
r h
]
· ⋅
]
+
]
]
Kinetic energy
2
0
0.5
t
t
W
KE V
g
 `
· ⋅ ⋅
. ,
Laminar skin friction empirical formula
1.328
f
C
RN
·
Lateral load factor
0
/
y y
N A g ·
Lift coefficient ( ) /
L
C L q S · ⋅
Lift coefficient
( )
2
0.00067506
L
C L M S δ · ⋅ ⋅ ⋅
Longitudinal load factor
0 x t t
N H V V g · +
! !
Longitudinal load factor
0
/
x x
N A g ·
Mach number
t
V
M
a
·
Mach number ( ) 1 M ≥
2.5
2
1
0.881285 1 1
7
C
q
M
P
M
]
 `
 `
]
· ⋅ + ⋅ −
] . , ] ⋅
] . ,
]
]
Mach number ( ) 1 M <
[ ] 1 3.5
5 1 1
C
q
M
P
]
¹ ¹
¹ ¹
 `
· ⋅ + −
] ' '
. ,
¹ ¹ ] ¹ ¹ ]
Mach number from equivalent airspeed
( )
e
SL
V
M
a δ
·
⋅
Normal load factor
0
/
z z
N A g · −
Normal load factor in climb
0
cos
t
z
V
N
g
γ
γ
⋅
· +
!
225
Normal load factor in turn (constant altitude, zero wind)
2
0
1
t
z
V
N
g
σ
 `
· + ⋅
. ,
!
Normal load factor in turn (constant altitude, zero wind)
1
cos
z
N
φ
·
Normal load factor times weight sin( )
z t g t
N W L F i α ⋅ · + ⋅ +
Pitch matrix (rotate about yaxis) [ ]
cos 0 sin
0 1 0
sin 0 cos
θ θ
θ
θ θ
]
]
·
]
] −
]
Potential energy
t
PE W H · ⋅
Pressure altitude above 36,089 feet
( )
36089.24 20805.84 ln
0.22336
C
H
δ
· − ⋅
Pressure altitude below 36,089 feet
( )
( )
( )
1 5.2559
1
6.87559 6
H
E
δ
]
−
]
·
−
Pressure ratio
SL
P
P
δ ·
Pressure ratio above 36,089 feet
[ ] ( ) { ¦ 4.806343 5 36089.24
0.22336
C
E H
e δ
− − ⋅ −
· ⋅
Pressure ratio below 36,089 feet ( )
5.2559
1 6.87559 6 E H δ · − − ⋅
Ram drag
r a t
F W V · ⋅
!
Range (approximate) ln
ts
te
W
R RF
W
 `
· ⋅
. ,
Range factor
t
t t
f
V
RF W SR W
W
· ⋅ · ⋅
Range for constant altitude (approximate)
( )
661.48
te
ts
t
W
t W
f
W
M
dt
R
W
W
δ
δ θ
 `
⋅ ⋅
. ,
· −
 `
⋅
. ,
∫
226
Range for constant altitude (approximate)
te
ts
W
t W
dt
R RF
W
· − ⋅
∫
Range for cruise at constant altitude
( )
661.48
te
ts
t
W
t W
f
W
M
dt
R
W
W
δ
δ θ
 `
⋅ ⋅
. ,
· − ⋅
 `
⋅
. ,
∫
Range for cruise at constant altitude
t
R V dt · ⋅
∫
Reynolds number
t
V l
RN
ρ
µ
⋅ ⋅
·
Reynolds number (7.101 6) RN E M l RNI · + ⋅ ⋅ ⋅
Reynolds number index
( )
2
110
398.15
T
RNI
δ
θ
] +
 `
· ⋅
]
. ,
]
Roll matrix (rotate about xaxis) [ ]
1 0 0
0 cos sin
0 sin cos
φ φ φ
φ φ
]
]
· −
]
]
]
Sideslip matrix [ ]
cos sin 0
sin cos 0
0 0 1
β β
β β β
− ]
]
·
]
]
]
Slender Body Theory ( ) 1 M ≈
2
2
2
L
L
L D
C
C AR C
AR
π
α
π
⋅
· ⋅ ⋅ ·
⋅
Specific excess power
( )
0
t
s E t x t
V
P H H V N V
g
]  `
· · + ⋅ · ⋅
]
. ,
]
! ! !
Specific range
t
f
V
SR
W
·
Speed of sound ( ) 661.48 a R T γ θ · ⋅ ⋅ · ⋅
227
Standard day density ratio ( )
4.2559
1 6.87559 6
C
E H
δ
σ
θ
· · − − ⋅
Standard temperature above 36,089 feet
0
T = 216.65 °K
Standard temperature below 36,089 feet 288.15 1.9812 3
C
T E H · − − ⋅
Standardized drag ( )
s t s t
D D D D
′ ′
· + −
Standardized excess thrust ( ) ( )
s t
ex ex ns s nt t
F F F D F D
′ ′ ′ ′
· + − − −
Standardized fuel flow
( )
fs ft fs ft
W W W W ′ ′ · + −
Standardized net thrust ( )
ns nt ns nt
F F F F ′ ′ · + −
Takeoff excess thrust ( ) cos( ) sin( )
ex t rw n t rw
F W L F D W µ θ θ + ⋅ ⋅ − · − − ⋅
Temperature correction to pressure altitude change
C
STD
T
h H
T
 `
∆ · ⋅ ∆
. ,
Temperature ratio
288.15
SL
T T
T
θ · ·
Theoretical tanker downwash angle
( )
( )
0
2
Lt
t
C
AR
ε
π
⋅
·
⋅
Thin Wing Theory (M > 1)
2
2
2
4 1
4
1
L
L D L L
M
C C C C
M
α
α
⋅ −
· · ⋅ · ⋅
−
Thrust horsepower
550
n t
F V
THP
⋅
· (where
t
V has units of feet/sec)
Thrust horsepower (user provided and n η )
( )
n
THP BHP η σ · ⋅ ⋅
Total energy E KE PE · +
Total temperature
( )
2
1 0.2
t
T T M · ⋅ + ⋅
228
True airspeed
( )
2 2 2
t bx by bz
V V V V · + +
True airspeed down
tD gD wD
V V V · +
True airspeed east
tE gE wE
V V V · +
True airspeed magnitude
( )
2 2 2
t tN tE tD
V V V V · + +
True airspeed north
tN gN wN
V V V · +
True airspeed vector
t g w
V V V · +
" " "
True airspeed vector [ ] [ ] [ ] [ ] [ ] 0
0
t tN
T T T T T
tE
tD
V V
V
V
β α φ θ ψ
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
· ⋅ ⋅ ⋅ ⋅ ⋅
' ' ' '
¹ ¹ ¹ ¹
¹ ¹ ¹ ¹
Turbulent skin friction empirical formula
2.58
10
0.455
(log )
f
C
RN
·
Turn radius (constant altitude, zero wind)
( )
2
2
0
1
t
z
V
R
g N
·
⋅ −
Turn radius (constant altitude, zero wind)
t
g
V
R
σ · !
Velocity rate corrections
0
0
0
i
i
i
bx bx x
by by y
bz bz z
V V r q l
V V r p l
V V q p l
¹ ¹ − ¹ ¹ ] ¹ ¹
¹ ¹ ¹ ¹ ¹ ¹
]
· + − ⋅
' ' ' ' ' '
]
¹ ¹ ¹ ¹ ¹ ¹
] −
¹ ¹ ] ¹ ¹
¹ ¹
Weight
0 t
W m g · ⋅
229
19.2 Constants
Conversion feet to meters = multiply feet by 0.3048 (exactly)
Conversion knots to feet/sec = multiply knots by 1.68781
Nautical mile ( NM ) = 1,852 meters
= 6,076.1155 feet
Reference gravity (
0
g ) = 32.17405 feet/sec²
Reference radius of the earth (
0
r ) (from the 1976 U.S. Standard Atmosphere) = 20,855,553 feet
Sea level standard temperature (
SL
T ) = 288.15 °K
Speed of sound at sea level standard day (
SL
a ) = 1,116.4505 feet/sec
= 661.4788 knots
Standard sea level pressure (
SL
P ) = 101,325 pascals (newtons/m
2
)
= 2,116.2166 pounds/feet²
Temperature in second segment of standard atmosphere (
0
T ) = 216.65 °K
Universal gas constant ( R ) 3,089.8136 feet²/(sec²°K)
Viscosity at sea level (
SL
µ ) = 3.7373⋅10
7
slugs/(feet sec)
230
This page intentionally left blank.
231
APPENDIX A
AVERAGE WINDS AND TEMPERATURES FOR
THE AIR FORCE FLIGHT TEST CENTER
232
This page intentionally left blank.
233
AVERAGE WINDS AND TEMPERATURES FOR
THE AIR FORCE FLIGHT TEST CENTER
The following average wind and temperature data were provided courtesy of the Edwards
AFB weather squadron. The data represents average values obtained on a daily basis over a
period of more than 30 years (1950s through 1980s). Figures A1 through A5 represent
average temperature deviation data versus month for 10, 20, 30, 40, and 50,000 feet pressure
altitude, respectively.
Temperature from Standar d: Pressure Altitude = 10,000 Feet; AFFTC Average
Data; Temperature Standard = 268.34 deg K
0
2
4
6
8
10
12
14
16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
f
r
o
m
S
t
a
n
d
a
r
d
(
D
e
g
K
)
Figure A1 Delta Temperature at 10,000 Feet
Temperature from Standar d: Pressur e Altitude = 20,000 Feet; Average AFFTC
Data; Standard Temperature = 248.53 deg K
0
2
4
6
8
10
12
14
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
(
D
e
g
K
)
Figure A2 Delta Temperature at 20,000 Feet
234
Temperature From Standard: Pressure Altitude = 30,000 Feet; Average AFFTC
Data; Temperature Standard = 228.71 Deg K
2
0
2
4
6
8
10
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
(
D
e
g
K
)
Figure A3 Delta Temperature at 30,000 Feet
Temperature from Standar d: Pressur e Al titude = 40, 000 Feet: AFFTC aver age
data; Standard Temperature = 216.65 deg K
4
3
2
1
0
1
2
3
4
Jan Feb Mar Apr Ma y Jun Jul Aug Sep Oct Nov Dec
Month
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
(
D
e
g
K
)
Figure A4 Delta Temperature at 40,000 Feet
235
Temperatur e from Standard: Pr essure Alti tude = 50,000 Feet : AFFTC Average
data; Standard Temper ature = 216.65 deg K
12
10
8
6
4
2
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
f
r
o
m
S
t
a
n
d
a
r
d
(
D
e
g
K
)
Figure A5 Delta Temperature at 50,000 Feet
Figures A6 and A7 present average windspeed and direction versus month. They are
presented at three different ambient pressure levels. These are in terms of pressures in
millibar (mb). The following are the corresponding pressure altitudes:
1. 200 mb = 38,661 feet,
2. 400 mb = 23,574 feet, and
3. 600 mb = 13,801 feet.
Wi nd Direction versus Month
180
200
220
240
260
280
300
320
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
W
i
n
d
D
i
r
e
c
t
i
o
n
(
D
e
g
r
e
e
s
f
r
o
m
T
r
u
e
N
o
r
t
h
)
P = 200 mb
P = 400 mb
P = 600 mb
Figure A6 Wind Direction
236
Windspeed versus Month
0
10
20
30
40
50
60
70
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
W
i
n
d
S
p
e
e
d
(
k
t
s
)
P = 200 mb
P = 400 mb
P = 600 mb
Figure A7 Windspeed
On a given day, the geometric height will not be equal to the pressure altitude. Figure A8
illustrates this difference for an average day above Edwards AFB. As can be seen, the
geometric height (on average) is always greater than the pressure altitude. This is due to the
fact (again on average) that the atmospheric temperature is greater than standard day for all
months of the year through 30,000 feet.
Geometric Height  Pressure Altitude versus Month
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
G
e
o
m
e
t
r
i
c
H
e
i
g
h
t

P
r
e
s
s
u
r
e
A
l
t
i
t
u
d
e
(
F
e
e
t
)
P = 200 mb
P = 400 mb
P = 600 mb
Figure A8 Geometric Height minus Pressure Altitude
237
APPENDIX B
WEATHER TIME HISTORIES
238
This page intentionally left blank.
239
WEATHER TIME HISTORIES
The following charts represent time histories of data for September through October
1998. On the charts, the terminology flight level (FL) is used. Flight level is pressure altitude
in feet divided by 100. Figure B1 shows the variation of delta temperature above standard
versus date.
Delta Temperature versus Date
4
2
0
2
4
6
8
10
12
14
16
22 Sep 26Sep 30Sep 4Oct 8Oct 12 Oct 16Oct 20Oct 24 Oct
Dat e ( 1998)
D
e
l
t
a
T
e
m
p
e
r
a
t
u
r
e
A
b
o
v
e
S
t
a
n
d
a
r
d
D
a
y
(
D
e
g
r
e
e
s
K
)
FL = 400
FL = 300
FL = 200
FL = 100
Note: /100
C
FL H ·
Figure B1 Delta Temperature Time History
Figures B2 and B3 illustrate the variation in windspeed and direction versus date at flight
levels of 100, 200, 300 and 400, respectively.
240
Wind Directi on versus Date
90
120
150
180
210
240
270
300
330
360
22 Sep 26Sep 30 Sep 4Oct 8Oct 12Oct 16Oct 20Oct 24Oct
Date (1998)
W
i
n
d
D
i
r
e
c
t
i
o
n
(
D
e
g
r
e
e
s
f
r
o
m
T
r
u
e
N
o
r
t
h
)
FL = 400
FL = 300
FL = 200
FL = 100
Figure B2 Wind Direction Time History
Wind Speed versus Date
0
20
40
60
80
100
120
22Sep 26Sep 30 Sep 4Oct 8 Oct 12 Oct 16 Oct 20Oct 24Oct
Dat e ( 1998)
W
i
n
d
S
p
e
e
d
(
K
n
o
t
s
)
FL = 400
FL = 300
FL = 200
FL = 100
Figure B3 Windspeed Time History
241
APPENDIX C
AVERAGE SURFACE WEATHER FOR
THE AIR FORCE FLIGHT TEST CENTER
242
This page intentionally left blank.
243
AVERAGE SURFACE WEATHER FOR
THE AIR FORCE FLIGHT TEST CENTER
Figure C1 shows the average surface temperature for the Air Force Flight Test Center.
Average Surface Temperatures
20
30
40
50
60
70
80
90
100
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
T
e
m
p
e
r
a
t
u
r
e
(
d
e
g
F
)
Maximum
Mini mum
Figure C1 Average Maximum and Minimum Surface Temperatures
244
This page intentionally left blank.
245
BIBLIOGRAPHY
1. Military Specification, Manuals, Flight, MILM7700D, 14 February 1990. (Out of print).
2. Bowles, Jeff V. and Thomas Galloway, Computer Programs for Estimating Takeoff and
Landing Performance, NASA TMX62, 333, Ames Research Center, Moffett Field,
California, July 1973.
3. Parks, Edwin K., Flight Test Measurement of Ground Effect for Powered Lift STOL Airplanes,
NASA TM 73, 256, Ames Research Center, Moffett Field, California, December 1977.
4. Aircraft Performance, USAF Test Pilot School, Edwards AFB, California.
5. Herrington, Russel M., et al, Flight Test Engineering Handbook, AF TR 6273, Air Force
Flight Test Center, Edwards AFB, California, revised January 1966.
6. Performance and Flying Qualities UFTAS Reference Manual, Air Force Flight Test Center,
Edwards AFB, California, October 1984.
7. Olson, Wayne M. and David Nesst, Digital Performance Simulation, Air Force Flight Test
Center, Edwards AFB, California, January 1986.
8. Anderson, John D., Introduction to Flight, Third Edition, McGrawHill, Inc., New York,
New York, 1989.
9. U.S. Standard Atmosphere, 1976, NOAAS/T 761562, National Oceanic and Atmospheric
Administration, October 1976.
10. Dunlap, Everett W. and Milton Porter, Theory of the Measurement and Standardization of
InFlight Performance of Aircraft, FTCTD711, Air Force Flight Test Center, Edwards
AFB, California, April 1971.
11. Liepmann, Hans W. and Anatol Roshko, Elements of Gasdynamics, John Wiley and Sons,
Inc., New York, New York, February 1965.
12. Nicolai, Leland M., Fundamentals of Aircraft Design, University of Dayton, Dayton, Ohio,
1975.
13. Chapra, Steven C., and Raymond P. Canale, Numerical Methods for Engineers, McGraw
Hill, Inc., 1985.
14. Olhausen, James N., “Use of a Navigation Platform for Performance Instrumentation on the
YF16,” Journal of Aircraft, Vol. 13, No. 4, April 1976.
15. Tippey, D. Kurt. 1985. “The INS Wind Calibration in Climb Algorithm.” Paper presented at
the 16
th
Annual Symposium Proceedings 1985, Society of Flight Test Engineers, Seattle,
July 29  August 2.
16. Sweeney, Tom, Performance and Flying Qualities UFTAS Link13 Users Guide, Air Force
Flight Test Center, Edwards AFB, California, February 1988.
246
BIBLIOGRAPHY (Continued)
17. Cheney, Harold. 1983. “Takeoff Performance Data Using Onboard Instrumentation.” Paper
presented at the 14
th
Annual Symposium Proceedings, Society of Flight Test Engineers,
Newport Beach, August 1519.
18. GGD Publication 92013, Geodesy and Geophysics Department, Defense Mapping Agency,
Edwards AFB, California, July 1992.
19. Pope, Alan. April 1964. Wind Tunnel Testing, Second Edition, John Wiley and Sons, Inc.
20. Roskam, Jan, Flight Dynamics of Rigid and Elastic Airplanes, Roskam Aviation and
Engineering Corporation, Lawrence, Kansas, 1976.
21. Etkin, Bernard. 1982. Dynamics of Flight, Second Edition. John Wiley and Sons.
22. Climatic Extremes for Military Equipment, MILSTD210A, U.S. Government Printing
Office, August 1957. (Out of print).
23. DeAnda, Albert, AFFTC Standard Airspeed Calibration Procedures, Air Force Flight Test
Center, Edwards AFB, California, June 1981.
24. Diehl, Walter, Engineering Aerodynamics, 1936.
25. Brown, W.G, “Measuring an Airplane’s True Speed in Flight Testing,” NACA Rep. TN
135, 1923.
26. Mair, W. Austyn, and David L. Birdsall, 1996. Aircraft Performance. Cambridge University
Press.
27. Smith, H.C. 1992. The Illustrated Guide to Aerodynamics, Second Edition. Tab Books.
28. Wood, Karl D. 1955. Technical Aerodynamics, 3
rd
Edition. McGrawHill Book Company.
29. Collinson, R.P.G. 1997. Introduction to Avionics. Chapman & Hall.
30. Jones, Robert T. 1990. Wing Theory. Princeton University Press.
31. Fox, David. 1995. “Is Your Speed True.” KITPLANES Magazine (February).
32. Dwenger, Richard, Wheeler, John and James Lackey. 1997. “Use of GPS for an Altitude
Reference Source for Air Data Testing.” Paper presented at the Society of Flight Test
Engineers Symposium.
33. Kimberlin, Ralph and Joseph Sims. 1992. “Airspeed Calibration Using GPS.” AIAA
924090. Paper presented at the 6
th
Biennial Flight Test Conference, August 2426.
34. http://www.navcen.uscg.mil/gps
35. Clark, Bill. 1994. Aviator’s Guide To GPS. TAB Books.
247
BIBLIOGRAPHY (Concluded)
36. NASA Allstar, www.allstar.fiu.edu/aero/research.htm
37. AIAA, www.aiaa.com
38. NASA Dryden, www.dfrc.nasa.gov
39. Denker, John S., See How It Fly’s, www.monmouth.com/~jsd/fly/how
40. Ojha, S.J., Flight Performance of Aircraft, AIAA Education Series, 1995.
41. Twaites, Bryan, ed. Incompressible Aerodynamics: An Account of the Steady Flow of
Incompressible Fluid past Aerofoils, Wings and Other Bodies, Dover Publications.
42. Anderson, John D. 1998. A History of Aerodynamics. Cambridge University Press.
43. Chanute, Octave. 1897. Progress in Flying Machines. The American Engineer & Railroad
Journal.
44. Lowry, John T., Performance of Light Aircraft, AIAA Education Series, 1999.
248
This page intentionally left blank.
249
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
Abbreviation Definition Unit
ADC air data computer 
AF acceleration factor 
AFB Air Force Base 
AFFTC Air Force Flight Test Center 
AGL above ground level ft
AIAA American Institute of Aeronautics and Astronautics 
AOA angle of attack deg
AOSS angle of sideslip deg
A acceleration ft/sec²
AF acceleration factor 
AR aspect ratio dimensionless
t
AR
aspect ratio of tanker dimensionless
D
A
acceleration in the down direction ft/sec
2
E
A
acceleration in the east direction ft/sec
2
N
A
acceleration in the north direction ft/sec
2
bx
A
X axis body acceleration ft/sec
2
by
A
Yaxis body acceleration ft/sec
2
bz
A
Zaxis body acceleration ft/sec
2
x
A
flight path longitudinal acceleration ft/sec
2
x
A
longitudinal acceleration
ft/sec²
y
A
flight path lateral acceleration ft/sec
2
y
A
lateral acceleration ft/sec
2
z
A
flight path normal acceleration ft/sec
2
z
A
normal acceleration (positive down) ft/sec
2
a acceleration ft/sec
2
a speed of sound kts
Note:
1. Velocity units in knots or feet per second.
2. Time in units of seconds or hours.
250
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
a temperature gradient °K/1,000 ft
a mean (average) acceleration ft/sec
2
SL
a
speed of sound standard day sea level 1116.45 ft/sec;
661.48 kts
α angle of attack deg
/ A C
α
angle of attack from the aircraft system deg
INS
α
angle of attack computed from INS data deg
BAA body axis accelerometer 
Btu British thermal unit 
BHP brake horsepower HP
b wing span ft
C Celsius deg
D
C
drag coefficient
dimensionless
min D
C
minimum drag coefficient 
L
C
lift coefficient dimensionless
Lb
C
break lift coefficient dimensionless
min L
C
lift coefficient at the minimum drag coefficient dimensionless
Lt
C
tanker lift coefficient dimensionless
fc
C
compressible skin friction drag coefficient dimensionless
fi
C
incompressible skin friction drag coefficient
dimensionless
cg
center of gravity pct MAC
cg center of gravity pct MAC
cm centimeters 
DGPS differential GPS 
D down 
D drag lbs
bw
D
drag of the aircraft body and wind lbs
251
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
s
D
standard day drag lbs
t
D
drag of the aircraft tail lbs
t
D
test day computed drag
lbs
s
D′
standard day predicted drag
lbs
t
D′
test day predicted drag
lbs
d distance ft
t
dV
change in true airspeed

t
dW
weight increment
lbs
dh change in altitude ft
dt
time increment sec
dB decibels 
deg degrees (either temperature or angle) 
E east 
EGI embedded GPS/INS 
E east 
E energy ftlbs
F Fahrenheit deg
FL flight level (ft/100)
FPA flight path accelerometer 
F Fahrenheit deg
*
F
summation parameter to be minimized 
e
F
propulsive drag
lbs
ex
F
excess thrust lbs
g
F
gross thrust lbs
n
F
net thrust lbs
nr
F
referred net thrust lbs
252
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
0 n
F
net thrust at zero speed lbs
/
n
F δ
corrected net thrust lbs
0
/
n t
F δ
referred net thrust
lbs
2
/
n t
F δ
referred (inlet) net thrust
lbs
ns
F
standard day net thrust
lbs
ns
F′
standard day predicted net thrust lbs
nslope
F
slope of thrust versus Mach lbs
nt
F
test day net thrust
lbs
nt
F′
test day predicted net thrust
lbs
r
F
ram drag lbs
rw
F
runway resistance force lbs
tsfcr
F
degradation factor for tsfcr 
1
F
nose gear load lbs
2
F
main gear load lbs
ft foot 
GPS Global Positioning System 
g
acceleration of gravity ft/sec
2
0
g
reference acceleration due to gravity 32.17405 ft/sec²
HUD headup display 
Hg mercury 
Hz Hertz cycles per second
H geopotential altitude ft
H
!
rate of change of geopotential height ft/sec
C
H
pressure altitude ft
E
H
energy altitude ft
253
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
d
H
density altitude ft
0
H
base geopotential altitude ft
h tapeline (or geometric) altitude ft
h
!
rate of change of geometric height ft/sec
AGL
h
height above ground level ft
w
h
height of wing above ground ft
ICAO International Civil Aviation Organization 
INS inertial navigation system 
In inches 
IHP indicated horsepower HP
i point number 
t
i
thrust incidence angle
deg
j iteration number 
K kelvin 
K ft thousand ft 1,000 ft
K Kelvin deg K
KE kinetic energy ftlbs
1 K parabolic coefficient of the drag polar dimensionless
2 K nonlinear coefficient of the drag polar dimensionless
kg kilogram 
km kilometers 
kt knot(s) 
LHV lower heating value Btu
L lift lbs
1
L
lift of the wing lbs
2
L
lift of the tail lbs
l characteristic length (in Reynolds number
formula)
ft
254
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
x
l
longitudinal (x) distance from cg ft
y
l
lateral (y) distance from cg ft
yy
I
moment of inertia about the ybody axis ftlbs/sec
z
l
normal (z) distance from cg ft
MAC mean aerodynamic chord 
MAX maximum rated thrust 
METO maximum except for takeoff 
MIL Military rated thrust 
M Mach number dimensionless
M moment ftlb
m mass slugs
m meter 
mbar millibar 
N north 
N/A not applicable 
NACA National Advisory Committee for Aeronautics 
NASA National Aeronautics and Space Administration 
NBIU Nose Boom Instrumentation Unit 
NTPS National Test Pilot School 
n/d nondimensional 
nam nautical air miles 
nm nautical mile 
N north 
N number of points in multiple regression 
x
N
longitudinal load factor g’s
y
N
lateral load factor g’s
z
N
normal load factor (positive up) g’s
η
propeller efficiency dimensionless
255
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
η
temperature probe recovery factor dimensionless
t
η
inlet pressure recovery factor dimensionless
P ambient (static) pressure lbs/ft
2
PE
potential energy ftlbs
SL
P
ambient pressure sea level
2,116.2166 lbs/ft²
a
P
ambient pressure lbs/ft
2
s
P
specific excess power
ft/sec
t
P
total pressure lbs/ft
2
t
P
′
total pressure behind a shock
lbs/ft²
p
roll rate deg/sec
pph pounds per hour 
q
pitch rate deg/sec
q incompressible dynamic pressure lbs/ft²
C
q
compressible dynamic pressure
lbs/ft²
R radius of a pullup ft
RMS root mean square 
R radius of turn or pullup ft
R
universal gas constant for air 3,089.8136 ft²/sec² °K
R range nam
/ R C rate of change of pressure altitude ft/sec
RF range factor nm
RN Reynolds number dimensionless
RNI Reynolds number index dimensionless
r yaw rate deg/sec
0
r
reference radius of the earth 20,855,553 ft
S south 
256
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
SFTE Society of Flight Test Engineers 
STOL short takeoff and landing 
S reference wing area ft²
SR specific range nm/lbs
SS sum of squares 
0 t
δ
referred pressure ratio
dimensionless
2 t
δ
referred inlet pressure ratio dimensionless
2 t
δ
total pressure ratio dimensionless
wet
S
wetted area ft
2
sec seconds 
TPS Test Pilot School 
T temperature °K
THP thrust horsepower HP
TSFC thrust specific fuel consumption lb/hr/lb
SL
T
sea level standard temperature 288.15 °K
a
T
ambient temperature (T = interchangeable
symbology)
°K
as
T
ambient temperature
°K
t
T
total temperature K °
0
T
base temperature ºK
t time sec
tsfc thrust specific fuel consumption lb/hr/lb
tsfcc corrected thrust specific fuel consumption dimensionless
tsfcr referred thrust specific fuel consumption lb/hr/lb
USAF United States Air Force 
cg
U
Xbody axis true airspeed kts
VSTOL vertical or short takeoff and landing 
257
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
V
!
rate of change of inertial velocity (ft/sec)/sec
C
V
calibrated airspeed
kts
D
V
down (z) inertial speed
kts
E
V
east (y) inertial (ground) speed kts
N
V
north (x) inertial speed kts
bx
V
longitudinal (xbody) axis airspeed kts
by
V
lateral (ybody) axis airspeed kts
bz
V
vertical (zbody) axis airspeed kts
cg
V
Ybody axis true airspeed kts
e
V
equivalent airspeed kts
g
V
groundspeed (usually horizontal component of
vector)
kts
g
V
"
groundspeed vector kts
t
V ∆
correction to be added to true airspeed kts
t
V
!
rate of change of true airspeed ft/sec
2
t
V true airspeed kts
tD
V true airspeed down kts
tE
V
true airspeed east kts
tN
V true airspeed north kts
t
V
"
true airspeed vector kts
ti
V
indicated true airspeed kts
v
V
vertical component of groundspeed vector kts
w
V
windspeed ft/sec
w
V
"
windspeed vector kts
258
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
wD
V
down (z) windspeed
kts
wE
V
east (y) windspeed kts
wN
V
north (x) windspeed
kts
W west 
W weight of an element of air lbs
Zf
W
zero fuel weight lbs
a
W
!
airflow lbs/sec
cg
W
Zbody axis true airspeed ft/sec
2
f
W
fuel flow
lbs/hr
( )
/
f
W δ θ ⋅
corrected fuel flow lbs/hr
fs
W
standard day fuel flow lbs/hr
fs
W′
standard day predicted fuel flow lbs/hr
ft
W′
test day predicted fuel flow lbs/hr
t
W
weight
lbs
/
t
W δ
weight over pressure ratio lbs
te
W
end gross weight lbs
ts
W
start gross weight lbs
wrt with respect to 
X independent variable 
1
XL
distance from cg to wing center of lift ft
2
XL
distance from cg to tail center of lift ft
Fn
X
distance main gear to thrust vector ft
GE
X
ground effect factor 
1
X
distance from nose gear to cg ft
259
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Continued)
Abbreviation Definition Unit
2
X
distance from main gear to cg ft
x
the x unknown =
wx
V
kts
Y dependent variable 
ˆ
Y
curve fit equation 
y
the y unknown =
wy
V
kts
1
Z
height of the body axis above ground ft
2
Z
height of the tail center of lift and drag
above body axis
ft
z the z unknown =
t
V ∆
kts
Symbol
σ
ambient density ratio dimensionless
σ standard deviation 
β sideslip angle deg
∂ partial derivative symbol 
θ pitch attitude deg
θ ambient temperature ratio dimensionless
V
θ
thrust vector angle deg
rw
θ
runway slope deg
2 t
θ
total temperature ratio dimensionless
δ ambient pressure ratio dimensionless
µ
viscosity slugs/ft sec
µ
runway coefficient of friction dimensionless
µ
coefficient of friction dimensionless
SL
µ
viscosity at sea level slugs/ft sec
ϖ angular rate of a pullup deg/sec
γ
flight path angle deg
γ
ratio of specific heats dimensionless
260
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
(Concluded)
Symbol
0
γ
gravity at sea level (function of latitude) cm/sec
2
φ bank angle deg
º degrees temperature or angle
λ engine losses factor 
ψ
heading angle (degrees from true north) deg
∆ increment 
∫
integral 
ϕ
latitude deg
φ roll attitude deg
∑
summation 
0
ε
theoretical downwash angle deg
τ thrust increase time constant sec
g
σ track angle deg from true north
261
INDEX
1976 U.S. Standard Atmosphere, 15, 16, 22,
31, 40, 174, 180
A
Accelerating or decelerating turns, 155
acceleration, 1
Accelerometer
accelerometer noise, accelerometer rate
corrections, 58, 60, 72
Aerobraking, 106, 112, 113
Airspeed, 12, 26, 30, 32, 35, 36, 37, 38, 83,
96, 100, 101, 104, 106, 111, 113, 116, 131,
134, 140, 150, 178, 246
Altitude
Constant altitude, Energy altitude, 13, 15,
17, 18, 23, 24, 25, 26, 28, 42, 55, 114,
120, 121, 134, 136, 140, 141, 166, 170,
171, 178, 201, 202, 219, 236, 246, 251
Ambient pressure, 82
Angle of attack, 67
Atmosphere, 17, 23, 40, 245
B
Braking
braking coefficient, braking forces, 3, 103,
106, 113
Butterworth filter
Fourpole Butterworth filter, 61, 63
C
Calibrated airspeed, 30, 83
Climb, 3, 144, 145, 146, 147, 149, 152, 181,
245
Cruise tests, 136
D
Deceleration, 3, 104, 154, 181
Density, 13, 26
Density altitude, 13, 26
Descent, 3, 108, 154, 181
Differential GPS, 121
Differential pressure, 33
Drag, 2, 4, 40, 41, 43, 44, 45, 46, 80, 81, 97,
98, 108, 111, 112, 113, 165, 169, 184, 185,
186, 188, 189, 190, 191, 192, 206, 207
Drag coefficient, 81
Drag due to lift, 184
Dynamic performance, 164
E
EGI, 114, 160, 179
Energy
kinetic energy, potential energy, 140
Equivalent airspeed, 37
Euler angles, 66, 73, 160
Excess thrust, 3, 57, 181, 182
F
Fuel flow, 4, 180, 182
G
Geometric altitude, 13
Geopotential altitude, 15
GPS, 2, 26, 30, 57, 58, 114, 115, 116, 122,
124, 125, 128, 129, 132, 134, 160, 218,
246, 250, 251
Gravity, 173
Groundspeed, 30, 129
I
INS, 26, 30, 58, 66, 71, 112, 114, 135, 136,
144, 146, 154, 156, 158, 160, 168, 172,
176, 218, 245, 250, 251
Instrumentation, 1, 2, 60, 245, 246, 254
L
Landing, 3, 75, 76, 103, 107, 109, 113, 245
Latitude, 174
Lift, 2, 4, 5, 40, 41, 44, 47, 82, 83, 84, 87, 94,
95, 97, 102, 108, 113, 189, 190
Lift coefficient, 82
M
Mach number, 4, 30, 32, 33, 35, 39, 41, 42,
43, 45, 47, 52, 80, 81, 111, 116, 122, 126,
129, 135, 136, 140, 141, 142, 144, 145,
148, 151, 152, 155, 156, 164, 165, 167,
168, 172, 175, 177, 178, 184, 185, 186,
187, 188, 189, 191, 192, 194, 195, 197,
200, 202, 203, 206, 209, 217, 254
262
Maximum thrust, 54
Military thrust, 208
Minimum drag coefficient, 103
N
NBIU (Nose Boom Instrumentation Unit), 59
Noise, 60
normal load factor, 152
P
Pitot tube, 33
Pressure altitude, 21
Pressure ratio, 213
Pullup, 170, 171, 172
R
Radar, 127, 134
Ram drag, 50
Range, 135, 136, 139, 140, 141, 142, 200,
201, 202, 203, 219
Range factor, 135, 140, 219
Range mission, 141
Rate corrections, 73
Refueling, 176
Reynolds number, 41, 42, 43, 80, 188, 194,
195, 199, 201, 202, 203, 213, 214, 215,
216, 217, 253, 255
Reynolds number index, 42, 194, 203, 216,
255
S
Skin friction drag coefficient, 188
SplitS, 167, 169, 170, 172
Standard atmosphere, 85
Standard day, 25
Standardization, 180, 183, 245
T
Takeoff, 3, 75, 76, 78, 86, 88, 97, 98, 99, 100,
101, 102, 113, 245, 246
Thrust, 2, 3, 6, 49, 50, 51, 53, 54, 81, 88, 92,
93, 102, 140, 145, 148, 193, 194, 195, 196,
197, 198, 204, 205, 206, 207, 208, 210,
211, 215, 218
Thrust runs, 81
Thrust specific fuel consumption, 193
Total pressure, 1
Total temperature, 1
True airspeed, 1, 30, 32, 125, 178
Turns, 155, 156
W
Weather, 117, 237, 239, 241, 243
Windspeed, 25, 30, 236, 240
AIRCRAFT PERFORMANCE FLIGHT TESTING
CHANGE FORM
Date:
To: Frank Brown, 412 TW/TSFT
Guide Revision No.: Page No.:
Reads As:
To Read:
Reason for Recommended Change:
Organization:
Name and Grade/Organization:
Signature:
FROM:
Frank Brown
412 TW/TSFT
195 E. Popson Ave.
Edwards AFB,
California 935246841
REPORT DOCUMENTATION PAGE
Form Approved OMB No. 07040188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (07040188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 222024302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DDMMYYYY)
2. REPORT TYPE
3. DATES COVERED (From  To)
September 2000
4. TITLE AND SUBTITLE
Final
N/A
5a. CONTRACT NUMBER
CRDA #99171FT0
5b. GRANT NUMBER
Aircraft Performance Flight Testing
6. AUTHOR(S)
5c. PROGRAM ELEMENT NUMBER 5d. PROJECT NUMBER 5e. TASK NUMBER
Olson, Wayne M., Aircraft Performance Engineer
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
5f. WORK UNIT NUMBER 8. PERFORMING ORGANIZATION REPORT NUMBER
Air Force Flight Test Center 412 TW/TSFT 195 E. Popson Avenue Edwards AFB, California 935246841
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES)
AFFTCTIH9901
10. SPONSOR/MONITOR’S ACRONYM(S)
AFFTC
11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION / AVAILABILITY STATEMENT
Approved for public release; distribution unlimited.
13. SUPPLEMENTARY NOTES
14. ABSTRACT
This document is intended as a reference source on the topic of aircraft performance flight testing. Formulas are derived for equations of motion, altitude and airspeed. It covers the various performance maneuvers, including takeoff, landing, cruise, acceleration, climb, and turn. Specialized tests to calibrate air data systems and to dynamically determine aircraft lft and drag are discussed. Lift, drag, thrust, and fuel flow i analysis methods are presented. Special topics include gravity models, aerial refueling, terrain following, and effects of temperature and wind. The text is primarily for conventional jet aircraft, however, many of the equations and methods are applicable to light civil aircraft.
15. SUBJECT TERMS
aircraft performance acceleration thrust
a. REPORT
models simulation drag turning flight fuel flow jet aircraft
c. THIS PAGE
air data GPS calibration
takeoff landing INS lift atmospheric effects
18. NUMBER OF PAGES
cruise performance climb performance
19a. NAME OF RESPONSIBLE PERSON 19b. TELEPHONE NUMBER
(include area code)
16. SECURITY CLASSIFICATION OF: b. ABSTRACT
17. LIMITATION OF ABSTRACT
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED
UNLIMITED
284
Standard Form 298 (Rev. 898)
Prescribed by ANSI Std. Z39.18
PREFACE
The author was employed at the Air Force Flight Test Center (AFFTC), Edwards AFB, California, from 1968 through 1993 as an aircraft performance flight test engineer. This document began, but was not finished, prior to his retirement in 1993. He endeavored to complete the document on his own and this text is the final result of that. He received a lot of help from the reviewers, which he mentions belowthey each made suggestions that improved the text vastly. The intent of this text is that it should provide a highly useful reference source for aircraft performance flight test engineers. It certainly should not be the only source of information. The bibliography contains just a few of the sources that the author has found most useful. Much of the material covered in this handbook can be found in slightly different forms in the bibliographies listed in the Bibliography section. Even though the Flight Test Engineering Handbook (listed in the Bibliography Section) was originally written in the 1950s and updated slightly in the 1960s, it still contains much useful information. The author utilized Everett Dunlap’s Theory of the Measurement and Standardization of InFlight Performance of Aircraft extensively as a reference source during his years at Edwards AFB. Also, the USAF Test Pilot School’s (TPS) Aircraft Performance manual was a valuable source, as well as the knowledge the author gained while a student at the USAF TPS. The emphasis here is on performance testing as conducted at Edwards AFB; therefore, low budget or light aircraft testing is not covered extensively. Very little is said about instrumentation, except that it is needed and should be as accurate as reasonably possible. The thrust discussion is kept to a minimum. A number of other possible topics are discussed lightly, if not at all. Items not necessarily complete are: 1. airspeed calibration in ground effect, 2. test planning, 3. test conduct, 4. how to fly the maneuvers, 5. use of parameter identification, 6. report writing, and 7. cg accelerometer system. This handbook is pieced together from writing the author has done going back as far as 1975. Much of it is from individual performance office memos which were written to standalone; therefore, you will see quite a bit of duplication. The same equation appears in several placesthe author tried to have the major derivation of the equation appear only once. For those of you who are familiar with the author’s style, you know he is big on theory and equations. Although it appears that there are a lot of intermediate steps in the derivations, the extra steps are appropriate to show where all the constants come from.
iii
Mac McElroy. which is discussed in this handbook. Mr. They are: 1. Milton Porter for teaching the author the mathematics that he applied to the cloverleaf method in a mathematics class at the USAF TPS. In addition. John Hicks from NASA. Frank Brown. Al DeAnda for teaching the author about calibrating airspeed. Virginia iv . for all his help in the preparation of this handbook. 2. 7. Mr. Mr. they deserve mentioning. pointed out that the AFFTC and NASA were using dynamic performance methods on the lifting body research projects years before those of us in the conventional aircraft business.The author worked with Mr. Messrs. Although the list is long. Mr. Finally. Dave Richardson. Mr. Lee for a short period of time in the early 1970s studying parameter identification. Mr. Dryden Flight Research Center. Mac McElroy. Alan Lawless of the National TPS and Mr. while reviewing a very early version of this text. Allen invented the “cloverleaf” airspeed calibration method. Jim Cooper. Allan Webb also reviewed the thrust section. the author would like to give sincere thanks to Mr. Pete Adolph. Clen Hendrickson. Charlie Johnson. In addition. Mr. Mr.Early versions of this text had three primary reviewers: Messrs. Randy Simpson of the Naval Air Test Center (now called the Naval Air Weapons Center). Mr. The author worked several months with Mr. 9. 10. Hart and Brown reviewed both the draft and final versions of this handbook. Ken Rawlings. provided significant comments that were implemented into the text. who in the middle 1970s taught the author about using inertial n avigation systems (INSs) for performance. Mr. Mr. As a result of Mr. Ron Hart. Lyle Schofield. Bob Lee . Jim Olhausen of General Dynamics on the YF16 and F16A. and Frank Brown. Mr. There were many individual engineers at Edwards AFB that the author would like to acknowledge in this handbook. 4. Frank Brown and many others for helping the author learn about test techniques and other aspects of flight test. Richard Colgren of LockheedMartin Skunk Works and Captain Timothy Jorris of the AFFTC provided excellent suggestions that were incorporated. Bill Fish for tutoring the author in propulsion (though propulsion is discussed lightly in this handbook). Bill Fish suggested adding the discussion of the ratio method of standardization and reviewed the thrust section. Willie Allen for teaching the author almost everything he knows about dynamic performance and flight path accelerometers. the INS became the primary source of flight path acceleration data on almost every large project at the AFFTC. 3. 8. McElroy looked at some early versions of this handbook. Messrs. Don Johnson. Jim Pape (who never found out the author did not know the difference between an aileron and an elevator when he first started working at Edwards AFB). Mr. Ms. 5. Mr. 6. Ron Hart. his successor at Edwards AFB. Mr. Simpson on developing dynamic performance methods in the early 1970s. Olhausen’s work.
com v . Frank Brown 412 TW/TSFT 195 E. The AFFTC would appreciate any suggestions for additional material. CA 935246841 Frank. contact either Frank Brown or the author via e mail with any comments. AFB.Brown@edwards. WA 986072340 Wayneoperf@home.mil Wayne Olson 3003 NE 3rd Ave. or any technical errors you may find. #222 Camas. This will not be the final version of this handbook. or if needed.O’Brien of Computer Sciences Corporation for the technical editing and final format o this f handbook. Popson Ave Edwards. clarification of existing material. A form to submit proposed changes and/or improvements is included in the back of this handbook.af. Following are addresses and email for each of them.
vi .This page intentionally left blank.
..................................................................7 Geopotential Altitude (H) versus Geometric Altitude (h) ..........................................1 Flight Path Axis..35 4............................................................3 Geopotential Altitude ..............38 5.................................................23 3......................................40 vii .6...32 4.....TABLE OF CONTENTS Page No....................................................................................................... 5 2.......................4 1976 U....................................2 Body Axis................................................................................................................................................................................................24 3.........................................19 3.........................6 Mach Number..............................................................10 SECTION 2.............2 Case 2: Linearly Varying Temperature.............1 Introduction – Altitude ............................................................................8 Calibrated Airspeed.................30 4..............................32 4................18 3..............................................5 Temperature and Pressure Ratio ..0 ALTITUDE.............................. 1....................................................................................................................4 The Nautical Mile .........5 Data Analysis ...........................16 3...............Nonstandard Day..0 AIRSPEED ................................................. 1.......................xii LIST OF TABLES ..............................................................................9 Effect of Wind Gradient.........4 InFlight Forces.........................iii LIST OF ILLUSTRATIONS.......10 Mach Number from True Airspeed and Total Temperature..............................xvii 1.......................16 3.......................................7 Total and Ambient Temperature.......................1 Introduction.37 4..8 Geopotential versus Pressure Altitude .........................................................................................11 Airspeed Error Due to Error in Total Pressure..........................2 Primary Instrumentation Parameters...............................................................................................................................1 Case 1: Constant Temperature ......................................................0 OVERVIEW .................................................................0 LIFT AND DRAG......... 1...........................................................................................................31 4...................................28 4..........2 Speed of Sound ....................................................................................... 5 2....................6 Pressure Altitude....................9 Equivalent Airspeed.............................................................................................................................................30 4............ Standard Atmosphere ..............................................25 3.............18 3......................10 Density Altitude.............3 Ground Tests .............................................................0 AXIS SYSTEMS AND EQUATIONS OF MOTION..............................................37 4................................................. 7 2.........5 True Airspeed...................13 3........................................ 1................0 REFERENCE ............3 True AOA and Sideslip Definitions..............................................................................................................S...............3 History of the Measurement of the Speed of Sound ...................................................................11 Pressure Altitude Error Due to Ambient Pressure Measurement Error....................................................................................2 Hydrostatic Equation .........................................................13 3................................35 4.......................................................................................30 4............................................................. PREFACE ............................ 1..1 Introduction.................................................................................. 8 2..............13 3....................................................................................26 3............................................................12 3............. 1 1 1 2 3 3 2..................................6........................................................4 Flight Maneuvers.....................................................................................40 5........................................15 3..........................1 Introduction – Airspeed ...............................32 4.........................
................................ 7....7 Air Force Flight Test Center Drag Model Formulation ............................ 6.................... 7................... 7..................................................................................................................................................... 6..... 7................................................8 Flight Path Accelerations ........... 75 8................. 5........................... 7............................................................................................................................................................................................................................... 78 8.......0 REFERENCES ..............12 Euler Angle Diagram......... 88 8..................................................................................... 7................5 Accelerometer Noise......................................................................................................................... 92 8......................................................................................................................................................................... 6.. 6....................8 Effect of Thrust Component...........................3 Developing a Takeoff Simulation.......................................2 GPS Method ...................................3 The Drag Polar and Lift Curve..... 75 8................ 6.................................................................2 Takeoff Parameters ......1 The Reciprocating Engine at Altitude .............................0 FLIGHT PATH ACCELERATIONS ........ 6......... 75 8. 7................... 40 41 42 43 44 45 45 49 49 50 51 51 51 53 54 54 55 57 57 58 58 58 60 66 66 71 72 73 73 73 SECTION 5.....................7 Calculating Alpha................................................................6 Inertial Measurement Method ......................4 Reynolds Number.........................9 EngineInoperative Takeoff............................................6 Effect of Wind on Takeoff Distance...........................................0 TAKEOFF ...................................................1 General............................................................................ 48 8............................................2 The Thrust Equation ....................................................................TABLE OF CONTENTS (Continued) Page No................8 Propeller Thrust ......................................8 The Terminology ‘Drag Polar’ .........................................................................................................................................................4 Status Deck........................................... 5.......... 7.........................3 InFlight Thrust Deck ................11 Calculating p...10 Idle Thrust Decelerations ...................................................... 5........ 7........................................................ 7................................................................................................................................... Beta and True Airspeed.............. 5.............................. 6................................... 5....................................................................................4 Flight Path Accelerometer Method ......................................................... q.................. 6...................................7 Takeoff Using Vectored Thrust.................................................................. 7.....................................................................................................................................9 Accelerometer Rate Corrections ................ 87 8.........................................................................................102 viii ........ 88 8........................ and r.....6 Thrust Runs ..........................8....................... 6...................10 Velocity Rate Corrections ............... 98 8.............................1 Introduction...................................................................... 80 8..........0 THRUST..............................................................................................................5 Effect of Runway Slope.................5 Inlet Recovery Factor .................................................................... 6.............................. 7..6 Idealized Drag Due to Lift Theories..........2 Definition of Lift and Drag Coefficient Relationships ................. 7.....................................................3 Accelerometer Methods ....................... 5.................1 AirspeedAltitude Method..................................................................5 Skin Friction Drag Relationships ..... 5....................................................................................................7 Thrust Dynamics.........4 Ground Effect...............
.....................................................................103 9...........................................139 11..3 Sawtooth Climbs.........................................0 LANDING.........................................................................4 Continuous Climbs .........................................115 10.........................................................115 10....................................................152 12................4 Landing on an Aircraft Carrier ..............3 Range .............................7 Slow AccelDecel.................4 Computing Range from Range Factor .................142 11........................................107 9..141 11.................................136 11...................0 ACCELERATION AND CLIMB........6 Range Mission.........2 Accelerating or Decelerating Turns .....................................................10 Air Force Flight Test Center Data Set..........................................................................................................................................144 12..........149 12...........................................8 Descent ......................132 11.......................................................................1 Introduction........8 Effect of Wind on Range ...........155 13................................................146 12...................................................................................................121 10........4 Pacer Aircraft......................................................................................0 CRUISE ......................................................9 Decele ration............................................145 12...........115 10...............................................................................3 General Concepts .................................Introduction ........................125 10.........124 10.......................................................................................................135 11.............................................................................................................155 ix .............. 9..........................................................................1 Acceleration......................................................136 11...........................................................................................................6 Acceleration Factor (AF).....9 Error Analysis ....................150 12..........................................135 11......................................2 Aerobraking................................................................................149 12....................154 12....................................................................................................1 Two Numerical Examples for AF ..112 9.....................................5 Constant Altitude Method of Cruise Testing ..............................2 Climb.....................................................................106 9.................................................................3 Landing Air Phase..............................................116 10..........................................8 The Flight Maneuver............126 10...............TABLE OF CONTENTS (Continued) Page No....6 AccelDecel...141 11............1 Introduction.............................148 12...............................................................0 REFERENCES .................................2 Groundspeed Course Method ........113 10...................11 Mathematics of the Cloverleaf Method......................................................................................................................................................................................................................................154 SECTION 12.........................................................................5 Climb Parameters.........................................6 Takeoff and Landing Measurement .............................155 13......5 Tower Flyby.......................6............7 The Cloverleaf Method ...................................................................119 10............................................................1 Historical Perspective ............................144 12.................................103 9.................................................................0 TURNING...............126 10..........1 Braking Performance ............................119 10.............................................................5 Stopping Distance Comparison .......................................................2 Cruise Tests............................142 12.......................................................................................................................109 9...................0 AIR DATA SYSTEM CALIBRATION .............................................................154 13..........................................7 Normal Load Factor During A Climb ..................................................................................................................
..............1 Normal Load Factor........7 Military Thrust.................................................................................................................................................1 Introduction..............................................................6 Thrust Specific Fuel Consumption...........................2........................................2..........................................................................6 Angle of Attack..........................................................................0 A SAMPLE PERFORMANCE MODEL...7 Turn Rate ..............................2 Acceleration/Deceleration ........3 ThrustLimited Turns .............................173 15.......184 17...................................................................2..........................................................................................................198 17....193 17.......0 STANDARDIZATION..............................178 15........................................................................................182 16.........................................................4 Cruise ...........170 14....................................................................................160 14..............................2 Increment Method .................3 Performance Degradation during Terrain Following ............................................................................................................................180 16...7 Vertical Wind........................................184 17...........................................................156 13.............................................155 13............................2 Drag Model..1 Introduction......................................................1 Minimum Drag Coefficient .................173 15..............................2...........................................164 14........4 Uncertainty in Performance Measurements............................................................9 Cruise....195 17......5 LiftLimited Turns ..........................................188 17....157 13.........3 Accelerating/Decelerating Turn..................................176 15................................................................................................................................................3 Windup Turn.........................182 17...................................................10 Range ............5 Thrust and Fuel Flow Model.182 16..........................0 DYNAMIC PERFORMANCE....................................................................................................................179 16...3 Skin Friction Drag Coefficient ..................................................................................................................2 Turn Radius.......4 SplitS.................................5 Pullup ........................................................................................................................................................................................2 Performance Degradation during Aerial Refueling ...................................................2...........................................................184 17.178 15.......2...........172 15....1 Climb/Descent....................................................................8 Winds Aloft...........180 16..............200 17................................189 17..................................5 ThrustLimited Turn...........................6.......................181 16.................167 14.........................................................11 Endurance..........164 14...................................................................................................184 17..............193 17.....................................1 Effect of Gravity on Performance..............4 Stabilized Turns .............................180 16....................159 13.................................172 14......................156 13....................................TABLE OF CONTENTS (Continued) Page No................ 13....1 Introduction.......6 Turn Equations ............................................................................................................8 Maximum Thrust .....................................2 Roller Coaster ..............................................................................................................................6 Wind Direction Definition........................................................197 17...........................................181 16............................................................................................................................................................159 13...................3 Ratio Method ................................................177 15...............................0 SPECIAL PERFORMANCE TOPICS ..................5 Sample Uncertainty Analysis ...................157 13...................6.........164 14..........................................................................................................................................................................167 14..............................................4 Drag Due to Lift.....................182 16......................203 x ......................................
.............................................................AVERAGE SURFACE WEATHER FOR THE AIR FORCE FLIGHT TEST CENTER................................................................220 19. ACRONYMS..............................................12 Acceleration Performance..WEATHER TIME HISTORIES.......TABLE OF CONTENTS (Concluded) Page No..................................................................................................................................204 17.....................................249 INDEX.........................13 Military Thrust Acceleration .........................261 AIRCRAFT PERFORMANCE FLIGHT TESTING CHANGE FORM xi .......213 18................................14 Maximum Thrust Acceleration...........................2 Constants........................................15 Sustained Turn............................. AND SYMBOLS ............................................................216 SECTION 18........231 APPENDIX B ................................245 LIST OF ABBREVIATIONS....1 Equations ...............................241 BIBLIOGRAPHY.......................................................................... 17.0 CRUISE FUEL FLOW MODELING...................................................207 17........................................................................................................................................215 18............................................................................1 Thrust Specific Fuel Consumption.203 17.............237 APPENDIX C ..........................................0 EQUATIONS AND CONSTANTS.....................................................................210 18..........................................................................................219 19.....................0 REFERENCE ..............AVERAGE WINDS AND TEMPERATURES FOR THE AIR FORCE FLIGHT TEST CENTER.....................................2 Multiple Regression .....229 APPENDIX A ..................................................220 19................................................
...............3 2..4 8.........53 Thrust Dynamics from an Air Force Flight Test Center Thrust Stand ........5 Title Page No............10 Axis System Angle Diagram.................................5 7.80 Lift Ratio InGround Effect .....................36 True Airspeed Error for 0...46 LifttoDrag Ratio versus Lift Coefficient........83 Takeoff Forces...............38 Ratio of Compressible to Incompressible Dynamic Pressure ......................... 7 Angle of Attack and Sideslip Definitions .....................................................................................6 7........4 7.............63 FourPole Butterworth Filter Group Time Delay..... Hg Error....................7 7............2 6...............65 ThirdOrder Polynomial Fit of Filtered Longitudinal Load Factor Data.................................................................................11 Element of Air ..3 4............................54 Air Force Flight Test Center Nose Boom Instrumentation Unit ........1 3...................................................2 3...............8 8........................................................................1 7.4 6...................................................2 5...............................................................................................................44 Drag Polar ..................................................1 4...1 2........................................................3 7...........001 in..............2 8................................ Aircraft Axis System .......76 Predicted Ground Effect Drag..............................86 xii ......................3 6....................................1 6...23 True Airspeed versus Calibrated Airspeed.......................................14 Logarithmic Variation of Pressure Ratio .......60 Longitudinal Load Factor – Unfiltered Data ............................................1 8......................22 Standard Atmosphere Temperature................64 Longitudinal Load Factor – Filtered Data.........3 5......61 Normal Load Factor – Unfiltered Data............62 FourPole Butterworth Filter Attenuation Characteristics........65 Euler Angles............................................ 8 InFlight Forces .........86 Takeoff Parameters...................47 Turbine Engine Schematic .2 7......................4 3..41 Skin Friction Drag Relationships.................3 8...................................................................2 2......................LIST OF ILLUSTRATIONS Figure No.................52 F15 Inlet Schematic .....2 5...................................................74 Takeoff and Landing Forces and Angles .......1 5................................................49 Normal Shock Recovery Factor .................................4 7.................... 2...................
...............................110 Groundspeed Course – Heading Method ....109 F/A18 with Tailhook Extended..........108 Landing Air Phase .......20 9..................10 8......................... Effect of Wind........................................9 9...10 10...................115 Groundspeed Method – Direction Method ..........................................................8 9.......... 93 Delta Tail Lift for Tail Area = 60 ft2 ..............2 9..................................12 8...............................................................5 9........S.........LIST OF ILLUSTRATIONS (Continued) Figure No............................................................1 9...............103 Stopping Distance versus Mu ( µ ) ............120 Altitude versus Grid Reading for Flyby Tower .................................107 Final Descent Rate versus Initial Descent Rate.......................3 10.................9 8............................13 8...............110 The U.........11 8......105 Braking Forces versus Calibrated Airspeed .. Nimitz...........................................19 8.....................................................116 Flyby Tower Grid.................................16 8............ 97 Takeoff Drag Model.............................................................101 Braking Forces ................................................................ 98 Takeoff Parameters versus Time ......104 Mu versus Groundspeed (Wet Runway) ....................14 8.... 91 Effect of Thrust Component on LiftOff Speed.......................18 8............... 99 Takeoff Forces versus Airspeed.....................S... 8.............................................7 8...................7 9.........................104 Deceleration versus Calibrated Airspeed..................15 8....17 8. 96 Takeoff Lift Model...............4 9........ 91 Angle of Attack at LiftOff .................................. 94 Delta Tail Lift for Tail Area = 80 ft2 ................... 88 F16 Dimensions .....................................................................................................................4 Title Page No...........................................................3 9.106 Total Resistance Force Comparison ....8 8... 95 Distance to LiftOff versus Airspeed .................6 9................100 Takeoff Forces versus Airspeed: Engine Inoperative ..................6 8.......................... 96 Calibrated Airspeed at LiftOff............................... 89 Distance to LiftOff ...2 10..............................1 10...........................................................120 xiii .............................. 92 Effect of Thrust Component on Distance to LiftOff ..........................................................
...........12 10.........................130 Groundspeed – Run 1b..............................126 Air Force Flight Test Center F15 Pacer ...................................................8 14.....................171 xiv .165 Roller Coaster Normal Load Factor ................................................................ Effect of 10Foot Error in Flyby Tower Altitude...................................................................................................170 SplitS Altitude Time History ...10 10.............................129 Groundspeed – Run 1a ................................157 Banked Turn Diagram.............................................7 13......................11 10.......................................121 Pressure Survey ......126 Position Error .14 10.....4 12......................................1 14.2 14...................................130 Groundspeed – Run 1c ................167 SplitS Drag Model.....................................5 12...................................169 SplitS Normal Load Factor ...........6 10...............123 AccelDecel Position Error Coefficient ..........153 Normal Load Factor Vectors In a Turn........166 Roller Coaster Altitude Time History...............9 10.............147 AC119G Sawtooth Climb Data ......................9 Title Page No...............170 Pullup Mach Number Time History......................................6 14....................................7 14...............................131 Specific Excess Power from Acceleration..........2 14.......................................................................1 12....3 14..........5 14......131 True Airspeed ....................................................................................147 AC119G Excess Thrust Data ......................................................5 10..2 12..................................................15 12..............................13 10.......................................................................150 Acceleration Factor – Constant Mach Number....166 Roller Coaster Mach Number Time History...........124 Cloverleaf Flight Maneuver ...........158 Drag Model.....................169 SplitS Mach Number Time History ...4 14..........................8 10............................. 10........6 12........145 AC119G Aircraft.................................123 AccelDecel Delta H ...................................152 Centripetal Acceleration Diagram......LIST OF ILLUSTRATIONS (Continued) Figure No.............................................................148 Acceleration Factor – Constant Calibrated Airspeed.......................................................................................3 12..................................7 10...............................................1 13........
........210 Maximum Thrust – Sustained Turn Normal Load Factor .............11 17.....7 17.............................Endurance ..................22 17....................185 Transonic Drag Increment......................................188 Drag Due to Lift Slope ......................................200 Maximum Range Factor ..................................................................................000 Feet...171 Subsonic Drag Increment .........000 Feet.............................197 Maximum Thrust .........000 Feet – Temperature Variation....18 17...................................205 Military Thrust – Specific Excess Power.........12 17..............202 Fuel Flow ................................................................198 Range Factor........20 17.............188 Skin Friction Drag Coefficient.14 17..............................9 17........207 Maximum Thrust Specific Excess Power ......192 Thrust Specific Fuel Consumption .....................5 17..........186 Summary of Delta Drag Coefficient............194 Military Referred Net Thrust .........................................................................................................27 Title Page No.......26 17......19 17......................................... Pullup Altitude Time History........................................196 Referred Net Thrust for Maximum Thrust........................191 Subsonic Drag Model.............................................................211 xv .............................25 17.......................................... Temperature Effect..2 17............................................................6 17...................185 Supersonic Drag Increment .8 17................3 17........21 17...............................................208 Maximum Thrust Specific Excess Power Temperature Effect at 30.............190 Drag Model at 0........................................................16 17...........................LIST OF ILLUSTRATIONS (Continued) Figure No.192 Drag Model – All Mach Numbers.......24 17....................17 17....................208 Acceleration Time – Variation with Thrust .........................................................8 Mach Number ..........10 17.4 17...............10 17..201 Range Factor – Altitude Effect............................203 Military Thrust Specific Excess Power.............15 17.................................13 17................. 14.....1 17..206 Drag at 10................................................................................................23 17......205 Military Thrust – Thrust and Drag at 10..........................196 Military Thrust........................201 Range Factor – Variation with Temperature................................
........233 Delta Temperature at 20................................................................................................000 Feet..............................234 Delta Temperature at 50..................236 Geometric Height minus Pressure Altitude.......................243 xvi .............................................235 Wind Direction ....233 Delta Temperature at 30......................................................218 Range Factor Variation with Altitude ................240 Average Maximum and Minimum Surface Temperatures .............................................................................000 Feet......................................................234 Delta Temperature at 40.......000 Feet.................................LIST OF ILLUSTRATIONS (Concluded) Figure No................3 18....................................219 Delta Temperature at 10...............................................215 Percentage Error in Thrust Specific Fuel Consumption .....000 Feet.........4 A1 A2 A3 A4 A5 A6 A7 A8 B1 B2 B3 C1 Title Page No........1 18....... 18.................................................................................000 Feet..........................................240 Windspeed Time History...236 Delta Temperature Time History .2 18............................... C17A Aircraft.............................................................235 Windspeed .....................................213 Thrust Specific Fuel Consumption .................239 Wind Direction Time History.....
................................................................. 17 Standard Atmosphere Pressure and Temperature .................................3 15.............1 7....................5 9...............................1 15.............2 17..............2 10................................1 3.........................4 Title Page No....................1 17..... 17 Edwards Average Weather Data for January.........128 Outputs .............................................136 Range Factor Versus Altitude for B52G..4 15......................1 8.......... 3..................140 Climb Ceiling Definitions ...178 Tabulated Drag Rise Data ....... 29 Reynolds Number Variation with Mach Number and Altitude ...................................................................113 Aircraft Average Measurements and Parameters ....... 97 Takeoff Parameters at Flight Events ........................... 26 Pressure Error Versus Altitude Error ....202 Range Factor Variation with Temperature...............................................................................113 Integration of Braking Results.........3 3.........................................................1 9............3 8.S.....................175 Effect of Heading on Drag Coefficient...............................LIST OF TABLES Table No.. 87 Forces at LiftOff Speed...........5 17........... 66 Takeoff Events..................................101 Ground Effect Parameters for F/A18 Carrier Landing.......................4 10.....2 3.203 Drag Variation with Temperature .... Wet....... 1976 U..............................................175 Effect of Heading and Speed on Normal Load Factor ..........112 Dry..............207 xvii ............3 9...2 15..........1 10.4 3.........3 11......................................................172 Effect of Latitude on Gravity at Sea Level... 42 Summary of Statistics for Longitudinal Load Factor.....100 Takeoff Parameters at Significant EventsEngineInoperative ...........................129 B52G Cruise Data .............................128 Inertial Speeds (GPS)....... 25 Energy Altitude Effect of Wind Gradient ........................1 15............2 8....................................174 Effect of Altitude on Gravity...............................187 Range Factor Variation with Altitude ............................... and Aerobraking Data Summary ...................................... 87 Effect of Runway Slope ...............................1 8.....146 Pullup and SplitS Initial and End Conditions..............2 12..........................111 Change in True Airspeed During Landing Due to Ground Effect...........................................................3 17...........1 11...........5 5.....4 8...2 9..................................................................................1 14......................176 Parameter Uncertainties......... Standard Atmosphere............................................
descent. On most aircraft programs. acceleration. We will compute the longitudinal acceleration of the aircraft parallel to the flight path. and static thrust runs. For flight test aircraft. Simulation and curve fitting may be utilized during the data analysis process. Specialized maneuvers called dynamic maneuvers are used to efficiently collect aircraft lift and drag data.1. We will make a list as follows: Total pressure. From total and ambient pressure then we compute the indicated airspeed. Indicated airspeed is a measure of the differential pressure. A temperature probe is used to measure the total temperature of the air. there is the collection of data during all phases of flight. Aircraft airspeed. Ambient (or static) pressure. deceleration. accelerating or decelerating turns. This is subject to errors called position errors. but a calculation usually based upon a set of fuel tank quantity measurements in flight. 1. The terminology is due to the fact that there is some ‘position’ on the surface of the aircraft where the ambient pressure error is zero or minimal. During flight. and landing. The aircraft will also be weighed at various times during the program to verify the calculations. Then. The phases of flight include takeoff. True airspeed is the physical speed of the aircraft with respect to the moving air mass. It involves ground tests such as calibrating instruments. and attitude. weighing the aircraft. cruise. we use inertial navigation system (INS) data to compute the 1 .2 Primary Instrumentation Parameters In a performance evaluation. Aircraft gross weight. This is not a single measurement. this is often from a nose boom. ambient pressure and total temperature we can calculate the true airspeed of the aircraft. The flight path is determined by the true airspeed vector. we will apply corrections to ambient pressure to be able to go from indicated airspeed to the corrected values for calibrated and true airspeed. Differential pressure is simply total pressure minus ambient pressure. From measured total pressure. The bad news is that for any given static source location. Total temperature. acceleration to climb speed. A measurement of the total pressure (in typical units of pounds per square foot) experienced by the aircraft. The fuel tank quantity weights are simply added to a known empty weight of the aircraft. and temperature measurement systems will be calibrated in flight. The empty weight will be computed for each flight based upon the particular configuration for that flight. Since we have position error in the ambient pressure. there can be hundreds of instrumentation measurements. Longitudinal flight path acceleration. thrust. only a few can be considered primary. and fuel flow.1 Introduction Aircraft performance flight testing is different things to different people. altitude. All data collected will be reduced to enable analysis of specific maneuvers such as cruise and to verify and update aircraft mathematical models for lift. However. the aircraft will also maneuver in sustained. the position error varies with speed. An attempt to measure the atmospheric ambient pressure (in same units as total pressure). altitude. drag. Taxi tests are performed prior to first takeoff. climb.0 OVERVIEW 1.
The airspeedaltitude method or GPS are also used. If there is one fundamental equation of aircraft performance. Normal acceleration: The acceleration perpendicular to the flight path is the normal acceleration. However. 1. then we get a second fundamental formula. By dividing longitudinal acceleration by the acceleration of gravity. Then. The principle of a thrust stand is quite simple. The aircraft should be weighed with zero fuel and with various amounts of fuel to check the numbers provided by the contractor. Taxi means simply to move the aircraft under its own power on the ground without achieving flight. The installation and calibration of all aircraft instruments should occur prior to flight. Thrust is computed from a variety of measured engine and atmospheric parameters. Divide normal acceleration by gravity to obtain normal load factor. The combustion process dramatically increases the temperature of the air and the air (plus the fuel) exits the tail pipe at a much higher velocity. This change in momentum and a pressure difference between the inlet and exit are the primary factors that produce thrust. while the first one is exact.3 Ground Tests Instrumentation calibration.longitudinal acceleration. The output of the total and ambient pressure probes can be groundtested using precision pressure monitors. Taxi tests. Much of the instrumentation can be checked after it is installed in the aircraft. The installed thrust of the engines can be measured directly on the ground on a static thrust stand. it would be the following: Drag = Net Thrust – Excess Thrust where: Drag = the net aerodynamic resistance parallel to the velocity vector. In this handbook. a comparison of thrust at zero speed over a range of power settings can be made with predictions. Static thrust. multiply the longitudinal load factor by the gross weight to obtain the excess thrust. The center of gravity (cg) can be determined in a weight facility where separate scales are available for the main and nose gear. we will discuss only turbine engines. While taxiing on the ground. Lift is the net aerodynamic force perpendicular to the velocity vector. keep in mind this one is only approximately correct. we get the longitudinal load factor. Thrust is produced during the process of air accelerating through the engine. the aircraft is tested. Aircraft weight and cg. However. Lift = (Normal Load Factor) x Weight Thrust. most of the equations of motion in this handbook are applicable to aircraft with other types of propulsion. If we ignore the small component of thrust perpendicular to the velocity vector. The first taxi tests 2 . The aircraft sits on a pad and is connected by cables to a load cell that measures load (thrust) directly in pounds of force. By operating the engine at various throttle settings. The propulsive force provided by the engine. The air entering the inlet is nearly brought to a stop and then accelerated through various turbine stages.
These tests are used in conjunction with climb tests to determine the optimum climb profiles. while in civilian testing. the obstacle clearance height is 50 feet. and MAX is the Maximum throttle setting with the use of afterburner. MIL (military) and MAX (maximum). this data can be used to assist in generating aircraft thrust/fuel flow and drag models. Idle is the minimum throttle setting. will quantify idle thrust at low speeds. the height is 35 feet for heavy aircraft and 50 feet for light aircraft. 3 . and temperature). The final test. will evaluate stopping performance as well as the ability of the brakes to withstand the high temperatures associated with maximum performance braking. altitude. In addition. Landing tests are used to measure the distance to land starting from clearing an obstacle (as in the takeoff test). Turning performance is conducted to both determine ability of the aircraft to turn and to assist in generating aircraft lift and drag models at higher lift and angleofattack values than what are obtainable in 1g flight. 1. with modern analysis methods. prior to first takeoff. and fuel used to climb to a cruise altitude. Engine thrust is evaluated at fixed throttle settings. Cruise testing is conducted to evaluate aircraft range. In addition. Braking tests performed during the landings or as separate tests. these settings are usually designated IDLE. distance. Acceleration tests are conducted during level 1g flight at fixed throttle settings. Liftoff is usually defined as when lift first becomes greater than weight. will be to rotate the aircraft to liftoff attitude. Thrust at these fixed throttle positions is primarily a function of flight conditions (speed. which are verified and updated using cruise and other data. MIL is the maximum throttle setting without the use of afterburner. the optimum range conditions are usually determined through analysis of drag and thrust/fuel flow models. Climb tests are flown to determine time. For military aircraft. combined with the static thrust data. For multiengine aircraft. engineout testing is also performed wherein one engine’s power is reduced to idle to simulate an engine failure during takeoff. 1. In USAF testing. Deceleration and descent tests are conducted to determine ability of the aircraft to decelerate and the fuel used in descent maneuvers.4 Flight Maneuvers Takeoff tests are performed to determine the distance required to liftoff and to clear an obstacle. rate of climb versus altitude is determined. The aircraft is flown in stabilized flight over a range of speed and altitude conditions in order to determine the best speed and altitude to achieve maximum range. Excess thrust (thrust minus drag) is measured versus speed at various altitudes. Taxi tests at higher throttle settings and approaching liftoff speeds will give an early indication of thrust and drag on the ground.5 Data Analysis Thrust. They are also used to update thrust and fuel flow models for fixed throttle settings over a range of altitudes and ambient temperature conditions. However. The idle taxi tests.would be accomplished in the lowest power setting called idle.
Fixed throttle refers to a specified throttle position like MIL. A second computer program is usually provideda prediction deck. Drag will be computed from thrust and excess thrust and modeled versus lift in nondimensional coefficient form. This is designated an inflight thrust deck. but rather computed from flight conditions and engine parameter measurements. and rpm (revolutions per minute). MAX or IDLE. Lift in the form of a nondimensional lift coefficient will be determined and modeled versus angle of attack and Mach number. Thrust is not measured directly. 4 . Thrust is then computed using an engine manufacturerprovided computer program as modified by the airframe contractor to include installation effects. The engine parameters needed usually include pressure. and plotted versus thrust and as a function of flight conditions. modeled. temperature. Drag. The flight test data analyst will compare the inflight thrust deck data to the prediction deck data. Engine fuel flow will be measured. Fuel flow data will be obtained both during the fixed throttle maneuvers (climb. and turn) and during cruise testing. analysis will be performed to attempt to ‘model’ this data.A secondary function is angle of attack (angle between the aircraft body xaxis and the airspeed vector). Lift. accel. Fuel flow. which will predict thrust without knowing any engine parameters (just flight conditions and throttle setting). Then.
The component of aircraft acceleration parallel to the flight path is the longitudinal acceleration ( Ax ). The small difference of these two altitude parameters is discussed in the altitude section. The component of aerodynamic force parallel to the flight path axis is defined as drag. The effect of assuming a constant g is dealt with in the gravity section. The longitudinal load factor ( N x ) is simply the Ax divided by the acceleration of gravity ( g ). The inertial velocity vector defines the inertial flight path axis. and = true airspeed (feet/sec). In this text.174 ft/sec² (foot per second squared).1) E = total energy (footpounds).4) Wt = m ⋅ g 0 PE = Wt ⋅ H where: m Wt H Vt = aircraft mass (slugs). Note: It is assumed that tapeline (or geometric) altitude ( h ) and geopotential altitudes ( H ) are identical.2.1 Flight Path Axis The true airspeed vector defines the flight path (or wind) axis. 5 . we are usually referring to the longitudinal or x component of the wind axis system.5 ⋅ t ⋅ Vt g 0 (2. and PE = potential energy (footpounds). KE = kinetic energy (footpounds). = aircraft gross weight (pounds). In conventional aircraft performance. assuming zero wind: 2 W KE = 0. Lift is the component of aerodynamic force perpendicular to the drag (or flight path) axis. g is assumed a constant at the reference gravity and given the value of 32. = geopotential altitude (feet).0 AXIS SYSTEMS AND EQUATIONS OF MOTION 2. [(pounds force)(seconds)2/(foot)].3) (2. The symbol g 0 will be used to denote the reference gravity. To derive the equations of motion we could start with the following energy relationship: E = KE + PE where: (2.2) (2. when the singular axis is used. Then.
6) Taking the derivative with respect to time (and ignoring wind) yields: dH E dt = dH / dt + Vt ⋅ dVt g dt 0 (2. The Cambridge Air and Space Dictionary (Reference 2.7 then becomes: ! ! Ps = H E = H + Vt ⋅ (Vt ) ! g0 Dividing by Vt yields: (2.Adding the potential and kinetic energy relationships (2. thus being usable for climbing. so the subscript f (f for flight path) is usually deleted on the flight path axis load factors. Equation (2.9) then becomes: ! ! N x = H Vt + Vt g0 (2.8) ! ! ! Ps Vt = ( H E Vt ) = ( H Vt ) + (Vt g 0 ) (2. accelerating or turning. H E = H + Vt 2 ( 2 ⋅ g0 ) (2.7) The derivative of H E with respect to time is called specific excess power and given the symbology of Ps .4) and dividing by Wt yields the following: Vt 2 PE KE E / Wt = + =H + 2 ⋅ g 0 ) Wt Wt ( (2.1) gives the following definition of specific excess power: “Thrust power available to an aircraft in excess of that required to fly at a particular constant height and speed.2) and (2. That is. 6 .” Equation 2. The accelerometer would be sensitive to both aircraft change in velocity ( dVt / dt ) and a component of gravity ( ( dH / dt ) / Vt ).10) In performance analysis.5) The energy per unit weight ( E / Wt ) is called energy altitude (or energy height) ( H E ). the axis system of interest is the flight path axis and not the body or earth axis. we use N x rather than N x f or even N xw (subscript w is for wind axis).9) Envision an accelerometer aligned perfectly with the longitudinal flight path axis and calibrated in units of g. Other references may use other symbologies.
The XYZ body axis system is an orthogonal axis system usually originating at the center of mass of the aircraft. then the flight path acceleration can be computed by transforming first through the angle of attack and then through the sideslip angle.12) 2 (V 2 bx + Vby + Vbz 2 ) (2. 7 . The Yaxis is positive out the right wing and the Zaxis is positive down. Vbz = body axis z component of the true airspeed.13) Vbx = body axis x component of the true airspeed.2 Body Axis The aircraft axis system (Figure 2. Vby = body axis y component of the true airspeed. The relationships for α and β as a function of the body axis true airspeed components are as follows: α = tan −1 (Vbz Vbx ) β = sin −1 (Vby Vt ) Vt = where: (2.1 Aircraft Axis System If the acceleration of the vehicle in the body axis is known. The Xaxis is defined through the center of the fuselage with positive being forward. and Vt = true airspeed.2.11) (2. Figure 2.1) is called the body axis system.
shows angle of attack ([AOA] or α ) and angle of sideslip ([AOSS] or β ) in relation to the body axis velocities.2. Figure 2. The following is the equivalent symbology for Figure 2. Wcg = Vbz Note: Positive directions are shown. In three dimensions.2 Angle of Attack and Sideslip Definitions AOA ( α ) is the angle between the Xbody axis and the projection of the true airspeed vector ( Vt ⋅ cos β ) on the XZ body axis plane. the α transformation matrix from the body axis to the flight path axis is as follows: cos α [α ] = 0 − sin α 0 sin α 1 0 0 cos α (2. Vcg = Vby c.2. AOSS ( β ) is the angle between the velocity vector and the XZ body plane. a.14) In three dimensions. U cg = Vbx b.3 True AOA and Sideslip Definitions The following illustration. the β transformation matrix from the body axis to the flight path axis is as follows: 8 .
as shown in equation 2.19) Ax . and with rate 9 .21) (2. There.18) {A} = [ β ]⋅ [α ]{Ab } where: (2.22) The topic of axis transformations is dealt with in more detail in the accelerometer section. flight path axis.18 assuming zero sideslip. down). rather than the above flight path accelerations.16 for the longitudinal load factor in the flight path axis yields equation 2. east.22.17.17 may be approximated by equation 2. we will deal with inertial axis (north.16 is as follows: (2.17) The vast majority of performance maneuvers produce very low sideslip and lateral acceleration such that equation 2. N x = Ax / g0 N y = Ay / g0 N z = − Az / g 0 Note the sign change on the Z component. (2. Aby . Ax = cos β ⋅ cos α ⋅ Abx + sin β ⋅ Aby + cos β ⋅ sin α ⋅ Abz (2. Ax ≅ cos α ⋅ Abx + sin α ⋅ Abz In matrix shorthand.20) (2. Az Abx . analysis is performed using the flight path axis load factors. and = three components of body axis accelerations. Usually. Ay . Abz = three components of flight path accelerations. equation 2.16) Multiplying the equation 2.20 through 2. cos β [ β ] = − sin β 0 sin β cos β 0 0 0 1 (2.15) The transformation of the acceleration from the body axis to the flight path axis is as follows (a subscript f [for flight path] will be dropped for the flight path axis): Ax cos β Ay = − sin β A 0 z sin β cos β 0 0 cos α 0 ⋅ 0 1 − sin α 0 sin α Abx 1 0 ⋅ Aby 0 cos α Abz (2.
Fg . it . Fe . simplified in that all forces are acting through a single point. L . This is called the point mass model. 10 .3 and 2. however. Transformations are made to the body axis where the rate corrections are applied. e.angle between xbody axis and the flight path axis.drag acting parallel to the flight path. D .corrections to accelerations and velocities in the body axis. d. f. a.angle between horizontal and xbody axis (not shown above). α . θ . and h.lift acting perpendicular to the flight path. Figure 2.4 InFlight Forces Figure 2.angle of attack .3 InFlight Forces The flight path axis is defined by the true airspeed ( Vt ) vector.3 illustrates the X and Z forces acting on an aircraft in flight. g. often equals zero.gross thrust – acting through the engine axis. A more complex model would distribute the lift and drag forces between the wing and tail. It is.angle between horizontal and the flight path. γ . What we might otherwise call the trailing edge flap of the wing provides the pitching moment that a tail usually would.thrust incidence angle (not shown) – angle above the xbody axis through which the gross thrust acts. 2.4 illustrate the basic forces and angles of a typical aircraft in flight. Most conventional aircraft simulations utilize this simplification.net propulsive drag – acting through the flight path axis. Figures 2. The tail may be a part of the wing as in an aircraft like the French Mirage. b. c.pitch attitude .flight path angle .
Figure 2. most will define α as the angle between the flight path axis and the xbody axis. be in a different direction and a different magnitude. which is the definition used in this handbook. However. However.24) Some airframe manufacturers will define α as the angle between the flight path axis and the wing axis. this is a three dimensional relationship that we can represent in vector notation as follows: " " " Vt = Vg + Vw (2. Fex = [ Fg ⋅ cos(α + it ) − Fe ] − D (2. in general.23) Fex = excess thrust. The true airspeed velocity vector and the inertial (or ground) speed vector will.25) 11 . The vector relationship between true airspeed and groundspeed is simply airspeed equals groundspeed plus windspeed.4 Axis System Angle Diagram Summing forces in the longitudinal or Xflight path axis: ∑ where: x W Fx = m ⋅ Ax = t ⋅ ( N x ⋅ g0 ) = N x ⋅ Wt = Fex g0 (2.
where: " Vt = true airspeed vector . let’s say you are flying due north. with zero sideslip. Wind direction. in the aero community. ed. if one uses a negative sign and is consistant with definitions. at 500 knots.26) (2.M. Or in this case. by meteorological convention. and " Vw = wind speed vector . SECTION 2.. For instance. 1995. Cambridge Air and Space Dictionary. However.B. That would mean the wind is 100 knots blowing from due north. and L = lift . " Vg = ground speed vector . This author considers plus to be the ‘correct’ sign.0 REFERENCE 2. The propulsive drag ( Fe ) is only in the longitudinal flight path axis so that its contribution normal to the flight path is zero. Airspeed equals groundspeed plus wind (plus is italicized to place emphasis). There is. some controversy as to the sign convention. Assume there is a 100 knot wind at 0 degrees. Cambridge University Press. Heading is the direction the aircraft is pointing. P.1 Walker. 12 . Summing forces in the normal or Zflight path axis: ∑F z W = m ⋅ Az = t ⋅ ( N z ⋅ g 0 ) = N z ⋅ Wt g0 N z ⋅Wt = L + Fg ⋅ sin(α + it ) (2. a pure headwind of 100 knots. If you have a 100knot headwind and a 500knot true airspeed then the groundspeed is 400 knots. is the direction from which the wind is blowing.27) where: N z = normal load factor . the results will come out the same.
but this difference grows in proportion the square of altitude from the center of earth. they could measure the outside ambient pressure. The equation for static equilibrium of the element of air is as follows (the unit dimension into the page ( dy ) is not shown in Figure 3. That altitude is the pressure altitude. In order to derive a relationship between pressure and pressure altitude. In the early days of flight. The pressure on the bottom of the element is P . This requires that sea level pressure is exactly the standard atmosphere value and that temperature is precisely standard day at all altitudes (not just at the altitude being considered). an altitude commonly used to compute pistonpowered light aircraft performance is density altitude ( H d ). which we will denote with the symbology H C . the technology was not available to measure altitude onboard an aircraft. Geometric altitude is the physical.000 feet is less than 200 feet. The length of a foot of geometric altitude does not vary as a function of temperature or gravity variation with altitude.2 Hydrostatic Equation We will derive the relationship between atmospheric pressure and altitude. it became necessary to define another altitude called geopotential altitude ( H ). generally. Envision a cubic element of air with unit horizontal dimensions ( dx and dy ) and a height equal to dh . The gravity model that has been used to define the geopotential altitude is a simplified model based upon reference gravity at sea level ( g 0 = 32. Density altitude and pressure altitude is the same on a standard day at the altitude being considered. just as pressure altitude is proportional to atmospheric pressure. the difference between h and H at 50. In this case. A standard atmosphere was defined which allowed the computation of an altitude that was proportional to the ambient pressure. However. Density altitude is useful for light aircraft primarily because engine performance is generally proportional more to density than to pressure for internal combustion engines. For the standard atmosphere model.174 ft/sec2) and gravity varying with altitude as per the inverse square gravity relationship.1 Introduction – Altitude There are several forms of altitude of interest in aircraft performance. linear altitude measured from mean sea level. For this text. where c stands for calibrated. Mean sea level is defined (from Britannica) as the height of the sea surface averaged over all stages of the tide over a long period of time.0 ALTITUDE 3. it is not required that temperatures be standard at all altitudes as was the case for H and Hc being identical. 3. The first altitude is geometric (or tapeline) altitude ( h ).3. Density altitude is proportional to atmospheric density. where the radius of the earth is over 20 million feet. Finally. The length of geopotential altitude foot varies with increasing altitude proportional to the change in gravity with altitude. all units will be in feet. H C and H are identical by definition. The pressure on the top of the element is P + dP .1): W = ρ ⋅ g ⋅ dx ⋅ dy ⋅ dz = weight of the element of air (3.1) 13 . As will be shown later.
= acceleration of gravity. = density.4) 14 . and = height increment.3) P ρ g h dh = pressure.1 Element of Air ( P + dP ) = P − ρ ⋅ g ⋅ dx ⋅ dy ⋅ dz = P − ρ ⋅ g ⋅ dh Since dx and dy are of unit length.P + dP W dh dx P Figure 3.2) dP = − ρ ⋅ g ⋅ dh where: (3. Using the inverse square gravity law: r g = g0 ⋅ 0 ( r0 + h ) 2 (3. = height. (3. and the height ( dz ) is equal to dh .
5) ρ= P where: (R ⋅T ) (3.356.3: P dP = − R ⋅T 2 r0 ⋅ dh ⋅ g0 ⋅ r0 + h 2 (3. Substituting 3.7) dP / P = − ( g 0 R ) ⋅ (1/ T ) ⋅ r0 ( r0 + h ) ⋅ dh (3.4 and 3.9.8136 ft²/(sec²°K).6 into 3. g0 = reference gravity (32. and = 9. Standard Atmosphere.089.772 meters.17405 ft/sec²).8) It is not a simple matter to integrate the above equation exactly.6) T = ambient temperature. and R = gas constant = 3. Value for R is converted from metric units using the 1976 U.855. = 6. The concept of a geopotential altitude was introduced to allow for the integration. g ⋅ dh = g0 ⋅ dH where: (3. and = geopotential altitude.9) g h H = gravity at altitude h .80665 m/sec2 (exactly by international agreement). 3. = tapeline (or geometric) altitude. 15 .S.3 Geopotential Altitude Geopotential altitude is developed from equation 3.5: (3.where: r0 = reference radius of the earth (20. Introducing the ideal gas equation of state: P = ρ ⋅ R ⋅T Solving for ρ in 3.553 ft).
2 to a geometric altitude of 86 kilometers. contained just one segment.S. The region up to about 17 kilometers (56. 3.” The temperature decreases rapidly with altitude in this region. To about 50 kilometers (164. the temperatures for the model atmosphere are included in Tables 3. The formulas are as follows: θ= T T = TSL 288. defined by the French in 1919. the transition to the next layer is called the tropopause.1 and 3. The constants in that segment are still the same today (as of 1976). The rising warm air meets the sinking cold air and the air tends to “turn over” like a “ball” – hence the term troposphere. g dH = ⋅ dh g0 Substituting 3. hence. 3.A tapeline foot is the same physical length independent of height while a geopotential foot expands with increasing altitude linearly with the corresponding decrease in gravity. A standard atmosphere model has been defined which contains only constant or linear temperature segments. The profile is presented in Tables 3. This standard atmosphere purports to represent an average temperature model of the earth’s atmosphere throughout the world and during the various seasons. respectively.11) (3. so we will not discuss those.15 (3. However.S. The first standard atmosphere. ‘turning’ and sphaira.000 feet). Standard Atmosphere model is (as of the writing of this handbook) the accepted temperature and pressure profile model in the United States. the temperature rises slowly in a region called the stratosphere. Altitudes higher than 50 kilometers are above the region of conventional aircraft performance. Standard Atmosphere The 1976 U. the ratio of ambient temperature to standard temperature at sea level and the ratio of ambient pressure to standard pressure at sea level. This means you can look up the integration formula in a table of integrals.10 into 3.000 feet) is known as the troposphere.6: (3.13) 16 . ‘ball’.1 and 3. Quoting from Britannica Online: “troposphere . One would pause between layers. These are.10) dP = − ρ ⋅ g0 ⋅ dH = − P ⋅ g 0 ⋅ dH ( R ⋅ T ) dP / P = ( − g 0 R ) ⋅ ( dH / T ) (3.3 and using 3.5 Temperature and Pressure Ratio We will define temperature ratio ( θ ) and pressure ratio ( δ ).2.12) The above formula can be integrated if T either is a constant or is linearly varying with geopotential altitude ( H ).4 1976 U.a term derived from the Greek words tropos.
000 55.940 0.4595 0.65 472.2166 11.000 15.000 60.3048 216.68 391.8969 0.000 232.43 248.34 258.852 278.000 20.53 238.9812 288.6877 0.987 0.6805 20.68 308. Table 3.62 228.2234 0.0000 0.000 65.78 117.43 497.49 785.9656 0.000 36.089.09001 0.0077983 Notes: 1.39805 71. The highest altitude in the table is an even 86.7519 0.65 216.617 0.65 216.000 154.5643 0.31 628.78 Pressure Ratio ( δ ) 1.3454 32.δ= where: P P = PSL 2116.65 2. and P = units of pounds/foot2.0000 270.14) T = units of degrees K.2353 0.6096 214.082632 84.65 216.386 N/A 186.000 10.15 2.8625 0.24 40.80 1455.7519 0.7937 0.0000 216.71 218.116.1 1976 U.000 65.22 1760.81 216.000 25.47 149. Table 3.000 ft) (°K) 0 0 1.7519 0.3048.S.65 216.000 45. The temperature gradient and base temperature in the first segment of the standard atmosphere has remained unchanged since the 1925 U.323 0.65 216.2 STANDARD ATMOSPHERE PRESSURE AND TEMPERATURE Geopotential Altitude ( H ) (ft) 0.8534 270.01 242.000 104.15 278.65 114.000 30.27 972.1145 0.1289 47.07078 0.65 Temperature Ratio ( θ ) 1.8534 228.21 190.65 18. 3.8320 0. 2.95 472. The standard atmosphere is defined in metric units.S.000 36. The exact conversion factor from meters to feet is to divide meters by 0.000 meters geometric (tapeline) altitude.31632 51.22 (3.089 0.000 Ambient Pressure ( P ) (lbs/ft2) 2116.0000 0.95 0.3711 0.8281 0.9312 0.00 5.24 268.05566 Ambient Temperature ( T ) (°K) 288.7519 0.7519 0. Standard Atmosphere.65 0.7519 0.7594 0.000 35.33 1194. STANDARD ATMOSPHERE Geopotential Geopotential Temperature Height Height Gradient Temperature Pressure (m) (ft) (lbs/ft2) (°K/1.7519 17 .199 0.1455 0.000 167.65 1.1851 0.000 50.2970 0.65 216.
12: (3.350 92.054 57.7882 The numbers in Tables 3.6 Pressure Altitude 3.51 221.1 and 3.1 Case 1: Constant Temperature T = T0 Substituting 3. there will be variation from that model (refer to Appendix A for what the average variation is for data taken above Edwards AFB).7565 0.08 225.15) dP / P = ( − g 0 R ) ⋅ ( dH T0 ) (3.7724 0.18) ( R ⋅ T0 ) P H = H0 − ⋅ ln g0 P0 (3.15 into the relationship 3.684 73.000 95.000 80.674 45.616. On any given day.56 224.05403 0. Since g0 .17) Solving for P in 3.2 (Concluded) STANDARD ATMOSPHERE PRESSURE AND TEMPERATURE Geopotential Altitude ( H ) (ft) 65.000 Ambient Pressure ( P ) (lbs/ft2) 114.04380 0.7671 0. R and T0 are each constant: ∫ − g0 − g0 dP = ln( P ) − ln( P0 ) = ⋅ H − H0 ) ⋅ ∫ dH = (R ⋅T ) (R ⋅T ) ( P 0 0 (3.03 222.000 100.02725 0.7618 0.01354 0.000 75.000 85.61 227.65 217.01076 Ambient Temperature ( T ) (°K) 216.13 Temperature Ratio ( θ ) 0.16) We will integrate using a table of integrals and relationships for natural logarithms.02155 0.2 represent the model atmosphere.8 70.01707 0.Table 3.123 28.17: P = P0 ⋅ e Solving for H : g0 ⋅ H − H 0 )} − ( R ⋅T0 ) {( (3.6.000 90.99 219.768 Pressure Ratio ( δ ) 0.19) 18 .7777 0.7820 0.7519 0. 3.656 22.608 36.03452 0.
and H 0 = 36.5 °C).68 pounds/ft2 at H = H 0 .6.22) Solving for P : P = P0 ⋅ 1 + a ⋅ ( H − H 0 ) T 0 g − 0 ( R⋅a ) (3.For the segment of the atmosphere from 11. c.24 feet (11.65 °K (69.000 m). 3. and a = temperature gradient (deg K/foot).000 meters (36. into the relationship (3.12) dP / P = ( − g 0 R ) ⋅ ( dH / T ) : g0 dP / P = − ⋅ dH R ⋅ (T0 + a ⋅ [ H − H 0 ]) (3. H 0 = base geopotential altitude.21) Integrating from a table of integrals: ∫ (a + b ⋅ x ) = b ⋅ ln (a + bx ) Then using the relationship ln(u ) − ln(v) = ln(u / v) : g (T0 + a ( H − H 0 )) ln P = − 0 P ⋅ ln 0 T0 ( R ⋅ a ) dx 1 (3.000 meters (65.089.23) Or solving for H : 19 . P0 = 472. Substituting.20) where: T0 = base temperature. again.617 feet): a. T0 = 216.7 °F or 56.2 Case 2: Linearly Varying Temperature Assume a temperature that varies linearly with altitude as follows: T = T0 + a ⋅ ( H − H 0 ) (3.089 feet) to 20. b.
2559 (3.255876 (round to 5.000 meters).28) Equation 3.26) (3.27) Solving for H : (1 5.87559 E − 6 ) (3.325 pascals (exactly.29) PSL = standard sea level pressure = 101.25) − g0 (R ⋅ a) = 32. The unit pascal has been defined as a newton of force per square meter.22).92 in.013.2248195 pounds force.116. substituting constants (from the international standard atmosphere) [for English units]: −a = T0 (1. Using the pressure ratio (δ) as defined in equation 3.8136 ⋅ (1. = 1. and = 29.116.9812 /1000 ) P P0 = (1 − 6.24) For the first segment of the standard atmosphere (zero to 11.2559 ) P 1 − P 0 H= 6.87559 E − 6 ) ( (3. ≅ 760 mm Hg.089.87559 E − 6 ⋅ H ) 5. zero to 36.− ( R⋅a ) g0 T P H = H 0 + − 1 ⋅ 0 a P0 (3.17405 = 5.14.26 is the definition of pressure altitude for altitudes from zero to 36.24 feet).9812 /1000 ) = 6. by international agreement).15 (round to 6.000 meters. δ =P where: PSL (3.8755856 E − 6 288. Hg 20 . In various English units: PSL = 2.25 millibar (mb).089 feet (zero to 11.2166 pounds/ft² (usually rounded to 2. A newton has units of (kg m/sec2). One newton is equal to 0.2559) 3089.
Substituting 3.31): δ = 0.9812 = − ⋅ H = 1 − 6.87559 E − 6 ⋅ H 288.24 )} (3.617 feet: g0 /( R ⋅ T0 ) = 32.34) For the temperature ratio ( θ ).089 feet pressure altitude.617 feet) is a constant temperature ( T =56.87559 E − 6 ⋅ H C ) 5. using equation 3.8136 ⋅ 216. The symbol H C is used for pressure altitude to distinguish it from the geopotential altitude ( H ).089 feet and 65.87559 E − 6 ⋅ H 288.2559 (3.15 288.2559) HC = 6.3048 meters per foot. 000 (3.22336 ⋅ e{ [ − 4.806343E − 5 (3089.31) For the temperature ratio ( θ ).000meter point is 36.87559 E − 6 ) ( (3.24 feet.15 288.32) The second segment of the standard atmosphere (11.35) 21 .806343 E − 5]⋅( H C − 36089. 000 (3.15 1.9812 = − ⋅ H = 1 − 6. The standard atmosphere is defined in metric units.000 meters) (36.089. Similarly: δ = (1 − 6. using equation 3.805. For the altitude segment between 36.000 to 20.5 degrees C) segment. the 11. English units require the conversion factor of 0.15 1.33) Computing δ for H = 36.84 g0 (3.30) The above is for zero to 36.089 to 65.29 into 3. Pressure altitude and geopotential altitudes are only identical for the model atmosphere.17405 = 4.65) R ⋅ T0 = 20.20 and substituting constants (from the international standard atmosphere): θ= T T0 1.28: 1 − δ (1 5. 089.20 and substituting constants (from the international standard atmosphere): θ= T T0 1. For instance.24 feet using the δ formula for the first segment of the atmosphere (equation 3.
Log(delta) versus Pressure Altitude [K Feet] 0 1 Log(delta): delta=Pa/Pasl 2 3 4 5 6 0 50 100 150 Pressure Altitude (ft*1. 22 . The standard atmosphere pressure ratio versus pressure altitude is nearly a straightline logarithmic function as can be seen in Figure 3. Standard Atmosphere can be derived by simply applying the above equations since all segments of the standard atmosphere are either constant temperature or linearly varying temperature versus pressure altitude. all the segments of the standard atmosphere are either constant temperature or linearly varying with altitude. As can be seen.S. As discussed earlier. etc.3 illustrates the linear temperature segments.00010946.1145. at 100K it is 0.01076. Figure 3.2 Logarithmic Variation of Pressure Ratio The logarithm in Figure 3. at 150K it is 0.The equations for any segment of the 1976 U. For instance at 50K the pressure ratio is 0.2 is base 10.000) 200 250 300 Figure 3.2. at each 50K point the atmospheric pressure decreases by a factor of 1/10th.
36 into 3. a = r0 .37 and solving for dH : 2 (3.7 Geopotential Altitude (H) versus Geometric Altitude (h) Using the inverse square gravity law and the definition of H: r g = g0 ⋅ 0 ( r0 + h ) g ⋅ dh = g0 ⋅ dH Substituting 3.36) (3.Standard A tm o sphere Te m p erature 300 250 P res sure Altitud e (ft*1000) 200 150 100 50 0 1 80 20 0 220 2 40 Stand ard T e m p erature (de g K) 2 60 28 0 300 Figure 3.38) Integrating gives the relationship between H and h (or tapeline). b = 1 and x = h .37) r 0 dH = r0 + h ( ) ⋅ dh 2 (3. From a table of integrals: ∫ (a + bx ) In our case.3 Standard Atmosphere Temperature 3. dx 2 =− 1 b ( a + bx ) Factoring out the r0 term in the numerator: 2 23 .
39) Multiply the first term in square brackets by r0 r0 and the second term by ( r0 + h ) ( r0 + h ) .40) (3. ( r0 + h ) −r0 2 H = r0 ⋅ + ( r0 + h ) ⋅ r0 ( r0 + h ) ⋅ r0 By factoring terms. ∆ tapeline ≅ ∆ geopotential): 24 .45) Since dh ≅ dH (i.41) At 50. for a difference of only 119 feet. H computes to be 49..Nonstandard Day A standard temperature may exist at a given altitude on a test day but there would never be a standard atmosphere at all altitudes except in computer models.000 feet tapeline altitude (the upper limit of most conventional aircraft performance testing).43) dP P = −( g0 / R ) ⋅ (dH / T ) test day There can be a significant difference between having a standard atmosphere and achieving standard temperature at a given altitude.39 and 3. we get: r H = 0 ⋅ h [r0 = 20. The pressure levels at a given pressure altitude are by definition the same whatever the temperature.16): dP P = −( g 0 / R ) ⋅ (dH C / TSTD ) standard day (3.40. we could equate the right sides of equations 3.44) where: T = Ttest day .42) (3. dH = T T ⋅ dH C STD (3. 3.24 percent. or 0. dH C / TSTD = dH / T (3.553 feet ] ( r0 + h ) (3.855. Therefore.881 feet.8 Geopotential versus Pressure Altitude . Using the basic dP / P relationship (3.H = r0 ⋅ ∫ 2 0 h dh ( r0 + h ) 2 1 1 2 = r0 ⋅ − + ( r0 + h ) r0 (3.e.
7 (3. To illustrate the effect of wind on climb performance we will take data from January at pressure altitudes of 13. Standard sea level pressure in millibars is 1013.574 feet (400 mb).8 241.2 1. As can be seen.45: ! ( 228.48) The physical rate of climb (the derivative of tapeline altitude) is 4. Then.4 Delta Temperature (deg K) 3.5 25 .7 degrees Kelvin (K).801 23. presume a test day temperature that is 10. 000 = 1.000 feet to 50.0 242. 043.000 feet every 10. Inserting these values into 3.3 EDWARDS AVERAGE WEATHER DATA FOR JANUARY Pressure Altitude (ft) 13.000 feet is 228. it is not uncommon to be off standard day by 10 degrees C or more.9 Effect of Wind Gradient Average windspeed and direction data for the AFFTC.7 43. This is average data for a time span of over 30 years.937 Standard Temperature (deg K) 260. can be found in Appendix A. This is typical for F16 and large transport aircraft.000 ft/min = rate of change of pressure altitude.574 Geometric Altitude (ft) 14. as a function of altitude for each month.4 Windspeed (kts) 28. Table 3.7 228.47) Sample calculation: Assume a climb through 30. We will conduct calculations for a climb speed of 280 knots calibrated airspeed ( VC ).4 percent higher than the rate of change of pressure altitude for being 10 degrees C hotter than standard day.0 degrees C hotter than standard day. Average temperatures for the Air Force Flight Test Center (AFFTC) at altitudes from 10.0 ) h= ⋅1.7 + 10.000 feet can be found in Appendix A. Standard day temperature at 30. dh = T T ⋅ dH C STD (3.801 feet (600 mb [millibar]) and 23. 3. Table 3.000 feet with dH C / dt = 1.0 Ambient Temperature (deg K) 264. for instance: ⋅ H = rate of climb ! h =T !C TSTD (3.3 contains the average meteorological data and computed variables.46) Or in a climb.065 23.25.
The inertial energy altitude. The airspeed vector is inclined with respect to the horizontal by the flight path angle while the winds are in the horizontal plane. as derived in the first section.1 440.937 feet.10 Density Altitude Density altitude is nothing more than an altitude on a test day that produces an equivalent density on a standard day.45318. whose power output is generally proportional to air density (i. b. and c. density altitude). When climb performance is measured using the altimeter (pressure altitude) large errors could be induced due to wind gradients. tailwind.4 shows the values of groundspeed and energy altitude for a headwind.312 feet.31 and 3. is as follows: HE = h + Vg 2 (2 ⋅ g0 ) (3. on an average day over Edwards AFB in January. the pressure and temperature ratio equations for the first segment of the atmosphere are appropriate.897 Headwind ( HE ) (ft) 18.937 Airspeed ( Vt ) (kts) 343. Table 3.004 feet.449 29. Comparing these numbers.5 Headwind ( Vg ) (kts) 314.507 Calculating the delta energy altitudes: a.4 ENERGY ALTITUDE EFFECT OF WIND GRADIENT Altitude (h) (ft) 14.612 feet. In making this comparison we have ignored the flight path angle. The density altitude parameter has been used primarily for reciprocating engines. and zero wind.065 to 23.4 396. In each case. 3.0 Tailwind ( Vg ) (kts) 372.453 Tailwind ( HE ) (ft) 20.194 32.e.0 percent compared to zero wind.065 23.Now.0 No Wind ( HE ) (ft) 19..7 353. Since the reciprocating engine is generally flown at altitudes below 11 km (kilometer). 26 .285 = 11. This is 11. This is over the geometric altitude range of 14. The relations (equations 3.285 30.449 = 11. The wind gradient effect can now be accounted for using GPS or INS data. Zero Wind ∆H E = 30.50720. This is why opposite heading climb data are obtained ("sawtooth climbs").308 feet greater flying with a tailwind than flying into a headwind. the change in energy altitude is 1.49) Table 3. the calibrated airspeed is the same at 280 knots.32) were derived above in the altitude portion of this section. Tailwind ∆H E = 32. Headwind ∆H E = 29.89719. we wish to compute the change in energy altitude for climbing directly into the wind (headwind) and with the wind (tailwind).194 = 12.9 percent compared to the headwind number or 6.
the temperature would have been: T = 288.6877 On a standard day. pressure ratio is a function of pressure altitude only and vice versa. That is.9312 = 268.δ = (1 − 6.2559] 1 − δ θ = Hd = 6.2559 Let’s give an example.87559 E − 6 ⋅ H C ) σ= = θ (1 − 6.3 = ( 268.7384 0.2559 θ = (1 − 6. δ (1 − 6.87559 E − 6 6. We are at 10.51) For the test day temperature of 100 degrees F: 27 .3°F θ = (1 − 6.87559 E − 6 ⋅ H C ) 5. However.87559 E − 6 ⋅ H C ) The first formula ( δ ) is valid for standard or any nonstandard day.3 − 273.8 + 32 = 23.15) ⋅1. one could define the density altitude ( H d ) as being directly proportional to density as defined by equation 3.87559 E − 6 ( ) (3.87559 E − 6 ⋅ H C ) 5.2559 (3. 000)5.87559 E − 6 ⋅ H C ) 4. We can compute density ratio ( σ ) for a standard day.2559] 1 − σ [1 4.50.9312 The standard day σ is: σ= 0.9312 Solving for H d [1 4.000 feet pressure altitude at 100 degrees F.87559 E − 6 ⋅10. On the other hand. the temperature ratio ( θ ) formula is valid only for standard temperatures.50) The above σ formula is valid only for standard day.15 ⋅ 0.2559 = 0. by taking the ratio of the above formulas.87559 E − 6 ⋅10. σ = (1 − 6.2559 = (1 − 6. The pressure ratio is: δ = (1 − 6.15 ⋅θ = 288.6877 = 0.87559 E − 6 ⋅ H d ) 4. 000 ) = 0.
computing H d we get: 0.000 feet pressure altitude is identical to the density ratio at a density altitude of 14. δ = (1 − 6.67 + 100 ) = 1. calculate the standard density ratio for 14. θ 0.5 shows the resultant pressure error for a 1foot error in pressure altitude.87559 E − 6 ⋅14.0790 Then. 28 . pressure altitude is only a function of ambient pressure and is independent of ambient temperature.2559 / 6. Using the standard atmosphere model formulas.87559 E − 6 ⋅14. the pressure altitude would also be 2.0790 518.52) H d = 14.6877 = = 0. and c. 607 feet versus 10.300 feet. 3.000 feet pressure altitude is the same as at 14.87559 E − 6 (3. the field elevation (geometric height) of the main runway (22/04) is 2. Equation 3. With standard atmospheric conditions.607 feet. we can compute what a 1foot change in altitude will produce in ambient pressure.5733 .θ= The σ for the test day would be: ( 459. That requires more than just being at standard temperature.607 feet pressure altitude on a standard day for that altitude.11 Pressure Altitude Error Due to Ambient Pressure Measurement Error At Edwards AFB. As we have derived. 607) = 0.6373 .000 feet for H C (pressure altitude).607 feet as follows: a. σ = δ 0.300 feet.8996 It checks! The density ratio for 100 degrees F at 10.6373 θ 1. b.67 σ= δ 0.0790 1/ 4.52 shows the density (or σ ) at 100 degrees F at 10.5733 = = 0. θ = (1 − 6.6877 H d = 1 − 1.2559 = 0. 607 )5.8996 . Table 3. To check on our calculations.
Hg ∆H C = 0. Looking at the inches of mercury column. The second limiting factor on altitude accuracy is the ‘position error.972 0. 210 = 1.250 0.820 0.15 0.0011=26 feet b. ∆P =30/16384=0.020 0. Hg) P (millibar) ∆P (millibar) 1.071 27.384: a.0005/0.037 0. However. Hg/(0.0011=6. Hg) ∆P (in.000 40. If full scale were 30 in. At sea level.536 ∆P = 30/65.0018/0.019 5.000 50.5 feet Therefore. Data recording system resolution is a limitation for any parameter. 12.45954 972. a 14bit system will get us to our goal of 1foot accuracy.0003 187. but let us use pressure altitude as an illustration.0010 931. 212 = 4.0 ft Therefore. For higher bit resolution the following numbers are computed: a.921 0.117 in. or 16 “bit” system.096= 0.0002=9.516 0.0002 115.0073/0.024= 0. Clearly.820 0.041 13.300 10. 14.029 8.6 feet c.43 0.0073 in.0004 300.014 0. our error due to recording system resolution is substantially larger at the higher altitudes.009 0.33 0.000 30.0018 ∆H C =0.0011= 1.6 feet d.’ discussed in the air data calibration section.012 3.056 20.49 0. Hg.18509 391.5 PRESSURE ERROR VERSUS ALTITUDE ERROR HC (ft) 0. First.68 0. 216 = 65.096 ∆P = 30/4. An 8bit system breaks full scale into 28 (or 256) parts.0011 in. it appears that at least at sea level. Hg/ft)=107 feet.013.029 in.536= 0.425 0.11446 242.0018 in. this is unacceptable for performance testing.577 0. is the number of digits recorded in the data stream. 29 .68770 1455.076 29.91963 1946. We have the same value for 214 =16. then the resolution of ambient pressure would be 30/256=0. However.630 0. let us see what happens at 50. The AFFTC pacer aircraft use a 16bit system. a 9foot error at 50. Hg ∆H C =0.0011= 0. It turns out that such accuracy level instrumentation is available.000 20. one can see that better than 1/1000th of an inch of mercury accuracy would be required to achieve 1foot accuracy in pressure altitude.885 0. Hg.0008 696.000 feet.890 0.034 0. Hg ∆H C =0. Hg ∆H C =0.024 ∆P = 30/1.538 0.00000 2116.0018/0. The data recording is an 8. this would be an altitude error of 0.22 0.0011 1.029/0.384 ∆P = 30/16.027 0.384= 0.540 0.750 0.000 δ P (psf) ∆P (psf) P (in.000 feet is considered acceptable. There are two other limiting factors on altitude accuracy.21 0.29695 628.0006 465.Table 3.006 Note: The pressure errors are carried to one extra digit than the pressure magnitude. 214 =16.117 in.0005 in.0 2. 10.
calibrated airspeed and equivalent airspeed are equal (or equivalent). east. Windspeed ( Vw ) is the speed of the air mass (wind) with respect to the ground. The first is that calibrated airspeed is ‘calibrated’ to sea level in the sense that it will be exactly equal to true airspeed at sea level.0 AIRSPEED 4. The c (calibrated) has two meanings. Calibrated airspeed ( VC ) is the speed displayed on a typical cockpit airspeed indicator. so that equivalent airspeed is a useful speed for structural testing.1) 30 . This is usually a scalar quantity. generally. which is dimensionless. 4. Once instrument and position error corrections are applied. That pressure is also designated compressible dynamic pressure. except for Mach number ( M ). This is due to obtaining groundspeed from INS (inertial navigation system) or GPS (global positioning system) data sources. However. True airspeed ( Vt ) is the physical speed of the aircraft with respect to the moving air mass. Impact pressure is the difference between total and ambient pressure. The value has instrument and position errors. though components of true airspeed can be computed using axis transformations using INS velocities and angles and windspeeds. The speed of sound is a function of the square root of the ambient temperature. The Mach number ( M ) is the ratio of true airspeed to the local speed of sound. in the real world. Mach numbers less than 1 are referred to as subsonic and those greater than 1 are supersonic. east and down components. In aircraft equipped with an ADC (air data computer). as mentioned above. For this text. the indicated airspeed becomes calibrated airspeed. The second is calibrated versus indicated. there is no such position so there will always be position errors of some magnitude. standard day. is a function only of the impact pressure. A pneumatic instrument (physically driven from pressure inputs) displays an ‘indicated’ value. and down components. There may be some ideal location to place probes where the errors are zero. Structural analysis is often in terms of incompressible dynamic pressure.4. At sea level. This is also a vector quantity with north. standard day.1 Introduction – Airspeed Aircraft speed can be expressed in several forms. but only at that condition. the units will be in either knots (nautical miles per hour) or feet per second. It is a function of only one parameterdifferential (or impact) pressure.2 Speed of Sound The speed of sound is computed by the following formula: a= (γ ⋅ R ⋅ T ) (4. but only at that condition. Groundspeed ( Vg ) is the physical speed relative to the ground and is usually expressed as a vector relationship with north. The instrument errors are errors due to the instrument itself. those corrections are usually already applied in the ADC so that the displayed airspeed is calibrated airspeed. Position errors are those due to the location of pressure probes. Calibrated airspeed. A measure of airspeed that is a function of incompressible dynamic pressure is called equivalent airspeed ( Ve ).
89 ⋅ 0. they specified the speed of sound at sea level standard day as 1116. These values were at 0 degrees C.” Very clever! In the 1650s.where: a γ R = speed of sound (ft/sec).4) From Britannica Online.3048 ⋅ 273. aSL = 1116. but could be computed from the other constants. In that report.089.4788 knots (usually rounded to 661.45) = 661.3) Then. The speed of sound is not directly defined. a = aSL ⋅ θ 4.4 meters per second.45 meters/second 288. a = 1116. Britannica reports a value of 331. which was amended to 331. and = 3.45 meters per second was obtained in 1942. define θ as the ratio of test day temperature to standard day temperature at sea level.40 (ratio of specific heats).4505 ft/sec (usually rounded to 1116. NACA (National Advisory Committee for Aeronautics) published Report No. The first precise value was obtained at the Academy of Sciences in Paris in 1738 at 332 meters per second.15 = 331. He “measured the time difference between spotting the flash of a gun and hearing its report over a long distance.15 °K.8136 ⋅ 288.3 History of the Measurement of the Speed of Sound (4.40 ⋅ 3089.2) = 1. For a sea level standard day.116.15] (4.48) For the speed of sound at temperatures other than standard sea level. In 1942.29 meters per second in 1986. 1235. = 1. Converting the NACA number to meters per second and to 0 degrees C: a.89 feet/second. the ICAO (International Civil Aviation Organization) agreed upon constants for use in a standard atmosphere. a= [1.4505 ft/sec = 340. Standard Atmosphere). the speed of sound in air was first measured by the French scientist Pierre Gassendi in the 1600s at 478.2941 m/sec 31 . two Italians (Giovanni Borelli and Vincenzo Viviani) obtained a much more accurate value of 350 meters per second. Then.15 In 1962 and again in 1976. T = 288. a aSL = ( γ ⋅ R ⋅T ) γ ⋅ R ⋅ TSL = T TSL (4. The speed of sound at sea level in English and metric units is as follows: a.8136 ft²/(sec² °K) (from the 1976 U.S.
NM Hour ⋅ = 1. feet/sec = 6.6878 knots 3. standard atmosphere of 1959. The relationship between true airspeed and the speed with respect to the ground ( Vg ) is: " " " Vt = Vg + Vw = true airspeed vector where: (4. 600 sec One would see the above conversion factor in textbooks published prior to the U.6 Mach Number Mach number ( M ) is defined as the ratio of true airspeed to the local speed of sound.5) " Vw = windspeed vector.5 True Airspeed True airspeed ( Vt ) is the physical speed of the vehicle relative to the moving air mass.45 /1.852 / 0.4 The Nautical Mile The nautical mile (nm) has been set.3048 = 6.6878 = 661. NM = 1. the conversion from knots to feet per second is as follows: a. aSL = 1. to exactly 1. 080.31 knots With the modern (as of this writing) values: b.89 /1. 4. The true airspeed is a vector quantity.116.48 knots 4.6889 knots Hour 3. M= Vt a (4. 076. Using the 1942 speed of sound and the early knots to feet per second conversion one gets: a. Therefore.080 feet.1155 feet Since a knot is 1 nm per hour. we can compute the number of feet per nautical mile. a. which had many of the same constants as the 1962 and 1976 atmospheres.6889 = 661.852 meters. 600 sec An early definition of a nautical mile was an even 6. It is called the British nautical mile. With that definition. aSL = 1. 076.6) 32 .115 NM ⋅ Hour Hour = 1. by international agreement. The conversion factor from feet to meters is also an exact number0.116. the conversion factor becomes: a. feet/sec = 6.S.3048 meters per foot.4.
the free stream total pressure ( Pt ) and the measured total pressure ( Pt′ ) are identical.13) 33 .We could compute Mach number from Pitotstatic theory with the simple expression for differential pressure ( q ) versus total pressure detected by a Pitot tube ( P′ ) and ambient C t pressure ( P ).10) Substituting γ =1. Differential pressure is also compressible dynamic pressure and often designated impact pressure.5 (4.8) Using Bernoulli’s Equation for M < 1 : qC (γ − 1) = 1 + ⋅M 2 P 2 γ (γ −1) −1 (4.11) Solving for M in equation 4. The prime on the total pressure is to denote a measurement behind a normal shock (for M ≥1).8 ⋅ M ) (4.5 2.5] q M = 5 ⋅ C + 1 − 1 P (4.40 for M <1: qc P = (1 + 0. For M <1.5 ) 2.12) For M ≥ 1 : qC P = (1.4 −1 ⋅ 2 ( −0.2 ⋅ M 2 3.9) And the Rayleigh Supersonic Pitot Equation for M ≥ 1 : (γ + 1) qC = P 2 ⋅M2 γ (γ −1) (γ + 1) ⋅ 2 1− γ + 2 ⋅γ ⋅ M ( ) 1 (γ −1) 1/ (γ −1) −1 (4.4 + 2.11: [1 3. qC = Pt′ − P Or dividing both sides by P : qC Pt′ P = P −1 (4.2 ⋅ M 2 ) − 1 3.
5 } by 2.502.Multiply by 1= (2. then use the value of M from the subsonic equation as the initial condition in the 34 . qC + 1 ⋅ 7 ⋅ M 2 − 1 P ( ) 2.5 and divide the second term in the { } brackets by the same 2.892929 Solving for M in the supersonic formula (4.4 2. then multiply both sides by the term 7 ⋅ M 2 − 1 ( ) 2.5 factor.5 .15) 1.5 + 2.5 = 166. For example: a.5 and collect terms.16) Note that one produces the identical value for qC / P when M = 1.5 2 Finally.9216 ⋅ M ) = 1.0 at that point.5 ⋅ (7 ⋅ M M7 2 − 1) 2.5 .9216 ⋅ M 2.0 = 0.5 ⋅ M ) 2 2. qC = 1.50/2.881285 ⋅ C + 1 ⋅ 1 − P 7 ⋅ M 2 (4.16).5 ⋅ 2.42.5 ⋅ 2. Multiply the first term { (1.5 ⋅ M ( P ( 2⋅3.9216) qC 7 = 166. divide both sides by 7 ⋅ M 2 ( ) 2.5 2.5 ⋅ 2.17) As can be seen.5 = 166. solve for M from the M on the right side.23.5 −1 (4.5 2⋅2. One method to approach the supersonic M calculation in a computer algorithm is first determine if M is indeed greater than 1. If M is greater than 1.14) = 1.5 2 qC + 1 ⋅ 7 ⋅ M − 1 P 7⋅M 2 (166.5 −1 (4.12). first add 1 to both sides.2 ⋅ M 2 )3. M appears on both sides of the equation.5 ⋅ 2.5 1 q M = 0.502. 2.8 ⋅ 2.5 − 1 P (7 ⋅ M 2 − 1) (4.52.11) or supersonic (equation 4.23.4 ⋅ 2.42.52.23.52.5 ⋅ 2.287560 ⋅ M = (7 ) ⋅ M [ ] 7 2.9215801 (round to 166. 2.0 by calculating M from the subsonic equation (4.0 is inserted into either the subsonic (equation 4.9216 ⋅ M 7 Then.5) ) ⋅ (−0. qC / P M =1.16) formula.50)2.
For VC < aSL : qC 2 V = 1 + 0. then the relationship between total temperature and ambient temperature ( T ) is as follows: (γ − 1) Tt = T ⋅ 1 + ⋅ M 2 = T ⋅ (1 + 0. we will leave out intermediate steps.supersonic equation. Assuming this probe is in the freestream with no heat loss (adiabatic).usually in just a few iterations. 35 . The subsonic and supersonic Mach number equations are used with the simple substitutions of ( VC / aSL ) for M and PSL for P . airspeed indicators were constructed with a single pressure input being the differential pressure ( qC ).2 ⋅ M 2 ) 2 (4. However.18) 4. sea level (661.5) + 1 − 1 VC = aSL ⋅ 5 ⋅ qC P SL (4. standard day.8 Calibrated Airspeed Historically.5 −1 (4. The gauge is “calibrated” to read true airspeed at sea level standard pressure and temperature. 4.7 Total and Ambient Temperature A total temperature probe is used to measure total temperature ( Tt ).20) For VC ≥ aSL : qC PSL = 166.5 −1 (4.21) Solving for VC and noting that the formula is similar in form to the M equation.19) (1 3. the condition for which the equations are used is no longer subsonic ( M <1) or supersonic ( M >1) but rather calibrated airspeed being less or greater than the speed of sound ( aSL ).9216 ⋅ (VC aSL ) 7 ⋅ (VC aSL )2 − 1 7 2. Then perform a simple iteration until M converges to a value .48 knots).2 ⋅ C a PSL SL 3.
000 1.2. We will leave it to the reader to make that comparison. then successively iterate on the above equation. Figure 4.600 1. Note that VC subsonic formula will converge in VC > aSL and not occurs on both sides of equation 4.000 feet. while the calibrated air speed is directly proportional to compressible dynamic pressure.000 ft H = 20. It should be emphasized that the supersonic formula is M >1. standard day.1 True Airspeed versus Calibrated Airspeed 36 .22.000 800 600 400 200 0 0 100 200 300 400 500 600 700 800 Calibrated Airspeed (kts) H = Sea Level H = 10.22) Notice the differences between equations 4. It just a few steps.5 1 qC ⋅ 1 − VC = aSL ⋅ 0. Figure 4.1 illustrates the difference of true airspeed versus calibrated airspeed.200 1.800 True Airspeed (Standard Day) (kts) 1. In summary.000 ft H = 50.17. the true airspeed is the physical speed of the aircraft with respect to the moving air mass. True Airspeed (standard day) versus Calibrated Airspeed 2.22 and 4. The solution is simply to use the to obtain a first iteration.881285 ⋅ +1 2 PSL V C 7 ⋅ aSL (4.000 ft H = 30. calibrated airspeed is about ½ of true airspeed.400 1.000 ft Note: At 50. The two measures of airspeed are identical at sea level.000 ft H = 40.
28) Therefore.10 Mach Number from True Airspeed and Total Temperature If one has an accurate direct measure of Vt .4.9 Equivalent Airspeed Equivalent airspeed is defined from the incompressible dynamic pressure formula.5 ⋅ ρ ⋅ Vt = 0.15 ⋅ M (4. q = 0. 4. For example: T Vt = aSL ⋅ 288.25) (4.23) (4. so the performance engineer needs to be able to convert Ve to parameters that are more useful. Notice that it is not a function of temperature. Besides equation 4. the equation 4.28 is a handy conversion between Ve and M .26.26) ρ 0 = ρ SL .24) (4. structural analysis is often performed in terms of equivalent airspeed (since it is a direct function of the incompressible dynamic pressure). Since Mach number is M= And σ = δ Vt a = Vt (a SL ⋅ θ ) ) (4. However. σ = ρ ρ SL Ve = σ ⋅ Vt 2 2 Ve = σ ⋅ Vt For the performance engineer. another useful equation is derived. then Ve = σ ⋅ Vt = ( δ θ ) ⋅ (a (a Ve SL SL θ ⋅M M= ⋅ δ ) (4. then M can be computed with the additional measurement of total temperature ( Tt ). there is no practical reason to use equivalent airspeed for anything.29) 37 . The direct Vt measure could come from laser velocimetry.27) θ .5 ⋅ ρ 0 ⋅ Ve 2 2 (4.
30 and 4. M is a function of ambient temperature ( T ).M= (V ⋅ 288.30) Recalling the total temperature equation 4.48 ⋅ T ) t (4.001 InHg Error in Total Pressure 1. That error analysis showed the effect of an error in ambient pressure on pressure altitude.31). Figure 4.0 0 200 400 600 800 1000 1200 1400 1600 1800 True Airspeed (kts) Delta True Airspeed (kts) H = 10. the technology to directly measure true airspeed was not generally available so one must rely on computing M from total ( Pt ) and ambient ( P ) pressure measurements.1 0. A similar analysis can be performed for an error in total pressure and its effect on the calculation of true airspeed.2 True Airspeed Error for 0.15 ) (661. Hg in the total pressure measurement.2 0.2 ⋅ M 2 and solving for T : ( ) T= Tt (1 + 0. An initial estimate of standard day might be chosen for the initial value of T for the iteration. Effect of 0. 4.001 in.7 0.2 shows that effect for an error of 0. Hg Error 38 .9 0.18.0 0.000 ft Figure 4. This is due to the way we have chosen to compute M using a measurement of Vt . At the time of this writing.31) Then.000 ft H = 30. one would iterate between the M and T equations (4.4 0.8 0.11 Airspeed Error Due to Error in Total Pressure An error analysis was presented at the end of the altitude section.000 ft H = 50. Tt = T ⋅ 1 + 0.2 ⋅ M ) 2 (4.5 0.001 in.6 0.3 0. In this case.
VC = f (qC ) calibrated airspeed.We have summarized the functional relationships derived in the altitude and airspeed sections as functions of three basic measurements: total pressure ( P′ ). P) Mach number. M = f ( Pt′ . c. and f. qC = Pt′ − P compressible dynamic pressure. M ) ambient temperature. H C = f ( P ) pressure altitude. a. and total temperature ( Tt ). ambient (or static) t pressure ( P ). Note that Mach number is obtained without a measurement of temperature. d. e. T ) true airspeed. Vt = f ( M . T = f (Tt . b. 39 .
the area is a reference wing area.0 LIFT AND DRAG 5.3) To show how the above equivalence is developed.5 ⋅ ρ ⋅ Vt = 0. (R ⋅T ) Figure 5.5 ⋅ 2 P (γ ⋅ R ⋅ T ) ⋅ M = 0. and S = reference wing area (feet²).S. 5.5 ⋅ ρ ⋅ Vt = 0.7 ⋅ P ⋅ M 2 2 (5. and c. we use formulas we previously derived. q = incompressible dynamic pressure (pounds/feet²). b. The force perpendicular to the true airspeed vector is the lift ( L ) force.2) D = drag (pounds). Defining q : q = 0. The lift and drag coefficients are defined as follows: CD = D / ( q ⋅ S ) drag coefficient CL = L / ( q ⋅ S ) lift coefficient where: (5.7 ⋅ P ⋅ M 2 .4 ⋅ M = 0. the force parallel to true airspeed ( Vt ) is the retarding force drag ( D ).1 illustrates the difference between the compressible ( qC ) and incompressible (q ) dynamic pressure.1 Introduction The aerodynamic force axis system used for aircraft performance is defined by the true airspeed vector.1).1) (5. a.2 Definition of Lift and Drag Coefficient Relationships Lift and drag are referenced to incompressible dynamic pressure and a reference area so that the coefficients are nondimensional. Octave Chanute in his 1897 book.5. uses the terminology resistance for what we now refer to as drag.5 ⋅1. Vt = γ ⋅ R ⋅ T ⋅ M . Assuming zero sideslip angle ( β ). L = lift (pounds). Progress in Flying Machines (Reference 5. ρ= P (R ⋅T ) .000 feet). 40 . The constants in the following equations are derived from the 1976 U. Standard Atmosphere (which are the same as in the 1962 U.S. Standard Atmosphere below 65. q = 0. In aircraft applications.
5 Figure 5.000675) A drag coefficient of 0.22 ⋅ δ ⋅ M 2 = 1481.2 1.5 2. α = f (CL . and b.7 ⋅ 2116.0 2 .3 The Drag Polar and Lift Curve (5.3 1. CD = f (CL .6) The drag polar and lift curve are usually presented as a function of lift coefficient and Mach number as follows: a.5) CD = 0.R atio o f C om pre ss ib le to In co m p re ss ible D yn am ic Pres s ure 1. 41 .00067506 ⋅ D (δ ⋅ M 2 ⋅ S ) (The constant is usually rounded to 0.4 1.0001 is defined as one drag count.6 1.5 1.0 M a c h N um b er 1.0 0. This is typically for a reference logitudinal center of gravity and Reynolds number or altitude. M ) lift curve . CL = 0.4) (5.00067506 ⋅ L (δ ⋅ M 2 ⋅ S ) 5. M ) drag polar .1 1.22) (pounds per ft2) q = 0.0 0.5 q c/qba r 1.1 Ratio of Compressible to Incompressible Dynamic Pressure More convenient forms for CD and CL are as follows: P = δ ⋅ 2116.3516 ⋅ δ ⋅ M 2 (5.7 1.2166 (usually rounded to 2116.8 1.
0 196.00 1.1027 0.0000 0.000 60.000 60. The numbers used are for standard day.1671 1.7985 2.15 288.4588 2.20 0.1027 0.00 1. we used Sutherland’s Law.15 288.0000 0.48 793.1010 8.2970 0.000 30.0000 1.00 1.00 3.65 216.0708 0. We define an index that is a ratio of Reynolds number to the Reynolds number at standard day.2970 0.1 gives a sense of the magnitude of RN . (T + 110 ) δ RNI = ⋅ 2 398. sea level at a given Mach number.0708 0.4010 0.00.8 223.7937 0.7101 1.6 340.5559 0.7937 0.4377 0.60 1.65 216.15 288.0708 0.1 132.0708 0. Then.71 216.1882 VC (knots) 66.3 396.0 390.60 1.0000 1.0000 1.0000 1.60 0.101E + 6) ⋅ M ⋅ l ⋅ RNI (5.0 626.) where: (5.000 δ 1. RN = (7.0000 1. = characteristic length (feet) ( l is usually the MAC [mean aerodynamic chord]). sea level values into the RNI equation you would get 1.9 661.000 60.0000 1.000 60.00 Altitude (ft) 0 0 0 0 0 30.5212 1.0000 0.60 2.7294 1.1027 0.0000 1.9) For a characteristic length ( l ) of 1.4 Reynolds Number Reynolds number is defined as follows: RN = where: ρ ⋅Vt ⋅ l µ (5.4202 4.65 θ 1.1027 0.000 30.0000 1.7397 0.65 216.000 60.7519 RNI RN / l (106/ft) 1. and l µ = viscosity (slugs/[feet sec]). which is a relationship for µ in terms of ambient temperature.15 θ (Note that if one were to insert standard day.0000 1.0000 1.0708 T (deg K) 288. Table 5. Table 5.10 0.15 228.7519 0.4010 0.0000 1.9 42 .8) RNI = Reynolds number index.7519 0.5.7519 0.65 216.4010 0.8474 4.60 1.71 228. To compute viscosity.0 110.7) RN = Reynolds number.0000 1.0 643.20 0.15 288.0.7519 0.1027 0.2970 0.9 430.71 228.2606 7.1 REYNOLDS NUMBER VARIATION WITH MACH NUMBER AND ALTITUDE Mach Number 0.
11) All of the sample problems in this text used equations 5.144 ⋅ M 2 ) −0.074 5 RN (5. S CD = C f ⋅ wet S (5.14.11. Cf = 1.328 RN (5.0001 CD ) per 2. equation 5. Cf = 0. Cf = Effect of Mach number: 0.15) Equations 5. 700 − 5 RN RN (5.13) A laminar flow empirical formula was developed by Blasius and shown in equation 5.15 are plotted versus the logarithm to the base 10 of Reynolds number in Figure 5. a 1drag count effect can be encountered.10 is a turbulent skin friction drag formula attributed to Schlichting.58 (5.5 Skin Friction Drag Relationships The following empirical flat plate relationships were developed by Ludwig Prandtl and others.The drag coefficient due to skin friction is typically as much as 70 percent of minimum drag coefficient and is a significant factor in the corrections to the drag polar. In Incompressible Aerodynamics (Reference 5.13 through 5. 43 .14) A transition formula between laminar and turbulent is attributed to Prandtl and Gebers and shown in 5. For 10 degrees K off standard day. It is typical that the Reynolds number correction is on the order of 1 drag count (0. Cf = 0.10) C f compressible = C f ⋅ (1 + 0. typically.13.2).000 feet of pressure altitude.65 (5.10 and 5.15. This is also a function of temperature. 5.12) An earlier friction drag equation is one developed by Prandtl and is shown in equation 5.455 (log10 RN ) 2.10 and 5. which cannot be ignored.074 1.2.
5 7.17) .0 5. “Wing Theory” (Reference 5. Jones.0 7.003 0.008 0. a. Subsonic M << 1 (1) Elliptic Wing Theory CL = 2 ⋅π ⋅α 2 1 + AR CDL = CL π ⋅ AR 2 (5.007 Turbulent: Prandtl Turbulent: Schlichting Laminar: Blasius Transistion: PrandtlGebers Skin Friction Drag Coefficient (Cf) 0.2 Skin Friction Drag Relationships 5.3).T.005 0.Empirical Skin Friction Drag Relationships 0. R.001 5. was written by a pioneer in the wing theory field.004 0.0 6.006 0.5 Log 10 (Reynolds Number) Figure 5.6 Idealized Drag Due to Lift Theories The following idealized theoretical drag due to lift models can be found in numerous aeronautical engineering textbooks listed in the Bibliography.16) Transonic M ≈ 1 (1) Slender Body Theory π CL = ⋅ AR ⋅α 2 Supersonic M > 1 44 CDL 2 ⋅ CL = π ⋅ AR 2 (5. One of the best textbooks (in the author’s opinion) titled.5 6.002 0.
5. by John D.a.3 shows two drag models plotted.(1) Thin Wing Theory CL = 4 ⋅α M 2 −1 CDL = α ⋅ CL = M 2 −1 2 ⋅ CL 4 (5. The second drag model represents that parabolic model plus a deviation from the pure parabola. Anderson. For a given Mach number and RN : CD = CD min + K1 ⋅ (CL − CL min ) + K 2 ⋅ (CL − CLb ) 2 2 (5.18) All of the above are idealized and are presented only for general trends.8 . A given point on a Cartesian (xy) plot can be defined by a radius and an angle. 45 .. The first drag model is a pure parabola. Third Edition (Reference 5. The K1 term in the drag polar model above is the pure parabola portion.8 The Terminology ‘Drag Polar’ The terminology ‘drag polar’ was first used by Eiffel. contributes nothing to the model until the lift coefficient exceeds this break lift coefficient. However. That historical note is found in Introduction to Flight. The term ‘polar’ is a reference to polar coordinates. a second source. The K 2 term is zero below a ‘break’ CL and therefore. Figure 5. One idealization made is symmetry (i. This is the same model used in the sample performance model section of this handbook for M = 0.k.4).) 5. a polar plot or a polar diagram). The break lift coefficient could be thought of as the point where flow separation begins and the drag model becomes nonlinear. lists Otto Lilienthal as the ‘inventor’ of the drag polar (a. wing is uncambered and at zero incidence angle.7 Air Force Flight Test Center Drag Model Formulation The following equations are drag model formulations that have been proven at the AFFTC to quite adequately curve fit actual flight test data..e.19) where: K 2 = 0 when CL < CLb .
132 ⋅ (CL − 0. The equation for this specific parabolic model is equation 5.132 ⋅ (CL − 0.30 0.2000 Drag Coefficient (CD) Nonlinear Model Tangent to Curve Figure 5.60) = 0 for CL < 0. 46 .00 0.2642 ⋅ (CL − 0.60 ) (5.3 and 5.4 show. the drag grows substantially after the lift coefficient increases beyond 0.0000 0.22).1800 0.02 + 0.10 0.02 + 0.8 .60 We can plot the ratio of lift to drag.40 0. A second parabola that adds to the first after the start of flow separation has been quite successful in curve fitting AFFTC drag model formulations.0400 0.20) 2 CD = 0.06 ) 2 2 (5.80 0.1600 0.21) (5. This model is a rough approximation to an actual F16A drag polar at M = 0.0200 0.20 and the equation for the nonlinear model is equation 5.23) Figure 5.1400 0.50 0.22) (CL − 0.70 Lift Coefficient (CL) 0. L D = CL CD (5. which is the same as the ratio of lift coefficient to drag coefficient.1200 0.0600 0.3 Drag Polar A secondorder parabola reasonably represents drag polar data only up to the point where flow separation begins.0800 0.90 0.Drag "Polar" 1.00 0.21 (modified by 5.20 0.1000 0.60 Parabolic Model 0.4 presents this lifttodrag versus lift coefficient for both the linear and the nonlinear model.06 ) + 0.6. As Figures 5. CD = 0.
000675 (5. The F16A reference wing area is 300 ft2.50 .297. this model has the following values for drag coefficient: a. the normal load factor can be calculated as follows: 47 . = = 126.890. CD = 0.5.L/D ver sus Lift C oefficient 12 11 10 9 Lift o ver D rag (CL/CD ) 8 7 6 Pa r abo lic M o d e l 5 4 3 2 1 0 0 . 720. a pressure altitude of 30.82 ⋅ 300. The aircraft has an angleofattack limiter.3 0 . Solving for lift and drag from equations 5.000pound aircraft (ignoring thrust component).0 No n lin e a r M o d e l L ift Co e f fic ien t (C L ) Figure 5. 0.297 ⋅ 0.5 0 .6 0 .4 LifttoDrag Ratio versus Lift Coefficient Very roughly.000 feet. Ignore the thrust component in lift and drag coefficient.82 ⋅ 300.25) For our 20.000675 (5.297 ⋅ 0.8 0 . and a Mach number of 0.5 ⋅ 0.24) D= CD ⋅ δ ⋅ M 2 ⋅ S 0.5077 ⋅ 0. At this limit lift coefficient.6: L= CL ⋅ δ ⋅ M 2 ⋅ S 1.000675 0.5 and 5.000 feet is 0. = = 42.4 0 .7 0.0 0 . These are reasonable values.1 0 . CL = 1. Let’s do a sample calculation. and b.5077 . Assume an airplane gross weight of 20.2 0 .80.000 pounds.000675 0. 0.9 1 . maximum thrust stabilized turns occur around 0. The pressure ratio ( δ ) at 30.8 lift coefficient. which corresponds to a lift coefficient of around 1.
07 0. δ ⋅M 2 ⋅S (5. 000. Solving for drag (which is equal to net thrust): D= CD ⋅ δ ⋅ M 2 ⋅ S 0.000675 (5. Robert T. 1990.2 5. ! V = −1.30) SECTION 5.29) Assuming all the negative excess thrust is in deceleration (constant altitude slow down turn): ! Vt = −1.4 48 . − 42.82 ⋅ 300. Jones. 1897. New York: McGrawHill. (5. = 6. Wing Theory. = −24.994.2130.25 = h + t Vt g0 20. ed. 000. CL = 0. 1960. The deceleration rate in knots per second comes to: Nx = ! −24. 1989.5 ⋅ 20.2130 ⋅ 0.297 ⋅ 0.994.25 ⋅ 32.297 ⋅ 0. Twaites. Since thrust equals drag in a sustained (or thrustlimited) turn.28) That would be a longitudinal load factor of greater than a 1 g.890. Bryan. the excess thrust is: Fex = Fn − D = 17. Incompressible Aerodynamics: An Account of the Steady Flow of Incompressible Fluid past Aerofoils.6878 knot ( ) ( ) (5.1 Chanute.5 g’s in maximum afterburner at these conditions.895. 5. New York. the drag coefficient comes to 0. and other Bodies. we can calculate the drag by first calculating the lift coefficient. Dover Publications.21).0 REFERENCES 5. 720. Third Edition.000675 ⋅ N z ⋅ Wt 0.8 knots sec ft sec 1. 720.174 ft sec2 = −23. Anderson.34 g's 20.895.L = N z ⋅ Wt = 126. = = 17.82 ⋅ 300.000675 ⋅ 4. 000. (5. John D.26) From the drag polar equation (5.000675 0. The American Engineer and Railroad Journal.27) At the maximum lift point. Wings. Octave. Princeton University Press. Introduction to Flight. Inc. = = 1. 0. → N z = 126. Progress in Flying Machines. Let’s say that someone told us that the aircraft could sustain 4.3 5.
We will discuss turbine engines and propellerdriven piston engines. so we will simply use a subscripte (e for exit). Figure 6. as well as variable cycle engines. or it can be exhausted separately. Engines that are more exotic include ramjet types. Other turbine engine types include low. The numerical designation of the exit plane varies with the engine complexity. and a low pressure section which drives the fan. In addition.1 Turbine Engine Schematic Vt 0 = Vt = true airspeed (ft/sec) Pt2 = ηr ⋅ Pt 0 (lbs/ft2) total (average) pressure at station 2 where: (6. referred to as bypass airflow. The usual station designation for the engine face is station two. 0 1 2 3 4 5 6 Figure 6.1).and highbypass ratio turbofans. We will start by giving the basic principles of turbine engine thrust.1) ηr = inlet recovery factor (addressed in more detail later). do we actually measure thrust using load cells.1 Introduction We will leave it to numerous other documents to discuss in detail the overall topic of propulsion. an afterburner (additional fuel added after the turbine section) may be added for additional takeoff or maneuvering thrust. where the bypass ratio varies with flight conditions and/or power level. The air flowing through the fan. since inflight thrust is a calculation based upon a number of separate measurements. In this text. Bypass ratio is the ratio of bypass to core airflow. 49 .0 THRUST 6. The term measurement is a misnomer. can be mixed with the core airflow following the turbine.6. A turbofan engine has two separate turbine sections: a high pressure section which drives the compressor. Air enters the engine at the face of the diffuser (Figure 6. the inlet. we are concerned just with the measurement of thrust. Only on the ground. and Pt0 = free stream total (average) pressure (lbs/ft2).1 represents a turbojet engine. either in an engine cell or during a static thrust run.
For turbofan engines an additional pressure times area term must be added to equation 6.2 The Thrust Equation The net propulsive force on the vehicle is called net thrust ( Fn ).5 (6. 6. ( ) Fn = Fg − Fr ! Fg = (Wa + W f ) ⋅ Ve + Ae ⋅ ( Pe − P ) ! Fr = Wa ⋅ Vt where: (6. this is ! generally a small term compared to the Wa + W f ⋅ Ve term. in most cases. however. Pe = pressure (average) across exit plane (lbs/ft2). The engine manufacturer will often provide an engine inflight thrust decka computer program with numerous inputs and outputs on engine performance and operating characteristics.4 below. Ram drag is the result of slowing the air from free stream to near zero speed at the inlet. W f = fuel flow (lbs/sec). Note that the total mass flow into the engine is airflow.3) (6.2 ⋅ M 2 ) ( pounds/ft2) 3. Ve = exit velocity (ft/sec) (average). Note: All of the velocities and pressures are integrated average values. A more precise engine thrust computation would take into account various bleed airs that extract air off the engine for cooling and other purposes. now we will think of it in units of pounds per second to be consistent with the airflow rate. while exiting the engine mass flow is airflow plus fuel flow.Pt0 = P ⋅ (1 + 0. than the ambient pressure. in layman’s terms. The term Ae ⋅ ( Pe − P ) in the equation for gross thrust. is the result of the pressure at the exhaust plane being higher.4 when the fan thrust is exhausted separately.5) ! Wa = airflow rate (lbs/sec) through the engine. Previously defined was the fuel flow ( W f ). 50 .4) (6. The basic thrust equation is gross thrust ( Fg ) minus ram drag ( Fr ). is thrust out the back. and Ae = cross sectional area of the exit nozzle (ft2). However. 6. The gross thrust. The terminology deck is left over from when this computer program was a stack of punched computer cards.2) where: P = ambient pressure (lbs/ft2).
predicts the performance (or status) of the engine usually with flight conditions and throttle position (or power lever angle). temperature. The inputs fall into two categories: a. and in the case where fuel flow is not input. rpm.0 due to shock waves in the inlet. and scrubbing drag. Computing an average total pressure requires several pressure measurements performed all across the inlet. the ηr is typically quite close to 1. Gross thrust will be degraded directly proportional to the reduction of ηr below its maximum value of 1. the deck is a black box with numerous instrumentation measurement inputs. The total pressure varies significantly over the face of the inlet. 6. The airframe manufacturer will add options to the deck to account for installation effects such as inlet spillage drag. At supersonic speeds. These computer programs are developed using proprietary prediction methods supplemented by engine test cell data. The engine rpm’s are the rotation rates of the rotating components. and ambient temperature ( T ). The second problem is components of these inlet rakes may break off in the inlet. we would disturb the flow in the inlet. In a normal shock inlet. The terminology recovery refers to how much of the free stream total pressure the engine inlet is able to recover.3 InFlight Thrust Deck The engine manufacturerprovided inflight thrust deck would vary in complexity. This violates the most fundamental rule of instrumentationdo not affect what you are measuring by the act of measuring it. rotating at different speeds. Engine parameters: fuel flow.4 Status Deck The status deck. From the Rayleigh supersonic Pitot equation: 51 . By measuring the total pressure in the inlet. For the complex augmented turbofans on the F15 and F16 engines.5 Inlet Recovery Factor The inlet recovery factor ( ηr ≤ 1. the status deck also predicts thrust. The recovery factor can be computed using the total pressure formula below. in some cases the status deck could have rpm and fuel flow as inputs and then would become an inflight thrust deck.0 ) is the total pressure loss factor at the engine inlet face.0 ). also fuel flow. A turbojet engine may have just a single rpm. 6. This deck may contain many of the same components as the thrust deck. In addition. Most importantly. Flight conditions: Mach Number ( M ). this recovery factor is about what one would see across an ideal normal shock. temperature. At subsonic conditions ( M < 1.0. First.0 (100percent recovery). then we can compute the recovery factor. A turbofan engine will have more than one turbine section. and fan and compressor rpm. The formula for that is the same as for the normal shock relationship for total pressure measurement in a nose boom. This poses two problems. This pressure variation is called distortion. causing engine damage or failure. pressure altitude ( H C ). airflow bleeds. and fuel flow that are inputs to the thrust deck. In addition. For the performance engineer. b. pressure. the decks are many thousands of lines of computer code plus extensive data table lookups. fuel flow or rotor speed may be input. The status deck will predict the pressure. the inlet recovery factor becomes less than 1.6. or prediction deck. built by Pratt and Whitney and General Electric.
3. in contrast. Pt2 Pt0 (6. which turn the flow through oblique shocks as shown in Figure 6.Pt2/P to 0.6) Pt0 = P ⋅ 1 + 0. has a series of inlet ramps.70 0.5 ⋅ P (7 ⋅ M 2 − 1) The free stream total pressure is just the subsonic formula.6.60 0.5 (6. the actual inlet recovery is modeled quite accurately by the normal shock equation.55 0.65 0. (6.90 0.2 is a plot of this relationship.6 Figure 6.75 0.80 0.7) ηr = Figure 6.6. the recovery factor is the ratio of these two: 3.45 0.2 1.2 is that for Mach numbers above approximately 1.2 ⋅ M 2 Then. The F16 has a normal shock inlet and at speeds above 1. the pressure losses become quite large (greater than 10 percent).2 2 .50 0.8) N o r m al S hoc k R eco ve ry Fac tor 1.85 R eco ve ry Fac tor . 52 .2 Normal Shock Recovery Factor The significance of Figure 6.40 1.0 1.0 2. The F15.00 0.95 0.8 M ach Num be r 2.9216 ⋅ M 2.4 1.4 2 .6 1. 7 Pt2 = 166.
collect data at IDLE. During the summer at Edwards AFB.6 Thrust Runs Checks of installed net thrust can be performed at zero speed using a thrust stand. back to IDLE. Going up and then back down in throttle determines if there is any thrust hysteresis (get a different value if increasing throttle versus decreasing throttle).3 F15 Inlet Schematic The net effect of this oblique shock inlet is that at Mach number = 2.0. MIL and MAX data at three temperatures. After the temperature increases some by late morning. repeat the whole procedure. the recovery factor of the F15 oblique shock inlet is slightly less than that for the F16. The most significant test points would be the fixed throttle points (IDLE. 6. Collect continuous data to observe stabilization times. you should get IDLE. then to MAX. A thrust stand may be as simple as a cable with a load cell. 53 . Get a few data points for the three fixed power points. and repeat at least once. This test most certainly should be performed on all performance test programs. In 1 day of testing. From the measured thrust stand values. The downside is the increased complexity of the inlet producing an increase in aircraft weight.92 versus only 0. since the throttle positions are not distinct and repeatable. it should not be necessary to collect the excessive amounts of data (10+ minutes at one condition would be considered excessive) that some propulsion analysts may desire. This will give you a range of ambient temperatures. start the engine(s).Figure 6. back to MIL. At subsonic speeds. A good test procedure might be to start the tests in the early morning when it is relatively cold. Refuel if necessary.72 for the normal shock inlet. do the process a third time in the afternoon. The suggestion. Intermediate throttle position data points are of less value. the inlet recovery factor is about 0. The thrust stand gives the only direct measurement of installed thrust. inflight thrust is a computation based upon a large number of measurements and a computer model of the engine to predict or estimate the thrust. After collecting that data in early morning. then go to MIL. proceed to shut the aircraft engines down and wait. This is probably due to the losses in turning the flow. since thrust stand time is costly. The importance of these points is that the direct comparison to both the inflight and status decks is possible. Finally. However. In contrast. that range of temperature could be as much as 50 degrees F (see Appendix C for average surface temperatures). one can compare to values of thrust from both the inflight thrust and status decks. MIL and MAX or whatever your fixed throttle points are called). For instance. is to concentrate on getting a number of fixed throttle data points and ignore the intermediate points.
This becomes significant when considering minimum interval takeoffs.8 Propeller Thrust In the examples.4 is an example of some actual throttle transient data taken on the AFFTC thrust stand. but assuming 100 percent thrust at time zero. 6.6. Thrust Lag versus Time 100 90 80 Percent Change in Thrust [MIL to MAX] 70 60 50 40 30 20 10 0 10 0 2 4 6 Elapsed Time (sec) 8 10 12 14 Figure 6. the time to liftoff at one particular condition was 41. was less than 1 percent. In addition. We do not wish to assume that is always the case. Figure 6. The common unit of output power of an engine is horsepower. For the same simulation. However. for instance. This data can be used to develop a thrust dynamics model for use with a takeoff simulation. affecting time to takeoff much more than distance to liftoff. it is unknown how much of the lag is due to lag in the instrumentation.4 Thrust Dynamics from an Air Force Flight Test Center Thrust Stand The thrust stand at the time this data was taken (late 1980s) had a 1 sample per second sample rate. Therefore.1 seconds (or over 5 percent).5 seconds using the simulation. using this thrust stand lag data allowed us to match the actual time to liftoff data very accurately. the time to liftoff was computed to be 39.7 Thrust Dynamics In an engine test cell. for the same liftoff speed. In the English system. it was assumed that thrust was derived from a jet engine. for this aircraft. The change in distance to liftoff. the engine manufacturer will perform throttle transients. To clarify. the effect of the engine lag occurs in the early portion of the takeoff ground roll. The typical aircraft is unable to stabilize at the start of a takeoff with maximum thrust. The equations of motion are just as applicable to an aircraft powered by an engine that drives a propeller. As an example. a throttle transient is necessary to initiate the takeoff. 1 horsepower was defined by James Watt in the 1700s to 54 .
We start with an indicated horsepower ( IHP ).1 The Reciprocating Engine at Altitude For the internal combustion engine. equation 6.29. For takeoff performance. From the definition of horsepower. Then. That produces the shaft horsepower or brake horsepower ( BHP ). One particular set of data gave a value of 1. 6.10) Each propeller manufacturer will usually provide propeller efficiency charts from which one can estimate η as a function of propeller rpm. reduce that by a factor to account for losses to the propeller ( λ ).12) Obviously. for that particular set of data. and flight conditions. THP = η ⋅ ( BHP) (6. As with jet engines. The AFFTC thrust stand is grossly underutilized for this purpose. an engine ‘rating’ will usually not include friction losses and transmission losses to the propeller.000 footpounds of work per minute.14. That factor is the propeller efficiency ( η ). In aircraft applications.13 could be used. pitch.11) (6. Using the density ratio ( σ ) as the density parameter. we will usually divide by 60 to get 550 footpounds of work per second.12 cannot be used at zero speed. the power output for any given engine speed varies with air density (for nonsupercharged engines).80 as a starting point in developing a propulsion model from flight test.13) Richard Von Mises in Theory of Flight suggests that some experimental data indicates that the σ factor would have an exponent ( n ) greater than 1. Then at speeds around liftoff. The result is thrust horsepower ( THP ).8. assume a value like 0. 55 . one can perhaps find similar charts for similar propellers.13 becomes equation 6.9) Then. the thrust horsepower equation as a function of altitude becomes: THP = η ⋅ (σ ⋅ BHP ) (6. If such charts are not available. This factor can be 10 percent or more. which is some fraction (up to maximum power of 100 percent) of the rating. If all else fails.equal 33. BHP = λ ⋅ ( IHP ) (6. there is the fact that the propeller cannot possibly convert 100 percent of the brake horsepower to propulsive force. Then. A thrust model might be just a linear interpolation of the thrust stand value and the liftoff value versus speed. equation 6. the equation for thrust horsepower in terms of thrust and true airspeed is as follows: THP = Fn ⋅ Vt (where Vt has units of feet/sec) 550 Fn = 550 ⋅ THP Vt (6. the static thrust could be measured on a thrust stand. equation 6.
29 = 0.29 e. the altitude degradation factor for this engine is 16. 56 .14) For instance.29 ⋅ BHP ( ) ( ) (6. b.4438 .5328 . σ σ = 0. θ = 0. δ = 0.8625 . Hence.833 . and 1. c.7 percent greater than what would be predicted by a straight density ratio factor.4595 . σ 1. d.THP = η ⋅ σ n ⋅ BHP = η ⋅ σ 1. for an engine at 20.000 feet pressure altitude on a standard day: a. σ = δ θ = 0.
2) Nx = where: (7. One can find other sources that use symbology of N xw and N zw ( w for wind) or N x f and N z f ( f for flight path). Hence. By taking derivatives of airspeed.0 FLIGHT PATH ACCELERATIONS 7. N x and N z are the symbology used to denote flight path axis longitudinal and normal load factor. Vt HE = H + (2 ⋅ g0 ) ! ! T H E = HC ⋅ TSTD Vt ! + ⋅Vt = Ps g0 Ps Vt 2 (7. = true airspeed (feet/sec).7.4) Even if you had zero errors in measured airspeed and altitude. 57 . The GPS yields an alternative method. Note: In this handbook. Fex = N x ⋅ Wt (7. Now. Excess thrust is the amount of the net thrust that is more than the amount needed to achieve equilibrium between net thrust and the drag of the aircraft. = acceleration of gravity (32. it becomes desirable to obtain the aircraft flight path acceleration by some means other than derivatives of true airspeed and pressure altitude. the airspeedaltitude method would have a weakness. You desire to determine the actual physical acceleration of the aircraft. we can compute the excess thrust ( Fex ). and = specific excess power (feet/sec). In addition. with a temperature correction to the pressure altitude.1) (7. many textbooks (including those listed in the Bibliography) will use simply N for flight path normal load factor. = geopotential altitude (feet). respectively. That weakness is the presence of winds.174 feet/sec²). = longitudinal load factor in the flight path (or wind) axis.1 AirspeedAltitude Method The classical method of determining the aircraft flight path acceleration is to differentiate airspeed and altitude using the energy altitude relationship.3) HE H Vt g0 Nx Ps = energy altitude (feet). as developed in the axis systems and equations of motion section. you will invariably have some derivative of wind included.
one can compute the groundspeed in the wind axis. 2 ! V f = Vgf + h2 (7. due again to the wind. the BAA is usually in an instrumentation bay away from the cg. is to perform a cloverleaf maneuver prior to the test maneuver to derive the winds. One could simply use the same equations as for the airspeedaltitude method. The longitudinal axis yields the local longitudinal acceleration and the normal axis the local normal acceleration. These use either the body axis accelerometer (BAA). The BAA uses a set of accelerometers placed somewhere within the body of the aircraft. An alternative to using a heading angle. the INS has been the primary accelerometer method used at the AFFTC. just simply insert the appropriate GPSderived accelerations into the airspeedaltitude equations. ! The groundspeed is the horizontal component of the GPS speed. The accelerometers are then subjected to body axis rates and corrections need to be made to subtract out rate effects. and altitude ( h ). then a flight path groundspeed ( Vgf ) could be computed as follows: Vgf = Vg ⋅ cos(σ g − ψ ) (7.3 Accelerometer Methods (V gN + VwN ) + (VgE + VwE ) 2 2 (7. 7. Corrections need to 58 .2 GPS Method A GPS unit will typically provide groundspeed ( Vg ). practically. The angleofattack vane is connected to a nose boom.6) Then. the above speed is the horizontal component of flight path inertial speed so a transformation is required. This would be appropriate for constant altitude maneuvers such as accels and turns. which may not be an available parameter on some projects. The formula is as follows: Vgf = 7. Nevertheless.5) However. Ideally. or an INS. however. uses the BAA method as its primary method.7.7) There are three different accelerometer methods used to measure flight path acceleration. Once the two components of wind (north and east) are determined.4 Flight Path Accelerometer Method The FPA consists of a twoaxis accelerometer that is aligned with an angleofattack vane. The parameter h is the GPS vertical velocity. If one had the additional parameter of heading angle (and assuming zero sideslip) available. At the time of this writing. NASA Dryden Flight Research Center. the flight path accelerometer (FPA). One catch is the track angle is not the same as the aircraft heading angle (ψ ). track angle ( σ g ). The cloverleaf maneuver is described in the airspeed calibration section. the accelerometers should be at the center of gravity (cg) of the aircraft.
Vt = true airspeed.8) (7.13) (7.1 represents an FPA unit (designated an NBIU [Nose Boom Instrumentation Unit]) developed at the AFFTC in the late 1960s.be made to the accelerations for not being at the cg (rate effects) and for being connected to an angleofattack vane that is not indicating the true angle of attack.11) (7. N xi = indicated longitudinal load factor.14) (7.12) (7.9) (7.15) q = pitch rate. LV = distance from accelerometer to aircraft cg (positive with the accelerometer forward of the aircraft cg). and N zi = indicated normal load factor. Figure 7. The flight path accelerometer correction equations (ignoring roll and yaw terms) are as follows: ! N x = N xi ⋅ cos(∆α ) − N zi ⋅ sin(∆α ) + LV / g 0 ⋅ q 2 ⋅ cos(α t ) − q ⋅ sin(αt ) ! N z = N zi ⋅ cos(∆α ) + N xi ⋅ sin(∆α ) + LV / g 0 ⋅ q 2 ⋅ sin(α t ) − q ⋅ cos(α t ) (7.10) α t = α i + ∆α + ∆α bb α i = measured angle of attack ∆α = ∆α q + ∆αu + ∆α lag LV ⋅ q ∆α q = tan −1 = pitch rate correction (Vt − LV ⋅ q ⋅ sin(α t ) ) ∆α u = upwash correction ∆α bb = boom bending correction ∆α lag = lag correction where: (7. 59 .
The data point was a stabilized cruise point.3 represent indicated longitudinal load factor ( N xi ) 60 .1 Air Force Flight Test Center Nose Boom Instrumentation Unit This unit is installed on the AFFTC F15B Pacer (at the time of this writing). The noise is from structural vibration at relatively high frequencies and lower frequency flight dynamic oscillations. No matter where one locates an accelerometer in the aircraft.Figure 7. Similar units are still being used for flight test in the late 1990s.2 and 7. Figures 7. we must deal with the noise in that data. it will be subject to substantial quantities of noise. Figure 7.5 Accelerometer Noise When we use an accelerometer to measure flight path accelerations. 7.2 is an example of some actual data from the first flight of the B1A in December 1974.
Mean = 0. On the B1A. It is called low pass because it passes low frequencies. This filter was a 4pole 30 Hz (cycles per second).0 0. The data were sampled at 64 samples per second. The 30 Hz is the cutoff frequency of the filter. This meant that our performance data had a substantial amount of longitudinal vibration data in it. a.3 0.6 0.7 0.2 1 0.and normal load factor ( N zi ).4 0.4 Indicated Normal Load Factor (Nz) 1. After the plots is a discussion of the characteristics of this filter. lowpass Butterworth filter.00831 b.2 Longitudinal Load Factor – Unfiltered Data The mean and standard deviation (sigma) of N xi are as follows for 58 data points.8 0. B1A First Flight Data: Flightpath Accelerometer: Indicated Nz 1.5 Time (sec) 0.6 1.01682 61 . In this case.4 0.6 0.2 0. The accelerometers were located in an AFFTC NBIU. the lowest longitudinal vibration modes were less than 10 Hz.0 Figure 7. The analog output of the accelerometers was filtered.8 0. the cutoff frequency was too high.1 0.9 1. Sigma = 0.
0047 ( ) 62 .3 Normal Load Factor – Unfiltered Data The mean and standard deviation for the N zi is as follows for the same 58 time slices: a. If N x was zero for this stabilized cruise point. then the above equation can be used to solve for upwash.5 Time (sec) 0.020 0.9 1.010 0.17) For this one data sample.6 0.3 0.1 0. Mean = 1.030 0.00831 = 0.010 0.8 0.030 0.020 0.2 0.0 0.16) ∆α = upwash angle.7 0.040 Indicated Longitudinal Load Factor (Nx) 0.0047 b.040 0. Sigma = 0.47 deg 1.000 0.060 0.B1A First Flight Data: Flightpath Accelerometer: Indicated Nx 0.2257 Ignoring pitch rate terms. the transformation equation for true flight path longitudinal load factor ( N x ) is as follows: N x = N xi ⋅ cos ∆α − N zi ⋅ sin ∆α where: (7.050 0.4 0. ∆α = tan −1 N xi N zi (7.0 Figure 7. the ∆α computes to be: ∆α = tan −1 0.
Figure 7.The attenuation of a filter is expressed in terms of decibel (dB).5 4. Figure 7.0 4.0 2. FourPole Butterworth LowPass Filter Attenuation 60 50 Transmission Loss . which is an output over input of 0.0 Frequency Ratio (frequency/cutoff frequency) Figure 7.5 5. By definition. The time delay is defined in terms of a parameter called the group time delay ( tdgroup ).5 3.0 0.4 FourPole Butterworth Filter Attenuation Characteristics At the time.708 or an attenuation of almost 30 percent.4 shows the attenuation for a fourpole Butterworth filter.0 3.Negative Decibels 40 30 20 10 0 0. The actual time delay ( ∆t ) is determined as follows: tdgroup ∆t = 2 ⋅π ⋅ fc (7. and = input. For this filter. the solution to the noise problem with B1A flight path accelerometer data was to change to filters with a much lower cutoff frequency.5 1. The definition of decibel is as follows: dB = −20 ⋅ log10 where: E0 Ei (7. The problem with that solution was that a filter with a low cutoff frequency also introduced substantial phase (time) lag.5 2. the cutoff frequency is at a dB = 3.5 represents the time lag function versus the frequency ratio.0 1.19) 63 .18) E0 Ei = output.0 .
5 2. The second was a moving average. To illustrate this. A span of 21 data points was chosen which would include the midpoint and 10 points on each side of the midvalue.5 3.0 was selected to avoid the very low frequency firstbody bending modes of this very flexible aircraft. the group time delay is 2. ∆t = 0.0 3.60 seconds. Since no dynamic performance maneuvers were performed on the B1A.0 Group time delay (sec) 2.2 second can be a source of significant errors for highly dynamic maneuvers such as the roller coaster. The first was a moving secondorder polynomial curve fit.5 4.014 sec for f c = 30 Hz b. A filter with a cutoff frequency of 2.1 times the cutoff frequency.5 1.0 1.0 Hz A time lag of 0.where: f c is the cutoff frequency in Hz. The actual time delays for the 30 and 2. These are shown in Figure 7. it would be more desirable to digitally filter the data. ∆t = 0.0 Hz filters compute to the following using the above equation.207 sec for f c = 2. this was not deemed a problem.0 0. To avoid a time shift error in accelerometer data.6.5 0. 64 . FourPole Butterworth LowPass Filter Group Time Delay 4.0 0 1 2 3 4 5 6 Frequency ratio (frequency/cutoff frequency) Figure 7.5 FourPole Butterworth Filter Group Time Delay At maneuver frequencies less than 0. a. the N xi was digitally filtered with two different methods.
016 0.7 plots the moving secondorder polynomial fit points.7 0.7 0.4 Time (sec) 0.1 0.006 0.010 0.1 0.004 0.2 0.000 0 0. (Moving Polynomial) Figure 7.5 0. Indicated Longitudinal Load Factor 0.6 Longitudinal Load Factor – Filtered Data Figure 7.4 Time (sec) 0.000 0.004 0.008 0.016 0.3 0. A thirdorder polynomial curve fit of the time history is also shown.8 Moving Polynomial Poly.8 Figure 7.014 Indicated Longitudinal Load Factor (N i) 0.7 ThirdOrder Polynomial Fit of Filtered Longitudinal Load Factor Data 65 .002 0.008 0.012 0.6 0.2 0.5 0.014 Longitudinal Load Factor (Nxi) 0.002 0.3 0.6 0.Indicated Nx data: Digitally Filtered: 21 Point Span 0.010 0.006 0.0 Second Order Polynomial Moving Average 0.012 0.
00140 Mean 1Sigma The average value of each of the three methods was identical to three digits (1 millig). The Euler angles are pitch.00831 0. Heading (rotate about the z axis [or yaw]) (transform through ψ ) 66 . That is because for any maneuver where variation in acceleration is not linear. VtE . ψ = true heading angle. The mathematics below will take you through the process to compute winds. then the transformations into the wind axis are performed.1 SUMMARY OF STATISTICS FOR LONGITUDINAL LOAD FACTOR Original Data 0. Once the winds are known. Define: a. d. 7. Although (for this data set) the simple moving average produced the greatest reduction in standard deviation. 7. b. θ = pitch attitude.θ . the INS outputs will be velocities and accelerations in the north. β = sideslip angle. The transformation must be performed in the exact order of β . Beta and True Airspeed The following matrices are used to transform the true airspeed from the flight path axis ( Vt ) to the earth axis ( VtN . roll. The two digital filtering methods reduced the standard deviation by about a factor of 10. and down direction and a set of angles called Euler angles. α . and true heading. Typically. φ = roll attitude.1 summarizes the mean values and 1sigma deviations from the mean for the different sets of data. it is preferable to use the moving secondorder polynomial fit.00233 SecongOrder Moving Minus ThirdOrder Fit 0 0.00853 0. and VtD ). c. φ .01682 Moving Average 0.6 Inertial Measurement Method The INS method involves transforming the earth axis inertial parameters of the INS into the aircraft wind (or flight path) axis. Table 7. the parabola will match the variation more accurately.7 Calculating Alpha. east. and e.ψ .00115 SecondOrder Polynomial Moving 0.Table 7. α = angle of attack.00848 0.
23) cos β [ β ] = sin β 0 − sin β cos β 0 0 0 1 (7.24) The matrix summary form of the transformation from the flight path axis true airspeed to the true airspeed in the earth axis ( N .22) cos α [α ] = 0 sin α Sideslip angle (transform through β ) 0 − sin α 1 0 0 cos α (7.25) 67 .cosψ [ψ ] = sin ψ 0 − sinψ cosψ 0 0 0 1 (7.20) Pitch (rotate about yaxis) (transform through θ ) cosθ [θ ] = 0 − sin θ 0 sin θ 1 0 0 cosθ (7. E . D ) is as follows: (VgN + VwN ) Vt (VgE + VwE ) = [ψ ] ⋅ [θ ] ⋅ [φ ] ⋅ [α ] ⋅ [ β ] ⋅ 0 0 (VgD + VwD ) From equation 7. (7.21) Roll (rotate about xaxis) (transform through φ ) 0 1 0 cos φ [φ ] = 0 sin φ Angle of attack (transform through α ) 0 − sin φ cos φ (7.25 we can solve for the winds.
In this case.31 using the winds determined during the wind calibration (equation 7.33) 68 . To take the transpose of these unique matrices reverse all the offdiagonal terms and keep all the diagonal terms the same. the [ β ] matrix derives from equation 7. reverse the order of the matrix multiplication and take the transpose of each individual matrix. During a typical wind calibration.28) (7. The five unknowns are the three components of wind (VwN . the transpose is the same as the inverse. and the bank angle (φ) are equal to zero.26 represents three equations with at least five unknowns. Equation 7.VwN Vt VgN VwE = [ψ ] ⋅ [θ ] ⋅ [φ ] ⋅ [α ] ⋅ [ β ] 0 − VgE V 0 V wD gD (7.32) (7.29) We now wish to perform the reverse transformation. For instance.26) The equation above is the general matrix formula. To transform the components. we will assume the vertical wind ( VwD ). VwE and VwD ) and α and β .27) (7.24 as follows: T [β ] T cos β = sin β 0 − sin β cos β 0 0 cos β = − sin β 0 0 1 T sin β cos β 0 0 0 1 (7. Then the α calculation reduces to the following: α =θ −γ ! γ = sin −1 h V = flight path angle t ! h = −VgD = rate of climb (7. the sideslip angle ( β ). that is.31) We can calculate all the velocities in the equation 7. to transform the components of true airspeed in the earth axis to the flight path.26) as follows: VtN = VgN + VwN VtE = VgE + VwE (7.30) The matrix formula is as follows: [ β ] ⋅ [α ] ⋅ [φ ] ⋅ [θ ] ⋅ [ψ ] T T T T T VtN Vt ⋅ VtE = 0 V 0 tD (7.
the angles were derived by geometry without the following matrix mathematics: cos β ⋅ cos α ⋅Vbx + sin β ⋅ Vby + cos β ⋅ sin α ⋅ Vbz = Vt − sin β ⋅ cos α ⋅Vbx + cos β ⋅ Vby − sin β ⋅ sin α ⋅ Vbz = 0 − sin α ⋅Vbx + cos α ⋅ Vbz = 0 Equation 7.36) Next.42 yields a formula for angle of attack.42) 69 . z ) are calculated in the following matrix manner: Vbx VtN T T T Vby = [φ ] ⋅ [θ ] ⋅ [ψ ] ⋅ VtE V V bz tD (7. these formulas should be the same as presented earlier.39) Multiplying out the above matrix yields three equations from which we will derive formulas for α and β . In the axis systems and equations of motion section. (7. When complete. transform the body axis to the flight path axis through angle of attack and sideslip angle as follows: [ β ] ⋅ [α ] T T Vbx Vt ⋅ Vby = 0 V 0 bz (7. y.38) cos β ⋅ cos α − sin β ⋅ cos α − sin α cos β ⋅ sin α Vbx Vt − sin β ⋅ sin α ⋅ Vby = 0 Vbz 0 cos α (7.41) (7.37) Expanding the alpha and beta transpose matrices and writing them out: cos β − sin β 0 sin β cos β 0 0 cos α 0 ⋅ 0 1 − sin α sin β cos β 0 0 sin α Vbx Vt 1 0 ⋅ Vby = 0 0 cos α Vbz 0 (7.40) (7.VtD = VgD + VwD Vt = (7.35) The airspeed components in the body axis ( x.34) 2 (V 2 tN + VtE + VtD 2 ) (7.
47 into equation 7. substituting equation 7. To transform the components.11 and 2. (7.42 to substitute for the term in the square brackets.48) We now wish to perform the reverse transformation.44 and 7. that is.42.40: (7.41 and 7.46: cos β sin 2 β cos β ⋅ ⋅ Vby + ⋅ Vby = Vt sin β sin β Vby = Vt sin β V β = sin −1 by V t Compare equations 7. we will use equations 7.44 into equation 7. to transform the components of true airspeed in the Earth axis to the flight path.43) α = tan −1 Vbz V bx Inserting the result for Vbx from equation 7. Replace Vbx in 7.45) cos 2 α sin 2 α cos β ⋅ ⋅ Vbz + sin β ⋅ Vby + cos β ⋅ ⋅ Vbz = Vt sin α sin α Collecting terms and using the trigonometric identity sin 2 α + cos 2 α = 1 : V cos β ⋅ bz + sin β ⋅ Vby = Vt sin α (7.46) Now.48 to equations 2.47) Finally. − sin β ⋅ cos 2 α sin 2 α ⋅ Vbz + cos β ⋅ Vby − sin β ⋅ ⋅ Vbz = 0 sin α sin α (cos 2 α + sin 2 α ) − sin β ⋅ ⋅ Vbz + cos β ⋅ Vby = 0 sin α Vbz cos β sin α = sin β ⋅ Vby (7. reverse the 70 .41 using 7.44) Vbx = cos α ⋅ Vbz sin α (7.sin α / cos α = tan α = Vbz Vbx (7.12.
z ) are calculated in the following matrix manner: Vbx VtN T T T Vby = [φ ] ⋅ [θ ] ⋅ [ψ ] ⋅ VtE V V bz tD (7.49) We can readily solve for the true airspeed components from the above.53) The velocities in the equations 7. the noise will be dramatically reduced. 7.51) (7. The matrix formula is as follows: [ β ] ⋅ [α ] ⋅ [φ ] ⋅ [θ ] ⋅ [ψ ] T T T T T VtN Vt ⋅ VtE = 0 V 0 tD (7. y. it is desirable to compute the accelerations by taking numerical derivatives of the inertial velocities.8 Flight Path Accelerations To compute the accelerations in the flight path requires first computing the accelerations in the N. we are 71 . angle of attack and sideslip are computed using equations 7.52) VgD (t + ∆t ) − VgD (t − ∆t ) (7. Note that the down (or z ) component involves subtracting out a gravity term. not the airspeeds. This is because the accelerations are sensing the high frequency vibrations of the aircraft and are usually quite noisy. The airspeed components in the body axis ( x. but the components in our wind axis system. The typical INS updates at 50 samples per second. The acceleration formulas are as follows: AN (t ) = AE (t ) = AD (t ) = VgN (t + ∆t ) − VgN (t − ∆t ) 2 ⋅ ∆t VgE (t + ∆t ) − VgE (t − ∆t ) 2 ⋅ ∆t 2 ⋅ ∆t − g0 (7.44 and 7.48. Even when the accelerations are available as a direct output of an INS. They are still inertial accelerations.53 are the inertial (or ground) speeds. E.order of the matrix multiplication and take the transpose of each individual matrix. However. E.50) From true airspeed and the body axis true airspeed components.51 through 7. Since the vertical component of acceleration is down. and D axis. The α and β are required in order to transform the earth axis accelerations to the flight path axis. If one simply samples the velocity data at no more than about 5 samples per second and then takes a derivative. and D axis. we will later transform these into the wind axis. We are computing inertial accelerations in the N.
i designation. (7. The transformation matrix formulation for accelerations is identical to that for velocities and is given below. b.actually adding in a gravity term. in conventional performance the standard sea level value of g ( g 0 = 32. y distance ( l y ) is positive out the right wing. at 5 samples per second. pitch rate. note that f designation is dropped for the flight path axis load factors.20 seconds. q = pitch rate ( rotation about y − axis ) ( + pitch up ) . If the noncorrected body axis accelerations are designated with a sub. Axf AN T T T T T Ayf = [ β ] ⋅ [α ] ⋅ [φ ] ⋅ [θ ] ⋅ [ψ ] ⋅ AE A A D zf (7.174 feet/sec2) is usually used. then the matrix correction equations are as follows: 72 . we normally work with load factors (acceleration over g) rather than the accelerations. the ∆t would be 0. and yaw rate in the body axis. and the z distance ( lz ) is positive down. The x distance ( l x ) is positive forward. Assume we have rate gyros. In addition. l y and lz from the cg of the aircraft. Assume that the accelerometers are at distances lx . N x Axf g0 N y = Ayf g0 N − A g z zf 0 Finally. r = yaw rate (rotation about z − axis) ( + nose right ) . For instance. and c.54) In performance. p = roll rate ( rotation about x − axis ) ( + right wing down ) . However. which give us roll rate. we will put the flight path accelerations on the left side of the equation. There is also a sign change on the normal load factor to account for the positive normal load factor convention. Define these as follows: a.55) 7.9 Accelerometer Rate Corrections The following corrections to accelerometers are presented without derivation.
12 Euler Angle Diagram (7. Then. the aircraft is rotated (positive right wing down) through the roll angle ( φ ). Then. q2 + r 2 Axb Axbi ! Ayb = Aybi + − ( r + p ⋅ q ) A A zb zbi − ( q − r ⋅ p ) ! ( ! ) (r! − p ⋅ q ) − (q + p ⋅ r ) l ! ( p + r ) ( p − q ⋅ r ) ⋅ l ! ( p + q ⋅ r ) − ( q + p ) l x 2 2 y 2 2 z (7. and r r − q l x 0 p ⋅ l y − p 0 l z (7. ! ! p = φ −ψ ⋅ sin θ ! ! q = θ ⋅ cos φ + ψ ⋅ cos θ ⋅ sin φ ! ! r = ψ ⋅ cosθ ⋅ cos φ − θ ⋅ sin φ 7. Vbx Vbxi 0 Vby = Vbyi + −r V V q bz bzi 7. in that order. It is critical that the order of rotation is just as described (ψ . then take numerical derivatives to compute accelerations.θ . the aircraft is pitched (positive up) through the pitch attitude ( θ ). θ and φ .8 illustrates the Euler angles. one would not rate correct the resultant accelerations. Starting with the aircraft heading north.θ . the i designation will be noncorrected velocities. q.10 Velocity Rate Corrections Rate corrections to the body axis velocities in the matrix format are presented in equation 7. φ ). 7. otherwise. 73 . a transformation is performed (positive east) through the heading angle (ψ ). φ ) are given.60) Figure 7.57) In the case where the Euler angles (ψ . we can compute the body axis rates using the following formulas. This Euler angle diagram pictorially illustrates the order of transformation.58) (7. one would get a different result.59) (7. N zb = − Azb g0 This author prefers to rate correct the velocities. Finally. Again. These will have been accomplished by axis transformations through ψ .11 Calculating p.57.56) Note: A sign change when computing normal load factor. a.
For this handbook. conventional aircraft would be any aircraft with a main gear. Dt = drag of the aircraft tail .acts along the aircraft flight path (this term is often lumped into the body drag for aircraft without a Ttail). Dbw = drag of the aircraft body and wing .acts through the center of gravity of the aircraft.acts perpendicular to the flight path.2 Takeoff Parameters Let us define the following forces. e. 74 .1 General This section will present the theory of takeoff and landing for conventional aircraft. During the ground roll.along the aircraft flight path axis. ‘conventional’ could include some aircraft that are considered STOL (Short TakeOff and Landing). a. Therefore. the flight path will be parallel to the runway. a nose gear.also acts perpendicular to the flight path. Wt = gross weight .1. (Not shown on the drawing [to avoid clutter] are gross thrust [ Fg ] and the engine inlet [or propulsive] drag [ Fe ]). distances. c. L2 = lift of the tail . d. and a single source of thrust at some angle of incidence it .0 TAKEOFF 8. 8. angles and coefficients as depicted in Figure 8.8 Euler Angles 8. One could derive equations that are more complex for a VSTOL (Vertical or Short TakeOff and Landing).Figure 7. L1 = lift of the wing . b.
o. j. XL2 = distance from the wing lift point to the tail lift action point. g.f. Requiring the summation of forces in the Xaxis to be zero: Fg ⋅ cos(θ + it ) − Fe = D + Frw + Fex (8. p.1) 75 . X 1 = distance from the nose gear to the aircraft center of gravity. θ = aircraft pitch attitude (angle between Xbody axis and horizontal). Z 2 = height of the tail center of lift and drag above the aircraft body axis. Z1 = height of the body axis of the aircraft above the ground plane. θ rw = runway slope. F1 = load on the nose gear (perpendicular to the runway). k. h. we can formulate the equations of motion for the aircraft during the ground roll. Fn = net thrust acting parallel to the flight path. Figure 8. n. X 2 = distance from the main gear to the aircraft center of gravity. i.1 Takeoff and Landing Forces and Angles Using the above diagram. l. F2 = load on the main gear (perpendicular to the runway). The equations are the same for either a takeoff or a landing. XL1 = distance from the center of gravity to action point of the wing lift (aerodynamic center of the MAC [Mean Aerodynamic Chord]). m.
and = coefficient of friction associated with the main wheels.6) Vg = groundspeed. we will account for the gravity component in the Wt ⋅ sin(θ rw ) term.3) µ1 µ2 = coefficient of friction associated with the nosewheels. N x = Ax / g0 ! Ax = Vg where: (8.5) (8. 76 . Ignore any pitch dynamics during the ground roll or any moment caused by the vertical component of gross thrust. Take moments about the main wheels.8) Require the summation of moments about the Yaxis to be zero. since the aircraft will pitch about the main wheels during the takeoff or landing ground roll. and Fex = excess thrust (positive forward).2) (8. Frw = total runway resistance = runway friction plus runway slope effect. the gravity component is included. Collecting terms: Fg ⋅ cos(θ + it ) − Fe = ( Dbw + Dt ) + ( µ1 ⋅ F1 + µ 2 ⋅ F2 + Wt ⋅ sin(θ rw )) + Fex Requiring the summation of forces in the Zaxis to be zero: (8. In the air. D = Dbw + Dt Frw = µ1 ⋅ F1 + µ 2 ⋅ F2 + Wt ⋅ sin(θ rw ) where: (8. On the ground. Note that the longitudinal load factor definition on the ground includes only the velocity derivative term. Fex = N x ⋅ Wt (positive forward) where: (8.4) N x = longitudinal load factor.where: D = total aerodynamic drag.7) L1 + L2 + F1 + F2 = Wt ⋅ cos(θ rw ) (8.
drag.F1 ⋅ ( X 1 + X 2 ) + L1 ⋅ ( X 2 − XL1 ) + Dbw ⋅ Z1 + Dt ⋅ ( Z1 + Z 2 ) + Wt ⋅ sin(θ rw ) ⋅ Z1 = Wt ⋅ cos(θ rw ) ⋅ X 2 + ( Fg ⋅ cos(it ) − Fe ) ⋅ Z1 + L2 ⋅ ( XL1 + XL2 − X 2 ) (8.12 in three by threematrix form as follows: 1 0 0 µ1 µ 2 Fex A1 1 1 ⋅ F1 = A2 ( X 1 + X 2 ) 0 F2 A3 (8. Once we have the excess thrust. Then we can rewrite equations 8. Collecting equations 8. However.12 as A1 .10 through 8.11) Wt ⋅ cos(θ rw ) ⋅ X 2 − Wt ⋅ sin(θ rw ) ⋅ Z1 + ( Fg ⋅ cos(θ + it ) − Fe ) ⋅ Z1 + L2 ⋅ ( XL1 + XL2 − X 2 ) (8.13) 77 . the lift and drag models may not be for inground effect. we can integrate the groundspeed derivative to obtain speed and distance versus time.12) − L1 ⋅ ( X 2 − XL1 ) − Dt ⋅ ( Z1 + Z 2 ) ( X1 + X 2 ) ⋅ F1 = We will define the terms in the square brackets in 8. If no inground effect corrections are available.9: Fg ⋅ cos(θ + it ) − Fe = Dbw + Dt + µ1 ⋅ F1 + µ 2 ⋅ F2 + Wt ⋅ sin(θ rw ) + Fex F1 ⋅ ( X 1 + X 2 ) + L1 ⋅ ( X 2 − XL1 ) + Dt ⋅ ( Z1 + Z 2 ) + Wt ⋅ sin (θ rw ) ⋅ Z1 = Wt ⋅ cos(θ rw ) ⋅ X 2 + ( Fg ⋅ cos(it ) − Fe ) ⋅ Z1 + L2 ⋅ ( XL1 + XL2 − X 2 ) L1 + L2 + F1 + F2 = Wt ⋅ cos(θ rw ) Rearranging the equations: Fex + µ1 ⋅ F1 + µ 2 ⋅ F2 = Fg ⋅ cos(it ) − Fe − Dbw − Dt − Wt ⋅ sin(θ rw ) F1 + F2 = [Wt ⋅ cos(θ rw ) − L1 − L2 ] (8. gross thrust.10) (8.10 through 8. A2 . The primary parameter of interest is the excess thrust from which we can compute the derivative of groundspeed.7 through 8.9) What we now have is three equations with three unknowns for purposes of simulating a takeoff or landing ground roll. and propulsive drag terms in the above equations. The three unknowns are the two normal forces on the wheels ( F1 and F2 ) and the excess thrust ( Fex ). and A3 . then some empirical predictions can be used until flight test results are available to create an inground effect model. It is assumed that one has a thrust and drag model for the lift.
15) (8. we make the following approximation: Fn ≅ ( Fg − Fe ) ⋅ cos(θ + it ) Fex + µ ⋅ F = Fn − D − Wt ⋅ sin(θ rw ) F = Wt ⋅ cos(θ rw ) − L where: (8. However. the contractor will provide an initial estimated model for lift and drag as a function of angle of attack ( α ). a point mass model will be assumed with all the forces acting through the cg of the aircraft. During the ground roll.16) (8.025 are also used.17) F = main gear load (assume all load on the main gear).14) 8. The thrust incidence angle is usually zero or small.3 Developing a Takeoff Simulation Usually.015 is usually assumed for the rolling coefficient of friction ( µ ). the angle of attack is equal to the pitch attitude ( α = θ ). or at least not known precisely. so we can ignore Dt in many cases. In addition.18) 78 . Therefore.13 can be solved by multiplying both sides by the inverse of the square matrix.17: Fex + µ ⋅ (Wt ⋅ cos(θ rw ) − L ) = Fn − D − Wt ⋅ sin(θ rw ) (8. Values of µ for a dry runway up to 0. Accounting for tail lift and drag becomes more important for modeling braking performance to determine the load distribution on the main gear and the nose gear.During the course of flight test. we measure excess thrust ( Fex ). Only the most precise simulations will typically account for a separate tail and body drag. Fex 1 F1 = 0 F 0 2 µ1 µ2 A1 1 1 ⋅ A2 ( X1 + X 2 ) 0 A3 −1 (8. Further. The A1 term is thrust minus drag minus the runway component of weight. since Fg >> Fe at low airspeeds. The matrix relationship in equation 8. we may need to iterate between the above equation and the solution of the above equation. one may need to supplement this model with empirical ground effect estimation. Combining equations 8. For takeoff performance. a value of 0.16 and 8. As mentioned before. the thrust and drag may be unknown. such as that found in the NASA takeoff and landing simulation program listed in the Bibliography.
H C . The suspicion is that this is from some early NACA work. They may also include Reynolds number (or skin friction drag) terms in the drag polar. One would also incorporate a fuel flow model to compute fuel used during takeoff. Smith and Technical Aerodynamics by Karl D. First. hAGL ) where: (8.2 is typical of a relationship defining the decrease in drag due to lift inground effect. In addition. We know (or assume values for) the other variables: gross weight. and = aircraft wing height above ground level. runway slope. Second.21) (8. T ) CL = f (α .C.Equation 8.20) [ Fn − D] = Fex + Wt ⋅ sin(θ rw ) + µ ⋅ (Wt ⋅ cos(θ rw ) − L ) From equation 8. we can compute the excess thrust during the ground roll of the aircraft. Wood. the engine is usually not at 100percent thrust at brake release so a thrust spool up factor needs to be supplied. The drag and lift models would be in the form of drag and lift coefficients versus angle of attack. to solve thrust minus drag (equation 8. Fex = [ Fn − D ] − Wt ⋅ sin(θ rw ) − µ ⋅ (Wt ⋅ cos(θ rw ) − L ) (8.18 can be used in two ways. The texts are The Illustrated Guide to Aerodynamics by H. One would be provided models for net thrust drag and lift.20).23) M HC T hAGL = Mach number. rolling friction. The equation is a curve fit of the points.22) (8. 79 . = pressure altitude (subscript C denotes calibrated). neither of which gave a source for the data. 8.19) (8. and aerodynamic lift. The parameter hAGL is needed to account for ground effect. This is to account for the fuel used for mission calculations. Typical model formulations are as follows: Fn = f ( M . = ambient temperature. The data points were taken from a curve found in two separate textbooks.19).19. hAGL ) CD = f (CL . The above are just typical model forms. to solve excess thrust (equation 8.4 Ground Effect Figure 8.
d.0 = height of wing above ground while aircraft on the ground (feet). Fnslope = 5. h. The model constants and equations are as follows: a. K Fn = 1 − K Fno ⋅ e − t /τ ( ) (8. b = 35 = wing span (feet). = thrust at zero Mach number (pounds).0 = thrust time constant (seconds). c. hw = 5. K Fno = 0.24) 80 .Figure 8. AR = 4. b. i. S = 300 = reference wing area (feet2). = start gross weight (pounds).0 = b 2 / S = aspect ratio. f.65 = thrust factor at zero time. g.000. τ = 2.000.2 Predicted Ground Effect Drag A very simplified model that approximates an F16 aircraft in military thrust was created to illustrate takeoff simulation. Wts = 25. e. Fno = 10.000 = slope of thrust versus Mach number (pounds).
For our model.0 (8. if X GE > 1.Thrust runs can be used to determine this thrust spool up factor. P δ = PSL where: (8.29) where: H C = 2.27) Drag coefficient ( CD ) is computed as follows: 2 1 ⋅ C − CL min ) CD = CD min + X GE (π ⋅ AR ) ( L (8.0500 = minimum drag coefficient.87559 E − 6 ⋅ H C ) 5.29 /100 b X GE = 1. at time = zero.2559 (8.28) where: CD min = 0. the thrust is 35 percent of zero Mach number thrust and increases exponentially with a 2. It may not be a simple exponential function as we are using here.05 = lift coefficient corresponding to minimum drag.2 was performed to produce an equation for ground effect.30) P = ambient pressure. ( h + hw ) X GE = 24. and CL min = 0. A curve fit of the data points in Figure 8. Ambient pressure ratio ( δ ) is as follows (formula derived in the altitude section): δ = (1 − 6.300 feet = initial pressure altitude.0 .25) (8.0 second time constant. Then the equation for the net thrust for this model becomes: Fn = K Fn ⋅ ( Fno + Fnslope ⋅ M ) W f = tsfc ⋅ Fn where: (8.12 ⋅ Ln + 108. and 81 .26) tsfc = thrust specific fuel consumption.
The formula is as follows: a.35 0.22 lbs/ft2.20.5 1. and later use flight test derived models.1 1.55 0.0 0.3.3 1.30 In both cases. The above lift and drag models are idealizations presented to illustrate general trends only.3 Lift Ratio InGround Effect The equation corresponding to the above curve is as follows: 82 . In a flight test project.4 CL(IGE)/CL(OGE) 1. Then. That yields Figure 8.6 Wing height above ground/wing span (h+hw)/b Figure 8.31) As with the drag coefficient. A lift coefficient factor inground effect was determined on two separate flight test projectsa fighter and a transportat the AFFTC. Lift coefficient ( CL ) is as follows (from elliptic wing theory): π ⋅ AR ⋅α CL = CL 0 + 1+ 2 AR (8.4 0.5 0.05 (5 percent). the ground effect factor at liftoff was about 30 percent. CL ( IGE ) CL (OGE ) = 1. In both cases.2 0.25 0.15 0.PSL = ambient pressure at standard day sea level = 2116.2 1. the lift ratio decreased to 1. the wing height to span ( h / b ) is about 0. an adjustment for ground effect needs to be applied to the lift coefficient. further assume that the relationship is base 10 logarithmic.1 0. Let us assume that by the time h / b increased to 0. one would initially use wind tunnel data.45 0.5 (half span).3 0. Lift Curve Ground Effect Factor 1.
0 Airspeed. Lift and drag (formulas in lift and drag section) are computed as follows: L = CL ⋅ δ ⋅ M 2 ⋅ S / 0. the last terms in our model are the runway resistance.35) Finally. µ = 0.0 deg/sec. The aircraft (or the simulated aircraft) will lift off the ground when lift is greater than weight. The lift first exceeds weight at an airspeed of 132 knots.CL ( IGE ) CL (OGE ) = 0. The angle of attack ( α ) is limited to a predetermined value. Calibrated airspeed is normally displayed in the cockpit and was discussed in detail in Section 4. We will assume zero runway slope. As will be shown in the later vectored thrust takeoff section. Frw = µ ⋅ (Wt − L ) Frw = 0.0 The angle of attack is held to zero during the ground roll until a rotation speed is reached.6282 ⋅ Log10 (( h + hw ) b ) (8.0 if L > Wt Combining terms: 83 (8. This rotation speed (in this simulation example) is at a calibrated airspeed of 100 knots. In this example simulation that value is 13 degrees.000675 (8.000675 D = CD ⋅ δ ⋅ M 2 ⋅ S / 0. The angleofattack profile used in this example computer simulation is as follows: α = α last + ∆α ∆t ⋅ ∆t where: ( ) (8.8609 − 0.32) With the following constraint: a. Then. the selection of 100 knots as the rotation speed is probably much too low for an actual F16.36) . the typical takeoff will rotate to some given angle of attack. CL ( IGE ) CL (OGE ) ≥ 1. Upon reaching the rotation speed.015 rolling coefficient of friction. 13 degrees α is reached at 130 knots calibrated airspeed.33) (∆α ∆t ) = 3. Then. In the numerical integration. that angle of attack is held until the aircraft generates enough lift such that lift is greater than weight and the aircraft lifts off the runway.34) (8.
Standard atmosphere is defined in the altitude section.15 ) = temperature ratio 84 (8.48 knots.40) ! γ = sin −1 h V flight path angle t (8.44) (8.43) Vt = true airspeed. Equation 8. N z . T = 288.0 Airspeed.39 through 8. During the air phase. and Vw = windspeed. however. We will assume windspeed equals zero.37) (8.41) From the N x .39) Vt ! During the ground roll.46) . are functions of airspeed and pressure altitude. the normal load factor equation is used.15 − (1. we can numerically integrate groundspeed ( Vg ) and geometric height ( h ). the hdot ( h ) term is zero.9812 /1000) ⋅ H C Vt = Vg + Vw where: (8. M = Vt / a a = aSL ⋅ θ = speed of sound where: aSL = 661. We have assumed a standard atmosphere for temperature.45) θ= T ( 288.38) (8.42) (8. The following equations were derived in Section 4. and γ equations (8. All of the forces. N z = cos(γ ) + Vt ⋅ γ! g0 (8.40 is derived in the section on normal load factor during a climb.Fex = Fn − ( D + Frw ) Fex = N x ⋅ Wt Nx = ! Vg g0 ! +h (8. (8.41).
000 Excess Thrust Force (lbs) 6.5) − 1 = calibrated airspeed } (8.2 ⋅ M 2 3.5 − 1 P a (8. VC = aSL ⋅ {5 ⋅ (q C PSL + 1) (1/ 3. A plot of thrust.000 4. drag plus the runway resistance terms and excess thrust versus calibrated airspeed.48) where: 2 PSL = 2116.000 8.4.4 Takeoff Forces The time history of the simulation is shown in Figure 8. is shown in Figure 8. 85 . qC = 1 + 0.000 Net Thrust 10.5.47) where: qC = compressible dynamic pressure.000 Drag + Runway Resistance 2.22 (lbs/ft ) = ambient pressure at standard sea level.000 0 0 20 40 60 80 100 120 140 160 180 Calibrated Airspeed (kts) Figure 8. Takeoff Forces Versus Speed 12.
0 0.6 0.3 132.0 0.0 16. from the initiation until completion of rotation.3 50.0 130.4 15.43 VC (kts) 0.1 shows the significant events during the takeoff. the angle of attack is increasing (at 3 degrees per second).3 19.7 26.4) can easily be correlated with the significant events in Table 8. which shows up in a dramatic rate of change of drag. Table 8.2 154. Once angle of attack stabilizes at 13 degrees. 86 .9 23.Takeoff Simulation Time History 200 180 160 140 Takeoff Parameters 120 100 80 60 40 20 0 0 5 10 15 Elapsed Time (sec) 20 25 30 Calibrated Airspeed (Knots) Distance (Feet/50) Angle of Attack (Degrees*10) Altitude (Feet) Figure 8.0 α (deg) 0 0 0 13 13 13 13 HC (ft) 0.5 Takeoff Parameters Table 8. the rate of increase of drag is reduced.1 100.0 167.6 19.0 0.1. For instance.1 TAKEOFF EVENTS Seconds 0.0 Event Brake Release/ Fn = 35 Percent 99Percent Thrust Rotation Initiated Rotation Completed LiftOff Lift>Weight OutofGround Effect Obstacle Clearance Height The inflection points in the drag versus calibrated airspeed plot (See Figure 8.0 8.0 50.
The west end of the runway is 21 feet higher than the east end.131 feet for a perfectly level runway. the absolute rate of descent is about 3 feet per second.24 10.50) Again.5 percent per degree of runway slope. and d = distance to liftoff (feet). A positive wind on this plot is a headwind.6 illustrates the effect of wind.131 3. The average acceleration is computed as follows: a = 2 ⋅ d / t 2 average (mean) acceleration (ft/sec2) where: (8.08 degree (21 feet elevation change in 15.99 10.52 0.49) t = time at liftoff (seconds).6 23. For a typical light aircraft the effect of runway slope is at least twice that amount. the effect of runway slope for this particular model is about 4.001 3.2 EFFECT OF RUNWAY SLOPE Slope (deg) 1.75 11. 87 .0 Distance (ft) 3.73 10.1 magnetic).164 3.5 Effect of Runway Slope Using the pseudo F16 model.0 2. Figure 8.0degree slope runway. The Edwards AFB main runway has an average slope of only 0. for a liftoff at 100 knots groundspeed with a negative 1.403 Time (sec) 22.32 degrees from true north (224. due to the much smaller thrust to weight ratio of the typical light aircraft. Table 8. one must take into account the fact of having a negative absolute rate of climb at liftoff for a negative slope runway.6 25.142foot takeoff distance compared to 3.000 feet).6 24.22 From Zero (pct) 4. the values of time and distance as a function of runway slope (in degrees) are shown in the Table 8. The rate of climb (or descent) with respect to the horizontal plane is as follows: ! h = Vg ⋅ sin(θ rw ) 8.247 3.0 0.0 0.06 As can be seen. The true heading for runway 22 is 238.56 9. this slope would produce a 3. For instance.8 Acceleration (ft/sec2) 11. Although the percentage change in acceleration is about the same for a positive or negative runway slope.5 1.00 2.0 24. For our F16 model. using the same pseudo F16 model.6 Effect of Wind on Takeoff Distance (8. The takeoff speed is 132 knots calibrated airspeed.29 4.2.8.
6 Effect of Wind 8.000 pounds. the simulation that will be presented here will be for sea level. There will be some degradation due to installation losses.7 Takeoff Using Vectored Thrust A limiting factor in takeoff distance for a highperformance fighter may be the ability to rotate the aircraft.25 feet. The engine on an F16 aircraft in maximum afterburner has a static sea level rating of about 25. Rotation would usually be achieved using the horizontal tail. c. We will presume that we have installed a nozzle with vectoring capability. we will assume zero losses. This does not mean the engine. the length of the F16 is 49. For the sake of using even numbers. however. X 1 = 8. In addition. b. X 2 = 4.7 illustrates forces and dimensions for an F16 aircraft. A full fuel F16 with no stores has a takeoff weight of approximately 25. when installed in the aircraft. The following dimensions are approximate values scaled from the diagram: a. The tail generates lift from dynamic pressure.4 feet (distance from weight vector to main wheel).5 feet (distance from main gear to thrust vector).000 pounds. X Fn = 14. 88 . Figure 8. As shown.7 feet (distance from weight vector to nosewheel).Effect of Wind on Liftoff Distance 30% 20% Change from Zero Wind (pct) Simulation Linear (Simulation) 10% 0% 10% 20% 30% 20 15 10 5 0 Wind Speed (kts) 5 10 15 20 Figure 8. produces that much thrust.
25 ft Figure 8. Requiring the summation of moments about the main gear to be equal to zero yields equation 8. We will also ignore the longitudinal forces. At zero airspeed. θV = thrust vectoring angle (+ nozzle up. We will assume that the lift and the weight act through the same distance ( X 2 ). we can solve for the vector angle that would be required to pitch the aircraft at zero airspeed.L Fn θV X1 F1 X2 F2 Wt X Fn 49. This is not generally the case. (W ⋅ X ) θV = sin −1 t 2 ( Fn ⋅ X Fn ) (8.51) Solving for the nosewheel force ( F1 ): F1 = Wt ⋅ X 2 − L ⋅ X 2 − Fn ⋅ sin (θV ) ⋅ X Fn ( X1 + X 2 ) (8. The assumptions made here are deleting higher order terms. to produce a pitch up).7 F16 Dimensions The forces are the same as for the conventional takeoff. A more complete simulation would not make these simplifying assumptions.53) 89 .51. With F1 equal to zero. The difference is that there will be thrust vectoring to produce a pitching moment to rotate the aircraft. lift ( L ) is zero. ∑ M = 0 = F ⋅(X 1 1 + X 2 ) + L ⋅ X 2 − Wt ⋅ X 2 + Fn ⋅ sin (θV ) ⋅ X Fn (8.52) Rotation will begin when the nosewheel force ( F1 ) becomes zero.
For the conditions we have chosen. we would need to rotate the nozzle 18 degrees to rotate the aircraft at zero airspeed using thrust alone. the nozzle angle required will be less due to wing lift.7° (25. Figures 8. Ignoring any tail lift. Once rotation begins. the vector angle computes to: ( 25.54) In round numbers. The engine vectoring would only be used to initiate rotation. standard day and with the aircraft model previously defined. 000 ⋅14.9 illustrate liftoff performance.4 ) θV = sin −1 = 17.51 becomes: ∑M = I where: yy ⋅ q = ( L − Wt ) ⋅ X 2 + Fn ⋅ sin (θV ) (8. For sea level. This 10degree per second rotation rate versus the previous 3degree per second rate was used in the simulation to minimize the distance traveled between initiation of rotation and liftoff.55) I yy = moment of inertia about the ybody axis. It was presumed that some sort of control system function accomplishes the rotation to avoid overrotation at these high rotation rates. The simulation assumed rotation was initiated at 90 knots and a rotation rate of 10 degrees per second was obtained. the vector angle can be decreased as the wing lift increases. and q = body axis pitch rate. equation 8.8 and 8. 000 ⋅ 4.5) (8. At higher airspeeds. The rotation was continued until liftoff attitude ( α = θ ) was attained. Overrotation means aft airframe ground contact. Then that attitude was maintained until liftoff ( L > Wt ). That assumes the engine is producing 100percent thrust at brake release. 90 .
Distance versus Vc 1.9 Angle of Attack at LiftOff 91 .000 900 Distance to LiftOff (ft) 800 700 600 120 125 130 135 140 145 Calibrated Airspeed at LiftOff (kts) Figure 8.8 Distance to LiftOff Takeoff: LiftOff Alpha versus Airspeed 16 Angle of Attack at LiftOff (deg) 14 12 10 120 125 130 135 140 145 Calibrated Airspeed at LiftOff (kts) Figure 8.
10 Effect of Thrust Component on LiftOff Speed 92 . During the ground roll.56) The complete formula is as follows: L = N z ⋅Wt − Fg ⋅ sin (α + it ) (8.57) Hence. Nz = L Wt (8. a portion of the aircraft weight is supported by the ground.8 Effect of Thrust Component In the previous simulation.8. (With Thrust Component) LiftOff Calibrated Airspeed (kts) 140 130 120 110 100 11 12 13 14 15 16 Angle of Attack at LiftOff (deg) Figure 8. A plot of liftoff speed versus angle of attack (Figure 8. solving for N z : Nz = (L + F g ⋅ sin (α + it )) Wt (8. which has been the subject of this entire section so far. Effect of Ignoring Thrust Component In Lift Axis 150 Ignoring Thrust Component With Thrust Component Poly. For instance. The simplified formula we used to compute normal load above is shown in equation 8. (Ignoring Thrust Component) Poly.58) We have presumed the thrust incidence angle it is zero. The effect of ignoring the Fg ⋅ sin(α ) term is quite dramatic. at the typical liftoff angle of attack for an F16 of 13 degrees α . which is only applicable after liftoff has occurred. Once the thrust vectoring has accomplished its task of rotating the aircraft. the term for Fg = 25.624 pounds of extra equivalent lift.56. the nozzle would be vectored to zero degrees with respect to the thrust axis.10) illustrates the effect.000 pounds yields 5. we have ignored the component of thrust.
11. Solving for the nose load: L2 ⋅ ( XL2 − {X 2 − XL1}) + Wt ⋅ X 2 − L1 ⋅ ( X 2 − XL1 ) F1 = ( X1 + X 2 ) (8. The liftoff angle of attack was varied to produce the variation in liftoff speed. We can take the nosewheel force equation and replace the thrust vector term with a tail lift term. Distance versus LiftOff Airspeed: Effect of Ignoring Thrust Component 1. we (the simulation) are able to liftoff at 116. Again. Without thrust vectoring.2 knots in only 618 feet. F1 = Wt ⋅ X 2 − L ⋅ X 2 − Lt ⋅ X t ( X1 + X 2 ) (8.11 Effect of Thrust Component on Distance to LiftOff At 13 degrees α .60) To rotate the aircraft using tail lift. we will ignore runway slope and vertical terms.The corresponding distances are presented in Figure 8. replace the terms above with the more general terms as shown in the C141 diagram (See Figure 8.59) Now. the tail lift ( L2 ) must be negative.000 900 Distance to LiftOff (ft) 800 700 600 Ignoring Thrust Component 500 With Thrust Component 400 100 105 110 115 120 125 130 135 140 145 150 Calibrated Airspeed at LiftOff (kts) Figure 8.61) 93 . taking moments about the main gear: ∑ M = 0 = F ⋅(X 1 1 + X 2 ) + L1 ⋅ ( X 2 − XL1 ) − L2 ⋅ ( XL2 − {X 2 − XL1}) − Wt ⋅ X 2 (8. However.1). the F16 would (for these conditions) not be able to rotate before approximately 130 knots.
Figure 8.62) For our aircraft model.4 = 18.1 0 0 W in g C L = 0 . Solving for the required tail lift: L1 ⋅ ( X 2 − XL1 ) − Wt ⋅ X 2 L2 = ( XL2 − {X 2 − XL1}) (8.50 is assumed along with sea level standard conditions and a gross weight of 25.13 is for a tail area of 80 ft2.000 pounds.3 0 W in g C L = 0 . DeltaTail Lift Tail C L= 1. Four values of wing lift coefficient are chosen to be 0.9 (8.4 0 1000 2000 3000 4000 110 115 120 125 130 135 140 145 150 155 160 C alib r at e d A irs p e ed ( k n ot s ) Figure 8.3 ⋅ ( L1 − Wt ) (8. 0.12 shows the results of plotting ∆Lift versus calibrated airspeed ( VC ) for a tail area of 60 ft2.12 Delta Tail Lift for Tail Area = 60 ft2 94 . ∆Lift = L2 − 0.64) Next.30 and 0.3*( L 1W eig ht)] (p o u nd s) 2 0 00 1 0 00 W in g C L = 0 .20.63) And: ( L1 − Wt ) ⋅ X 2 = 0.40. Figure 8.10.3 ⋅ L − W L2 = ( 1 t) ( 1 t) ( XL2 − X 2 ) (8.303 ⋅ L − W ≅ 0. A tail lift coefficient of 1.5 + 4. Then: XL2 = X Fn + X 2 = 14. 0. we can compute the difference between the tail lift ( L2 ) and the opposing lift from weight ( Wt ) and wing lift ( L1 ).2 0 W in g C L = 0 . the angle of attack ( α ) will be zero.Rotation will occur when the nose load ( F1 ) equals zero. but the wing will provide lift due to having flaps down configuration.5 .65) During the aircraft takeoff ground roll. Ta il S =6 0 ft^2 4 0 00 3 0 00 D elta L ift [L 20. we have assumed XL1 = 0 and we will assume the tail force acts at the same point where we assumed the thrust vector acted.
due to thrust being much greater than drag. 95 .5.10 0 Wing CL= 0. This is much greater than drag at liftoff speed. Figure 8. we have assumed a constant 25. S tail = 80 ft^2 4000 3000 Delta Lift [L20.3*(L1Weight)] (pounds) 2000 1000 Wing CL= 0.40 and a tail area of 80 ft2.000 pounds of thrust. The high rotation rate in the simulation was necessary to achieve reasonable liftoff speeds. For this aircraft simulation. The line is approximately a straight line and is such.14 shows the results. For instance.40 1000 2000 3000 4000 110 115 120 125 130 135 140 145 150 155 160 Calibrated Airspeed (knots) Figure 8. for a wing lift coefficient of 0. The 10degrees per second rate is much greater than a normal rate of about 4 degress per second.14).30 Wing CL= 0.13 Delta Tail Lift for Tail Area = 80 ft2 The points on the plots where the ∆Lift becomes positive is the minimum speed for rotation. By varying the rotation speed. the minimum rotation speed is about 119 knots (from Figure 8. The rotation rate was assumed 10 degrees per second in each case. which produces a nearly constant acceleration versus speed.13).20 Wing CL= 0. we can generate a plot of distance versus speed for liftoff (Figure 8.Delta Tail Lift CL tail = 1.
200 1.14.000 900 800 130 140 150 Calibrated Airspeed at LiftOff (kts) 160 170 Figure 8. the aircraft model) was accelerating.Distance versus Vc at LiftOff 1. Figure 8. LiftOff Vc versus Rotation Vc: Thrust = 25. the limiting factor in liftoff was the rotation rate.15 shows rotation speed versus liftoff speed and illustrates just how rapidly the aircraft (in this case.15 Calibrated Airspeed at LiftOff 96 .000 lbs 170 Calibrated Airspeed at LiftOff (kts) 160 rotation=10 deg/sec Linear (rotation=10 deg/sec) 150 140 130 110 120 130 Calibrated Airspeed at Rotation (kts) 140 150 Figure 8.14 Distance to LiftOff versus Airspeed In each data point in Figure 8. The liftoff occurred before 13degrees α was achieved.100 Distance to LiftOff (ft) 1.
554 At rotation for 130 knots. Takeoff Model: CL versus Alpha 2. Figures 8.2 knots per second.1420 Lift (lbs) 1.795 22. That is why we needed such a high rotation rate.0 0 2 4 6 8 10 12 14 16 Angle of Attack (deg) Figure 8.3 FORCES AT LIFTOFF SPEED Fn (lbs) 25.0 1. the speed is increasing at 17. in order to achieve a reasonable liftoff speed.16 Takeoff Lift Model 97 .0501 0.10 1. for an excess thrust of 22. We must emphasize here that the model used was not an accurate F16 model.16 and 8.Table 8.0 13.000 25.3 shows the forces at 130 knots calibrated airspeed.420 0. but merely an approximate model used to illustrate takeoff principles.6 Lift Coefficient (CL) 1.437 Frw (lbs) 345 9 Fex (lbs) 23.8 0. The equations for the lift and drag models were presented earlier.4 0.716 24.795 pounds.0 CL CD 0.2 0. Table 8.369 Drag (lbs) 860 2.000 α (deg) 0.17 are plots of these equations.
we will discuss takeoff of a twoengine aircraft with an engine failure at some point during the takeoff ground roll.20 0. We will make simplifications. Ground: α → CL → CD (8. In the real world.16 Drag Coefficient (CD) 0. such as assuming an instantaneous loss of thrust on the failed engine. one is able to compute lift coefficient. Experience has shown that this nonlinearity will begin at lift coefficients on the order of 0.17 Takeoff Drag Model In computing drag on the ground. 8. However.12 0.66) Once liftoff occurs. We will use the same pseudo F16 aero model.Takeoff/Landing Drag Model: CD Vs Alpha 0.not to generate an accurate simulation. The purpose herein is to illustrate basic principles . Then.67) The lift and drag model used for this analysis is an idealized linear model. and finally drag coefficient. then compute lift coefficient. Ignoring the component of gross thrust: Air: CL = 0.000675 ⋅ N z ⋅ Wt →C (δ ⋅ M 2 ⋅ S ) D (8.04 0 2 4 6 8 10 12 14 16 Angle of Attack (deg) Figure 8.08 0. Let us presume a very simple thrust model for each engine as follows: 98 . we will assume two engines instead of one. you start with a given angle of attack.9 EngineInoperative Takeoff In this section. there will be deviations from the linear model caused by flow separation at higher angles of attack. you start with lift coefficient and compute drag coefficient. You can also measure angle of attack.50.
we will simulate a takeoff at high altitude where the performance would be minimal if one engine were to fail.18 Takeoff Parameters versus Time Takeoff forces versus calibrated airspeed up to an altitude of 100 feet are presented in Figure 8. 000 pounds. Fn δ = 5. We will assume 10. Takeoff Parameters versus Time 180 160 140 120 Takeoff Parameter 100 80 60 40 20 0 0 10 20 Elapsed Time (sec) 30 40 50 Calibrated Airspeed (kts) Distance (ft/100) Angle of Attack (deg*10) Altitude (ft) Figure 8.6877 ).19.000 feet pressure altitude ( δ = 0. Figure 8. 99 .18 is a time history of a simulation for our 25. Now.a. The plot is for both engines operating.000pound aircraft model with both engines operating.
277 3. We will now assume that one engine fails at exactly the initiation of rotation ( VC = 130 knots).50 ! V (kts/sec) 4.864 6 47.20 illustrates the same parameters as shown in Figure 8.0 161.0 6.94 1. 50 feet AGL (above ground level) 6.600 2.82 11.0 134.800 0 130.000 2.000 Force (lbs) 4. 100 feet AGL Event* Time (sec) α (deg) VC Fn D + Frw (lbs) 375 1. Table 8.32 3.872 5 44.0 165.881 3.850 *The numbered events are as follows: 1.19 Takeoff Forces versus Airspeed For liftoff and 50 feet.265 ! h (ft/sec) 0 0 0 3.8 6.725 13.206 2.877 3 33.585 Fex (lbs) 6.671 4.000 Net Thrust Drag + Runway Resistance Excess Thrust 3.000 5.000 0 0 20 40 60 80 100 120 140 160 180 Calibrated Airspeed (kts) Figure 8.990 3.05 The twoengine case in Figure 8. Figure 8.4 TAKEOFF PARAMETERS AT FLIGHT EVENTS (kts) (lbs) 1 0 0 0 6.96 4.877 4 39.6 6. Liftoff 4.71 1.0 150.4 presents takeoff parameters.000 1.19.877 2 31. Table 8.19 was presented primarily as a baseline of comparison for the following engine failed case. Brake release 2.575 13.100 13.441 3.3 6.7 feet) 5. Initiate rotation 3.550 13.TwoEngine Takeoff Forces 7.6 6.26 2.000 6.423 3. 100 . Outofground effect ( hAGL = 19.41 24.502 5.
Engine Failure Takeoff Forces 8. that excess thrust is being used for climb at the expense of airspeed.20 Takeoff Forces versus Airspeed: Engine Inoperative Table 8.425 *The numbered events are as follows: 1.71 0.877 2 31.00 13.000 6.000 3.20 As can be seen.0 Liftoff 4.70 13.80 0 130.1 31.0 153. by the time altitude equals 100 feet the aircraft is slowing.0 50 feet AGL (above ground level) 6.502 5.133 3.5 duplicates Table 8.000 5.01 0.877 2.000 7. Table 8.0 154.70 0.0 100 feet AGL D + Frw (lbs) 375 1.38 0.438 4 68.0 Outofground effect ( hAGL = 19.438 3 33.000 1.0 3.0 6.79 0 130.206 1.32 1.671 2.0 147.04 ! V (kts/sec) 4.7 3. except we will add an event (2.436 5 100.206 2.6 3.432 6 109.63 3.05 13. 101 .7 feet) 5. which is immediately after we fail one engine in the simulation.232 935 552 299 325 ! h (ft/sec) 0 0 0 0 0.0 132.5 TAKEOFF PARAMETERS AT SIGNIFICANT EVENTSENGINEINOPERATIVE VC Fn α Time (kts) (lbs) Event (sec) (deg) 1 0 0 0 6.884 3.4 for the same events.100 Fex (lbs) 6.96 4.49 7.1).503 2.1 Engine failure 3.000 0 0 20 40 60 80 100 120 140 160 Calibrated Airspeed (kts) Net Thrust Drag + Runway Resistance Excess Thrust Figure 8.0 3.0 Initiate rotation 2.0 Brake release 2.00 13.6 3. Although excess thrust is increasing slightly.000 Force (lbs) 4.000 2.
This is just one possible option. The drag reduction would be accomplished by raising the gear. idle thrust decelerations may be performed. Then.In case of an engine failure in such a scenario. One would accelerate the aircraft on the runway to some high airspeed. conduct a lowg turn (to minimize drag) and return to base for landing. We can solve for drag ( D ) in the equation found in the Developing a Takeoff Simulation subsection and then put D into coefficient form. Then. cut the throttle to idle and allow the aircraft to freely decelerate. 8. one would need to reduce the drag and pitch over to reduce rate of climb. The aircraft flight manual would contain the recommended emergency procedure.69) 102 . Lift and drag coefficients are discussed in the lift and drag section of this handbook.10 Idle Thrust Decelerations To assist in the development (or verification) of a takeoff and landing simulation. D = [ Fn − Fex − Wt ⋅ sin(θ rw ) − µ ⋅ Wt ⋅ cos(θ rw ) ] + µ ⋅ L (8.
one can see the effect of braking coefficient of friction ( µ ) upon stopping performance.000 8.9. In Figure 9.000 pounds. when one has an 8.1 Braking Performance Using the same aero model as for takeoffs.000 0 20 40 60 80 100 120 140 Calibrated Airspeed (kts) Figure 9. the µ for maximum braking is typically between about 0.1 Braking Forces For a dry runway. Figure 9.0700 to account for additional drag devices (such as spoilers) activated during braking. you usually will not conduct a maximum performance stop just to minimize tire and brake wear.50. representing Idle thrust.35 and 0. Braking Forces: Mu = 0.0700.35.000 Force (lbs) 0 Net Thrust 2. Minimum drag coefficient has been increased from 0.000foot runway.000 Drag + Runway Resistance 4. more representative of landing weight.300 feet with zero wind.300 feet pressure altitude with initial speed of 130 knots groundspeed.2 shows the distance as a function of µ for the 20.000 4.000pound aircraft at 2.0 LANDING 9. this is a typical dry runway value. 103 .000 6.1. The initial groundspeed was 130 knots for a calibrated airspeed of 124. the coefficient of friction has been set to a constant 0.35. The gross weight has been reduced to 20.0500 to 0. Fn = 600 lbs 8. However. The thrust has been set to a constant 600 pounds. The pressure altitude is 2. Cd= 0.8 knots.000 2.000 Excess Thrust 6.
Braking Deceleration 10 9 8 7 6 Deceleration (kts/sec) 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 Mu = 0.Dry Runway: Distance versus Mu 3.50 Mu = 0.25 to 0. Hydroplaning is where the tires ride on a film of water and never contact the runway.30 Mu = 0. Figure 9.3 Deceleration versus Calibrated Airspeed For wet runway conditions.50.000 1. Figure 9. The test 104 .10 Calibrated Airspeed (kts) Figure 9.000 0.4 represents actual test data.20 Mu = 0. This is especially true at high speed where hydroplaning may occur.40 Mu = 0. the µ is much less than for dry runway conditions.500 3.35 0.40 0.500 2.000 Stopping Distance (ft) 2.2 Stopping Distance versus Mu ( µ ) For the braking coefficient range of 0.3 illustrates the deceleration (knots per second) versus calibrated airspeed.30 0.500 1.45 0.25 0.50 Mu: Braking Coefficient of Friction Figure 9.
381E06x3 + 1.05 0.326E01 0. The forces are computed using the mu or µ from Figure 9. the data will not extend to the full range of the desired simulation.15 0.was on a wet runway.047 if Vg > 130 knots . y = 3.20 0.5. Using the curve fit beyond the range of its data should be avoided by use of limits. µ = 0.336 if Vg < 10 knots . Wet runway forces are shown in Figure 9. and b. The data points are average values of the actual data and the line is a fourthorder polynomial curve fit of the data points.4.4 Mu versus Groundspeed (Wet Runway) Figure 9.00 0 20 40 60 80 100 120 140 Ground Speed (Knots) Figure 9.1. The limits that will be used in applying the curve fit will be the curve fit values at the extreme points as follows: a.4 shows the braking coefficient computed from braking tests.30 Friction Coefficient (MU) 0. µ = 0. 105 . with the water applied using water tankers. Invariably.35 0. A limit would be where the curve fit value (y) would take on some predetermined constant value if the x value exceeds the highest (or lowest) value used in the curve fit. A warning is appropriate for using curve fits in simulations.25 0.811E04x2 .1.736E09x4 .40 0.137E02x + 4.10 0.
aerobraking may be used on a dry runway simply to reduce wear on the brakes and tires.000 6. Second is the control power available to hold the aircraft up at an angle of attack.000 0 20 40 60 80 100 120 140 Calibrated Airspeed (kts) Net Thrust Drag + Runway Resistance Excess Thrust Figure 9. 106 . That is a factor of more than three times longer for a wet runway. instead of immediately pushing over to a 3point attitude to begin braking. aerobraking may be used. The ability to perform aerobraking is limited by at least two factors. "your results may vary.6 illustrates the difference in total resistance for aerobraking versus 3point braking.” 9. in many cases.000 8.5 Braking Forces versus Calibrated Airspeed The simulation for our wet runway model produces a total distance of 7.236 feet for our dry runway model using a constant µ of 0. In addition. First is the tail scrape angle. This compares to a distance of 2. However.000 4. the aircraft is held at a high pitch angle (to produce a high angle of attack) to maximize the aerodynamic drag. which limits how high of an angle of attack may be held. in order to reduce the ground roll. but as the saying goes.000 6. the 3point braking has more resistance except at high airspeed. For this simulation.000 4.000 Force (lbs) 0 2.059 feet.35. That is typical. Figure 9.2 Aerobraking When one is faced with a wet or icy runway. aerobraking can be more effective.000 2. Upon touching down.Braking Forces: Wet Runway 8.
3 Landing Air Phase The landing air phase will be discussed using the same aircraft model we have used for the takeoff discussion and the landing ground roll.600 1. The simulation used an estimated CL of 1. Then.05 (out of ground effect) for an angle of attack of 13 degrees. Vt ⋅ γ! g0 (9.2) Each aircraft is flown differently and different pilots may have slightly different pilot techniques. However.15 degrees alpha 2.Drag + Runway Resistance Comparison: Aerobraking versus 3point Braking 2. The simulation will be conducted by first computing the initial conditions.400 100 105 110 115 120 125 Calibrated Airspeed (kts) Figure 9. We can compute the initial speed (Mach number).40). For our simulation. N z = cos(γ ) + Then. that angle of attack is 13 degrees.3) 107 .200 2.400 Drag + Runway Resistance (lbs) 2. we can compute Mach number as follows when we also have given the weight and altitude: M= 0. From angle of attack we can estimate the lift coefficient ( CL ).800 1.000 1.6 Total Resistance Force Comparison 9. a typical final approach technique is a constant angleofattack descent. by assuming that the flight path angle ( γ ) is initially constant ( γ! = 0 ).1) N z = cos(γ ) (9.600 3Point Braking Aerobraking. The normal load factor equation is the same as for takeoff (equation 8.000675 ⋅ N Z ⋅Wt δ ⋅ S ⋅ CL (9.
7 Final Descent Rate versus Initial Descent Rate The aircraft simulation predicted that. Fn = D + Fex = D + N x ⋅Wt (9.3 is solving for Mach number from equation 5. the pilot would take action to touch down with stick. we will assume that true airspeed is constant. However. The initial conditions chosena runway pressure altitude of 2. the descent rate will decrease. the aircraft would not touch down at any initial descent rate less than about 11.300 feet at a standard day and an obstacle clearance height of 50 feetare what might be typical with a postmission weight of 18. the initial descent rate is varied. This is an ideal computer simulation.6 in Section 5. throttle or speed brake.7 illustrates the dramatic impact of ground effect. Figure 9. not a real airplane. Further.Equation 9. which would decrease 108 . A constant angle of attack of 13. A pushover would decrease angle of attack. In the real world. for the conditions specified.2 ft/sec. Final Descent Rate versus Initial Descent Rate 18 16 Final Descent Rate (at 0 feet) (ft/sec) 14 12 10 8 6 4 2 0 10 12 14 16 18 20 Initial Descent Rate (at 50 ft) (ft/sec) Figure 9.4) We can then solve for the net thrust that would be required to have true airspeed constant at the beginning of the landing descent.0 (Lift and Drag).5) Having performed these computations. the ground effect will increase the lift and hence.000 pounds.0 degrees is maintained and thrust is held constant. the additional lift is a factor of 1. initially. the simulation enters ground effect at 16 feet (AGL) and at touchdown.30. For this aircraft model. The longitudinal load factor equation then gives: Nx = ! ! ! h Vt h + = Vt g0 Vt (9.
0 4.5 5. you are already at near idle thrust and the small additional thrust increment could be insufficient. When the same simulation was performed with ground effect terms eliminated.8 Landing Air Phase 9. Figure 9. Maryland.074 feet. Finally. Reducing thrust will also cause a deceleration.9 is a U.8. The total distance from 50 feet to touchdown was computed to be 1. a pullup would also eventually get you on the ground.lift.5 4. the lift will decrease and when lift becomes less than weight.4 Landing on an Aircraft Carrier The following text is the result of information given to the author by Page Senn and Richard Huff of the Naval Air Weapons Center. By pulling up sufficiently to dramatically increase drag.7 feet. Rate of Descent.S. for a difference of 96 feet or nearly 10 percent of the air distance. Notice the curvature in the final phase of the altitude versus time. The wing is roughly half the distance between the pilots eye and the tailhook as can be seen from the photo. 109 . Interestingly. The simulation computations were begun at 50 feet AGL (above ground level). speed brake can be used to slow down and reduce lift.0 3.1° and glideslope = 3. With a lower airspeed.0 Elapsed Time from 50 ft (sec) Altitude (ft) Rate of Descent (ft/min) Distance (100 ft) Figure 9.5 degrees (or γ = −3. thereby increasing descent rate. however. the vertical height from the tailhook to the pilot’s eye is 16. but only the last 20 feet are shown. Last 20 Feet of Landing Descent 20 18 Altitude. A time history of the descent for the landing simulation is shown in Figure 9. Patuxent River.5° )].5 3. you will descend. At landing attitude [ α = 8. A pushover to about 10 degrees angle of attack would suffice. Distance/100 16 14 12 10 8 6 4 2 0 2. the total distance changed to 978 feet. the aircraft will decelerate. Navy photo of an F/A18 with its tailhook extended. The situation we will discuss is the landing of an F/A18 on a Nimitz class carrier.
Nimitz.2 feet.08 feet for no flare. The landings are accomplished from the aft deck while the carrier is maintaining forward speed to give a minimum wind over the deck of 15 knots.S. Figure 9. With the minimum windspeed of 15 knots.S. Nimitz The distance from the ramp to the target hook touchdown point is 230. A more normal wind is 25 knots.9 F/A18 with Tailhook Extended Figure 9.5degree glideslope. For the 3. Figure 9. For the F/A18 at 33.10 is a Navy photo of the U. this yields a groundspeed of 131 knots (14615) assuming standard day temperature. the airspeed is 146 knots.000 pounds. this computes to a hook to ramp clearance of 14.Hence.S. We will use that height to make estimates of ground effect. We can calculate the time from passing over the ramp to tailhook touchdown as follows: 110 .S.10 The U. the wing height above the tailhook is about 9 feet.
2.000675 (9.174 ! Vt = g 0 ⋅ N x = ⋅ N x = 19.04 sec speed (ft/sec) 131 ⋅1. For an outofground effect drag coefficient of 0. we calculate ∆D using equation 9.27 76. for a weight of 33.2 9.06 ⋅ N x (knots/sec) 1.22 72. Also shown is an estimate of percentage reduction in drag from Figure 8.5degree glideslope.2132. The average sink rate from the ramp to target hook touchdown computes to 13. For the F/A18.7 ft/sec). The wingspan of the F/A18 is 40.8 0.1 shows the height/span (h/b) of the aircraft versus distance along the deck from over the ramp to tailhook touchdown.13 seconds. 000 g 0 Vt g 0 32.1 0.∆time = distance(ft) 230. we can estimate the change in speed by integrating. δ = 1. A wind of 25 knots.6 200 10. This assumes that all of the drag change goes into acceleration and none into changing the rate of descent.7) For sea level standard day.34 82.000 pounds we calculate longitudinal load factor and then the derivative of velocity.8 50 20.50 91.1 GROUND EFFECT PARAMETERS FOR F/A18 CARRIER LANDING Distance Percentage Traveled Wing Height Drag Point Over Deck (ft) (ft) h/b (pct) Ramp 0 23.0 and airspeed of 141 knots yields a Mach number ( M ) of 0.0 0.7. Table 9. Now.42 87.9 0.0 0. Table 9. Airspeed and Mach number relationships are found in Section 4 (Airspeed).4 feet. This compares to the nominal sink rate 14 fps.1 Note: The percentage drag is an estimate of the drag as a percentage of the outofground effect drag. for instance.25.57 94. would produce a time of 1.0 0.8) For a groundspeed of 126 knots (212. Then. From ∆CD .4 150 13. to estimate ground effect. One form of the relationship between drag and drag coefficient is derived in the lift and drag section and is repeated below: ( ∆C ∆D = D ⋅δ ⋅ M 2 ⋅ S ) 0. we will assume a constant descent rate based upon on a 3. 111 .5 fps (ft/sec). ! ! h V ! V ∆D = t + = t 33.4 100 17. the gear limit is 25 fps and testing at Patuxent is accomplished up to 20 fps.2 = = 1.6) Since 15 knots of wind is the minimum.6878 Nx = (9. the time will generally be longer.6878 (9.6 Hook Touchdown 230. We can estimate the change in speed of the aircraft due to ground effect.
The aerodynamic drag coefficient during aerobraking at 13 degrees angle of attack was determined to be about 0.0356 0.24 0.63 ∆C D 0.0258 0.25.0 91. tests were also conducted to determine aerobraking drag and dry runway braking coefficient.5 Stopping Distance Comparison During the same series of tests that produced the braking coefficient of friction data in Figure 9. it is deflected upward creating an updraft for the oncoming aircraft.985 ft/sec (9.49 0. the aircraft first encounters an updraft.577 0.0698 Note: Above data based upon an Navyprovided number.91 0.4 0.0 82.2 72. then a downdraft.16 141.8 0.0478 0.71 0.94 0. The ship is traveling at a minimum of 15 knots.00 141.174 0.0106 0.30 141. and then a sudden loss of any vertical wind as it encounters the aft deck.0 87.0316 150.2 CHANGE IN TRUE AIRSPEED DURING LANDING DUE TO GROUND EFFECT Distance Traveled (ft) Percentage Drag (pct) ! ∆Drag ∆time ∆Vt Vt (sec) Nx (lbs) (kts) (kts/sec) 351 0.48 141. Navy tests did indicate a 1 to 2 knot increase in INS groundspeed during landing.06 851 0.68 0.3 summarizes the data for wet runway. For a nominal landing gross weight.880 0. was not a Another factor in landing on a carrier is the wind over the deck. values of lift coefficient were determined from either predicted models or flightdetermined. This Vt (kts) 141. we can calculate the change in speed by integrating the speed derivative as shown in Table 9.19 1.0586 230. the air flows downward aft of the ship.7 ⋅ sin(−3.0 94.1 0. In addition.47 0. dry runway.14 outofground effect drag coefficient of 0. Then.9) Now.4 0.30.08 0.0436 200. The dry runway braking coefficient ( µ ) was found to be in the vicinity of 0.6 0.35. when that air contacts the sea below.2. and aerobraking.5°) = −12. 112 .! h = Vg ⋅ sin γ = 212.09 1. Aerobraking can be maintained until approximately 70 knots calibrated airspeed.14 1.0176 0.10 1. limited by available horizontal tail power.0570 1. Table 9.6 0. There is a downdraft (negative vertical wind) immediately aft of the deck.34 0. the touchdown speed is 135 knots calibrated airspeed.0130 50. 9. Table 9.06 141.20 582 0.0216 100.0 76. So.4.
dry runway 3point braking provides the greatest deceleration at all speeds.088 5.4 0.660 Aerobraking ! Vt (kts/sec) 6.10) Using equation 9. this was from Askania cameras. The cameras tracked the aircraft on film at up to four frames per second.000 feet.350 Figure 9.17 125 7.300 In addition. Plus.47 2. The film contained azimuth 113 .90 0.6 Takeoff and Landing Measurement In the past (prior to this handbook). First.4 should be made.58 2.11 5.48 2.4 INTEGRATION OF BRAKING RESULTS Airspeed Dry ! VC Vt (kts) (kts/sec) 135 7. Since thrust was a small contributor to the distance integration. At the AFFTC. by aerobraking for the first 20 knots (135 to 115) the difference in distance is only just over 100 feet. However. by using aerobraking down to 100 knots. the distance to stop on a wet runway can be reduced by more than 1.635 N/A N/A A few observations from Table 9. WET.768 3.25 4. For this small increase in stopping distance.510 2.3 DRY.71 3.8 and integrating versus time to compute distance yields Table 9.34 2.95 100 6.920 5.446 1.283 Wet ! Vt (kts/sec) 2.095 0.63 50 6.4.950 2. AND AEROBRAKING DATA SUMMARY Lift Coefficient CL Drag Coefficient CD Braking or Rolling Coefficient ( µ ) 0. runway slope and wind were assumed zero and standard day conditions at sea level were used. an idle thrust model was provided by the engine manufacturer.500 feet from the runway. The equation for excess thrust ( Fex ) then simplifies to the following: Fex = Fn − D − µ ⋅ (Wt − L ) (9. we will ignore thrust incidence. Table 9.015 3Point Braking: Dry 3Point Braking: Wet Aerobraking 0.20 0.20 0.06 115 6.17 Note: N/A – not applicable Dry Distance (ft) 0 307 598 992 1.095 0.12 N/A N/A Aerobraking Distance (ft) 0 386 705 1. 9. Second. a substantial reduction in energy absorption by the brakes can be achieved – thereby increasing the service life of the brakes.63 2. much of takeoff performance utilized external tracking.81 80 6.Table 9.41 0 6. Askania was the brand of the particular cameras located in towers near each end of the main runway and about 1.45 3.693 2.90 Wet Distance (ft) 0 873 1.04 5.
and altitude. the onboard inertial velocity data can be integrated to provide distance. with the advent of INS and GPS. and computerprocessed.13) i = the i ' th time sample. read. an INS will internally integrate the accelerations at a sample rate of at least 50 samples per second. Improved integration results would be produced using a moving secondorder polynomial curve fit. velocity. a data process used by the AFFTC. the vertical velocity at the liftoff point would be subtracted out. (9. By sampling the INS velocities at no more than 5 samples per second.11) Altitude would be determined by integrating the vertical velocity. it is recommended to differentiate the velocities rather than use a direct output of the INS. The topic of noise in accelerometer data is discussed within the flight path acceleration heading of the excess thrust section. For the height integration. The GPS does not have a null error. the longitudinal acceleration can be determined with something as simple as a central difference derivative method. To compute acceleration. distance. That is because the INS is sensitive to body axis vibrations of the aircraft and the acceleration data will be very noisy due to this vibration. The precise determination of the liftoff point would involve additional onboard instrumentation such main gear loads or wheel speed. The computer output included time. 114 . acceleration. beginning at the point where liftoff occurred. A new device called an EGI (embedded GPS/INS) combines the outputs of both an INS and a GPS using a filter. it is necessary to subtract out any null error. Ax (i ) = where: (V (i + 1) − V (i − 1)) (t (i + 1) − t (i −1)) g g (9. Then. you can essentially average out the noise in the data.and elevation data. Since the INS is subject to small drift errors. d = ∫ Vg ⋅ dt where: Vg = horizontal component of groundspeed. Typically. For the horizontal distance. this is obtained by simply collecting data when the aircraft was stopped. ∆h = ∫ Vv ⋅ dt = altitude above the liftoff point where: (9. The film was developed. Now.12) Vv = vertical component of groundspeed.
1) As long as wind is unchanging. Brown titled.” Diehl states.2 Groundspeed Course Method The course would consist of two parallel lines connected by a line perpendicular to those two lines. airspeed indicators must be calibrated by runs up and downwind over a measured course. Vt = ∆Distance ∆Time (10. Then.1 Groundspeed Course – Heading Method 115 . “Measuring an Airplane’s True Speed in Flight Testing. This is an error greater than 1 percent and even more for speeds less than 100 knots. The course at Edwards. If the groundspeed is 100.G. 10.135 (1923) by W. Note: A positive wind vector direction is the direction from which the wind is blowing. The aircraft heading (direction nose is pointing) would be the same as the course heading in method one as shown in Figure 10.1 Historical Perspective In Engineering Aerodynamics (Revised Edition. then by trigonometry we could compute that the true airspeed is 101. The aircraft would drift from the line due to any crosswind. is 4 miles long. it does not enter into the problem since true airspeed is parallel to the course.1.0 knots and there is a 15knot wind exactly perpendicular to the aircraft’s inertial speed vector. “In general. The first problem is minimized with the advent of GPS to determine groundspeeds.N. These points are a known distance apart. We rarely use the groundspeed course method at Edwards because of its lack of accuracy at high speeds and variable surface winds. Figure 10. This requires a visual hack of when the aircraft crosses the horizontal lines marked on the ground.1 knots. In 1923. The way to determine true airspeed is to simply use a stopwatch to time the aircraft between the start and end lines. for instance. true airspeed is determined by the following. T. Walter Diehl discusses the calibration of airspeed indicators.0 AIR DATA SYSTEM CALIBRATION 10.” We later knew this as the groundspeed course method. However. opposite heading passes are not needed.10. 1936). it is common to conduct passes in opposite headings just to get an average. speeds of order of 100 knots were achievable. Diehl points out that such tests should not be done when the crosswind exceeds 15 knots as that would have resulted in an error in airspeed of more than 1 percent. Then. He references NACA Rep.
The basic measurements are total pressure ( P ). Then. indicated pressure altitude and indicated total temperature. You would need to be able to correct for crosswind if you flew these opposite direction passes as recommended in AFFTC Standard Airspeed Calibration Procedures (Reference 10. static pressure ( P ). The opposite direction pass would be as shown in Figure 10.2. GPS groundspeed is very accurate (0. The opposite direction or track angle passes would have the aircraft fly directly down the groundspeed line with the aircraft pointing into the wind to account for crosswind. In addition. the basic measured parameters on the test aircraft are the actual measured values of indicated airspeed. but also pressure altitude and total temperature. Now. one could determine the component of groundspeed parallel to the course. Vt = (V g1 + Vg 2 ) 2 (10. due to not having to make crosswind corrections. however. pressure altitude ( H C ). The equations for these parameters are included in the airspeed and altitude sections of this handbook. The correction equations are as follows: 116 .With the use of GPS. and Mach number ( M ). one would need to conduct opposite heading passes to average out the wind. and total temperature ( Tt ).2 Groundspeed Method – Direction Method 10. On some limited evaluations. The t static (or ambient) pressure and total pressure are used to compute calibrated airspeed ( VC ).1 m/sec). The data reduction in Reference 10. the average true airspeed is simply the average groundspeed. Note: A positive wind vector direction is the direction from which the wind is blowing.3 General Concepts The terminology ‘airspeed calibration’ actually involved the determination of corrections to be added to not only airspeed.1). You would avoid the problem of visually determining the time passing points on the ground.2) Note a distinction between conducting opposite heading (direction the nose is pointing) and opposite direction (ground track direction) passes. Figure 10. With Mach number and total temperature. the true airspeed and ambient temperature can be calculated. The opposite heading method is preferable.1 ignores crosswind.
The total temperature probe is also subject to an error called a probe recovery factor ( η ). The ground temperature measurement would be the source during tower flyby tests. However.00 for a welldesigned system. and ∆Tti = instrument correction to total air temperature. Tt ) on a test aircraft the i subscript referred to a measurement that had not been corrected for any instrumentation errors. Appendix A contains weather balloon data from the Edwards 117 .4) (10.6) If. or from a ground temperature measurement. Too many performance engineers ignore this valuable source of information. one has an ambient temperature source ( T ) and a total temperature measurement ( Tt ) one could solve for η in the above equation and could calibrate the probe. in practical application with modern probes a value of 1. The other source could have been from another aircraft with a calibrated total temperature probe.98 to 1.2 ⋅η ⋅ M 2 ( ) (10. However. The ambient temperature ( T ) would have been from another source. When dealing with the three basic measurements ( Pt .0 is frequently used.3) (10. from a weather balloon.VC = Vi + ∆ViC + ∆V pC calibrated airspeed H C = H i + ∆H iC + ∆H pC corrected pressure altitude (10. P . ∆H pC = position error correction to pressure altitude.5) Tt = Tti + ∆Tti total temperature where: ∆ViC = instrument correction to indicated airspeed. The relationship for total versus ambient temperature is as follows: Tt = T ⋅ 1 + 0. in flight test. When the parameters are instrument readings that not uncorrected for instrument and position errors then the modifier ‘indicated’ should be applied. the modifier ‘calibrated’ on calibrated airspeed needs to be retained to distinguish it from true airspeed. The value for η is typically 0. ∆V pC = position error correction to instrument corrected airspeed. However. The terminology ‘position error’ refers to the premise that there is some location on the aircraft to locate a sensor such that there would have been zero error in that measurement. Weather balloon data would not be used as a primary calibration source. However. there is no single position that would yield zero error at all Mach number and angle of attack. The modifier ‘corrected’ on pressure altitude is often dropped in practice. it makes an excellent check on your data system. ∆H iC = instrument correction to pressure altitude. The Tt is the test aircraft’s measured total temperature.
So.10) At the AFFTC.7) (10. the symbology used here for ambient pressure ( P ) will be shown as ( Ps ). A position error parameter called delta p over q is defined as follows: ∆Pp / qCic = where: ( P − Pi ) qCic (10. The two pressure measurements could both have ‘position’ errors as follows: Pt = Pti + ∆Pti (10. With the assumption of zero total pressure error. There is also data from a sampling of 1 month of weather soundings. The data illustrates average values of winds and temperatures versus month.AFB weather squadron. However. A study conducted at Edwards AFB in the 1960s indicated that balloon temperature accuracies were on the order of ±2 degrees C. The s would denote static.) A positive correction to be added to ambient pressure would produce a negative correction to be added to both calibrated airspeed and to pressure altitude. For purposes of this handbook static and ambient are considered the same thing. a sign convention has been that a positive sign on ∆Pp would produce a positive correction to be added to both calibrated airspeed ( ∆VC ) and pressure altitude ( ∆H C ). In general. both of the pressure measurements are subject to errors. (One can avoid the confusion of a sign change by thinking of ∆Pp as being a positive correction to be added to the compressible dynamic pressure ( qC . one would need to change the sign on the ambient pressure correction as follows: ∆Pp / qCic = − ( P − Pi ) ( Pi − P ) qCic = qCic (10.9) qCic = indicated compressible dynamic pressure.8) P = Pi + ∆Ps Often. In that case. the correction to be added to compressible dynamic pressure simplifies to the following: ∆qC = −∆Pp (10. and ∆Pp = error in ambient pressure (position error). it is often assumed that there is zero total pressure error. all of the Pitotstatic error is in the ambient pressure measurement.11) 118 .
119 .5 Tower Flyby The tower flyby method of airspeed calibration consists of flying along a flyby line on the lakebed and passing by an observation tower perpendicular to the flyby line some 1. With a grid on a window. Figure 10. The AFFTC pacer aircraft have onboard computers. = standard day temperature (°K) at H p tower . which calculate instrumentation and position errors then add these corrections to the indicated values to present calibrated values. The pacer’s computed values of calibrated airspeed ( VC ). The position errors are the difference between the measured (or indicated) Pitotstatic parameters and the true values.379 feet away (at Edwards AFB). the computed ambient temperature is lumped with the pressure parameters and called Pitotstatic parameters. Just for simplicity. A pressure altitude measurement in the tower is used to determine the zero grid line pressure altitude. This has the potential of passing on errors from another pacer. To avoid that problem the new pacer is also tested using the tower flyby. and Tstd T = test day ambient temperature (°K).3 shows an actual photo of an aircraft (F18) passing by the Edwards AFB flyby tower. The pacer will fly in formation with the test aircraft. An observer in the flyby tower watches the aircraft pass by the tower. 10. it needed to be calibrated before it could be utilized in the airspeed calibration of test aircraft. and ambient temperature ( T ) are compared to those three parameter values from the test aircraft. ∆htower = geometric height of aircraft above the zero grid line measured by the tower.10. the pressure altitude of the aircraft is computed as follows: T H C a / c = H p tower + ∆htower ⋅ std pressure altitude for the aircraft T where: (10.12) H p tower = pressure altitude measured at the zero grid line in the tower. pressure altitude ( H C ). acceldecel. the observer is able to compute the aircraft’s altitude above the tower zero grid line as the test aircraft passes in front of the grid on the window. One of the methods used in calibrating a pacer aircraft is to fly against another pacer aircraft. Before pacer aircraft became the standard for Pitotstatic measurement.4 Pacer Aircraft An aircraft that is utilized in the airspeed calibration of a test aircraft is called a pacer aircraft. The test aircraft’s Pitotstatic measurements are referred to as indicated values until a set of corrections can be determined by simply comparing to the pacers calibrated computed parameters. Then. and cloverleaf methods.
1 Figure 10. Altitude versus Grid Reading y = 31.1) represents flyby tower data.422x 360 340 320 300 280 260 240 220 Altitude (ft) 200 180 160 140 120 100 80 60 40 20 0 0 1 2 3 4 5 6 7 8 9 10 11 12 Grid Reading (in) Extracted from Reference 10.4 Altitude versus Grid Reading for Flyby Tower 120 .3 Flyby Tower Grid Figure 10.4 (Reference 10.Figure 10.
consider a 90degree F day at Edwards.4 degrees K.0 0.422 times grid reading and at the very best a guess to the nearest 0. a method of accelerating 121 . Effect of a 10Foot Error in Flyby Tower Altitude 2.5 Effect of 10Foot Error in Flyby Tower Altitude 10. H C a / c = 2. In addition. Assuming the pressure altitude is equal to the geometric altitude.5 0.1 inch grid is possible.0 1. The test day temperature of 90 degrees F equates to 305. the error in altitude would be 14 feet. then the accuracy of the flyby tower data is about ±3 feet. So. assume the aircraft flew by the tower at a geometric height of 200 feet as follows: a.500 feet. = 2.305 + 200. Too often.5 Error in Calibrated Airspeed (kts) 2. To illustrate the error that could result. The geometric altitude of the zero grid line of the flyby tower is 2.6 AccelDecel It is difficult to obtain stabilized airspeed calibration data in the transonic regime. then the standard day temperature computes to 283.305 + 186. the temperature correction is ignored.4 If one ignores the temperature effect. at supersonic speeds. ⋅ 283.305 feet. Next. Accuracies of better than 3 feet have been demonstrated with differential GPS (DGPS) over the flyby line at Edwards.0 0 100 200 300 Indicated Airspeed (kts 400 500 600 Figure 10.491 305.Since ∆h = 31. which is a normal summer day.5 1. fuel consumption is very high.6 = 2. That is an optimistic figure.5 illustrates the effect of a 10foot error in pressure altitude on calibrated airspeed at a pressure altitude of 2.6 degrees K. This error is computed based upon the assumption that there is zero error in total pressure. Figure 10.
Figure 10. the windspeed and direction. 2. altitude condition. The collection of data points near 30. The altitude survey can be as few as three data points to yield a straightline fit. typically) from the start condition. Figure 10.6.6 and 10. one supersonic acceldecel data set is shown from data collected at the same time as AFFTC data set one. Acquire a few additional data points at the same indicated Mach number. Mach number.000 feet pressure altitude are from a subsonic acceleration corrected using the pacer curve. and total temperature. The method is as follows: a. 122 . and VwE = f (h) . ambient temperature (computed from Mach number and total temperature) and tapeline altitude (radar or GPS). Compute also.8 is the position error parameter versus indicated Mach number. You now have the following functions: 1. The data processing involves computing corrections to be added to airspeed. All of the required equations have been presented in previous sections. and aircraft true airspeed.7 is correction to be added to indicated pressure altitude. Perform an altitude survey over a small range of altitude (±1. the aircraft is accelerated from this known calibration subsonic point through the transonic and into the supersonic regime where the calibration is not known. T = f ( h) . 3. but at different altitudes.7. just presented with different parameters. In Figures 10. That data set is in the discussion of the cloverleaf method. The start condition is some Mach number. groundspeed and direction.and decelerating starting and ending at subsonic speeds (where the airspeed calibration is known from the tests previously described) is used. The four functions above are quite accurately represented by a straightline curve fit. 4. Then. altitude. The extreme data points are stabilized points while the other points are from a subsonic acceleration. The data are corrected using a position error curve previously determined from pacer and tower flyby data. Measure pressure altitude. d. Those data points are shown in the Figure 10. Both plots are the same data.000 feet. The assumption is made that all of the error in the air data comes from the ambient pressure. b. VwN = f (h) . c. Figure 10.6 is a plot of a pressure survey taken prior to a supersonic acceldecel. H C = f (h) where h = tapeline altitude.
000 28.94734x .6 Pressure Survey Delta H versus Indicated Mach Number 1.18223 31.Subsonic Pressure Survey y = 0.0 1.7 0.500 31.500 31.500 30.1 1.500 GPS Altitude (ft) 32.000 32.5 Cloverleaf Delta H data Cloverleaf Delta V data Accel Delta H Method Accel Delta V Method Decel Delta H Method Decel Delta V Method 0.000 Pressure Altitude (ft) 30.138.500 29.3 Indicated Mach Number Figure 10.8 0.2 1.000 29.6 0.000 31.9 1.400 1.000 Figure 10.200 1.500 33.000 800 Delta H (ft) 600 400 200 0 200 0.7 AccelDecel Delta H 123 .500 30.
References 10.6 0. California.5).05 0.9 1. The development that makes this method dramatically more economical for flight test is GPS.03 0.04 0.07 0. The method was originally dubbed the cloverleaf method due to the pattern prescribed in the sky.8 AccelDecel Position Error Coefficient Section 10. One no longer needs to track the aircraft with radar.0 1.01 0 0. This handbook will present the mathematics of this method and some substantiating data.01 0. The method has been applied with reasonable success by the NTPS. Ideally.2).Introduction In the early 1970's.7 is an edited portion of a paper titled.3 Indicated Mach Number Figure 10. for instance. making the math much simpler (Reference 10.8 0.02 0. It does not require that each pass be executed at the exact same airspeed or at precisely 120 degrees apart. The idea is as follows: One assumes that wind remains constant while the aircraft performs consecutive turns to produce three passes through a common airmass. the AFFTC developed a new method to calibrate airspeed.4. is the nonlinear mathematical solution. This method is more commonly referred to as the cloverleaf method. uses a method where the passes are 90 degrees apart. Besides the two components of wind (north and east). which reduces test time and required test resources.2 1.1 1. the passes should be equally spaced in heading (or 120 degrees apart) and at the same indicated airspeed. 10. and there is a reduced cost for data processing. The National Test Pilot School (NTPS).7 The Cloverleaf Method .06 0.Delta P/qcic versus Indicated Mach Number 0. “PitotStatic Calibration Using a GPS MultiTrack Method” (Reference 10. there would be an unknown error in true airspeed that would need to be computed.3 and 10. in Mojave. What this handbook will contribute beyond that which the NTPS has already contributed.5 Cloverleaf Delta H data Cloverleaf Delta V data Accel Delta H Method Accel Delta V Method Decel Delta H Method Decel Delta V Method Delta P/qcic 0. The test 124 . They involve the solution of three nonlinear equations in three unknowns.7 0.
the heading sequence would be 1. The data in this handbook was available at 1 sample per second.8 The Flight Maneuver Figure 10. Then a righthand turn to a heading of ψ 2 is accomplished and repeats another data collection. but the data reduction involves track angle. it will not discuss air data systems at any length. historically.22 psf). In that case.navy. on the other hand.19 knot). For the sake of this handbook. which are referred to as ‘calibrated’ airspeed ( VC ). The test is performed by first collecting stable data along a heading of ψ 1 .usno. The military specification for velocity is 0. it will be shown how one can derive the unknown error in Vt and the north and east components of wind. which invokes the principle that the vector sum of groundspeed plus windspeed is equal to airspeed.2. True airspeed ( Vt ) requires computations involving total pressure ( Pt ).S. In addition. The aircraft was flown on heading. See for instance. That input is a differential pressure between total and ambient pressure. The whole sequence should be performed in one continuous sequence.3. the U. The terminology ‘true’ airspeed is used to avoid the confusion with the cockpit indicator readings. and total temperature ( Tt ). Lefthand turns could also be used.8 illustrates a sequence of cloverleaf maneuvers. since the heading angles do not have to be exactly 90 degrees apart. This handbook will not discuss the theory and operation the GPS system. is more complex. the primary piece of information required of GPS is the accuracy of the velocities and at what update rate they are available. A final righthand turn ends up at a heading of ψ 3 and a final collection of data. The cockpit indicator. Vt .10 meters per second (0. Heading is the direction the aircraft is pointing while track is the angle of the aircraft groundspeed vector. 125 . For those not familiar with calibrated airspeed. Only a few seconds of data are required to acquire average airspeed and groundspeed data.html. The true airspeed. ambient pressure ( P ). Since it is easier to relate to windspeed magnitude ( Vw ) and direction (ψ w ). By solving three equations in three unknowns. This handbook will attempt to explain and demonstrate the validity of a method to calibrate true airspeed ( Vt ).points do not have to be flown as precisely. In addition.mil/gps. 10. the cockpit airspeed indicator only measures actual airspeed on a standard day (59 degrees F) at sea level standard pressure (2116. Navy web site http://tycho. could be constructed mechanically with only one pressure input. the references and bibliography contain just a few of the numerous information sources on these topics. Heading could also be considered the direction of the true airspeed vector when the sideslip angle is zero. Both subjects have been written about at length.3 sequence for the right hand turns.2 instead of the 1. the north and east components will be converted to magnitude and direction.
A discussion of pacer aircraft can be found in References 10.9 Error Analysis This method is a true airspeed calibration method. ambient pressure ( P ). If the data were acquired at roughly equally spaced angles. The NTPS. then the method should produce reasonable results. Finally. has demonstrated that a separation of 90 degrees produces quite adequate results.9 Cloverleaf Flight Maneuver On 19 August 1997.1 and 10. in fact. three cloverleaf runs were performed using an AFFTC F15B pacer aircraft. Each run consisted of three separate passes at track angles about 120 degrees apart. the data 126 . ground speed ( Vg ).000 feet pressure altitude and indicated Mach numbers of 0. These runs were performed at nominal indicated conditions of 30. USAF S/N 132 (Figure 10. Notice that the headings are separated by the ideal value of 120 degrees. a second righthand turn was performed to a track angle of 135 degrees (SE quadrant).Figure 10. Then a lefthand turn was performed bringing the aircraft around to a track angle of 255 degrees (SW quadrant). 0.10 Air Force Flight Test Center F15 Pacer 10. the first pass was at a track angle of 15 degrees (NE quadrant).6. In round numbers.10).7. total temperature ( Tt ). and track angle ( σ g ). In many cases. and 0. Figure 10.6. The first two measurements come from pressure transducers.8. There are five measurements: total pressure ( Pt ).
The error in computed true airspeed for an error in total pressure then is: d. total and static pressure are computed from altitude and airspeed. ( ∆Vt )/( ∆Pt ) = (484.432 psf.0147957.484. while keeping the other parameters the same. in. and c. The last two parameters are either GPS or radar measurements. 127 . Pt = 957. The laboratory calibration accuracy for pressure transducers is about ± 0. Hg = 70.001 in. one will use these numbers and pick a typical condition near the test conditions of the data shown in this handbook. Tt = 272. Therefore. At those conditions (and carrying out computations to beyond usual resolution): a. add the same error (0. for the laboratory accuracy of 1milliinch of mercury (0. b.071 psf) and about ± 0. Ambient Temperature = 242.959 knots (true airspeed).999 knots.001 in. Pa = 628. a. Pt = 958. Keep in mind this is the error slope at just this one set of conditions.944) = 0. Vt = 484. To examine ambient pressure errors.565 (knots/psf) = 0. c. Vt = 484.001 in. Hence. Hg (0.98 °K. the conversion factor is as follows: a.0 °K.044knot error in true airspeed.044 knots per 0. Hg) to ambient pressure. Pressure Altitude = 30. The third one is from a total temperature probe.001 in. Hg Total Pressure. Since we are working with two different units on pressure. Hg "error" to Pt b. and d. In that case.0147 computing true airspeed c. Mach number = 0.10 °K for temperature probes. Hg) the error in total pressure results in a 0.726 psf add 0.source may be altitude and airspeed.800.001 in.959)/(958.944 psf. b.000 feet.999 .
375 635. b.959 639.22 16.41 257. P = 628.459 266.48 390. Vt = 485.959)/(0.a. Subscripts a.7 311.5 3 985.898. the error in each leg might be either to the nearest knot or to the nearest onetenth of a knot. For the AFFTC data.83 506.750 29.4 (deg K) 243.39 326.3.1 through 10. an error in the aircraft parameters equal to their laboratory accuracies would produce errors in Vt of less than 0. b.1degree error in total temperature produces a true airspeed error as follows: a. Hg Ambient Pressure.2 243.1 knot.1 261.482 637.898 .89 (knots/deg K) = 0.686 29. 2.5027.10 Air Force Flight Test Center Data Set The results for the 19 August 1997 data are summarized in Tables 10.959)/(628.174 275.1) = 0. and c denote separate passes. Note that the numbers are displayed to at least one digit more than their accuracy level. For this particular flight condition.627 222.07 16.1 2 878.001 in.089 knots per 0. then.861 (knots/psf) = 0.14 127. Table 10.1 AIRCRAFT AVERAGE MEASUREMENTS AND PARAMETERS Run Number Pti (psf) Psi (psf) Tti (deg K) H Ci (ft) VCi (kts) Ti 1 806. some of the results will be presented to greater than 0.6 Table 10.20 1. Vt = 484.432) = 0.08 545.88 257. A 0.1knot resolution.2 INERTIAL SPEEDS (GPS) Run Number 1 2 3 Notes: Vga σ ga Vgb σ gb Vgc (kts) 370.5027628. 128 .76 471. In the case of using a hand held GPS unit.7 Note: The subscript i denotes indicated value.65 18. The ground speed error is likely to be just the readability of the data. c. 10.51 258.1 °K Total Temperature. Errors in ground speed will produce errors in true airspeed proportional to the error in the ground speed on each leg of the method.74 465. b.79 σ gc (deg) 127.23 (kts) (deg) (deg) (kts) 409. 29.80 128. ( ∆Vt / Tt ) = (485.048484.067 knots per 0.048. but this does not imply that that accuracy level has been achieved.0 243.484.26 431.606 260. ( ∆Vt / ∆P )= (484. Runs 2a and 2b used radar data.
004 30.87 Vw (kts) 48.Table 10.7088 0.14.75 0. (Pacer Curve) 0.07 8.73 5. and 1c). Run number one consists of three separate passes (1a.6927 0.11 Position Error Groundspeed time histories for run number one are depicted in Figures 10.85 0.2 242.93 45.65 0.32 4.5 0.03759 The pacer corrections are known to a high degree of accuracy.74 222.95 Indicated Mach Number Figure 10.8322 ∆Vt (kts) 6.03098 0.030 0.020 0.040 delta P/qcic Pacer Curve 0.5947 0. These corrections are in the form of a curve of the parameter ∆P / qCic versus indicated Mach number. These corrections are applied to pacer data any time the pacer is used to calibrate another aircraft.080 ∆H C (ft) 185 318 453 ∆VC (kts) 3.86 T (°K) 242. both of which have been smoothed in this case with a 19point secondorder polynomial curve fit.4 242.7 0.9 0.54 223.935 30.94 10.8119 M 0.49 ∆P / qCic 0.045 0. These compare radar data and GPS data.025 0. F15 Pacer Position Error 0.03793 0.86 ψw (deg) 223.8 0.11 is a plot of the three cloverleaf data points with a comparison with the pacer curve. This parameter is often referred to as the position error parameter.01 46.050 0.1 HC (ft) 29.3 OUTPUTS Run Number 1 2 3 Mi 0.12 through 10. 1b. Figure 10.035 Cloverleaf Points Poly.6 0. 129 . They are at the same aim airspeed but at different groundspeeds.6054 0.55 0.
F15: Run 1a
413
412
411
Radar GPS
Groundspeed (kts)
410
409
408
407
406
405 0 10 20 30 Elapsed time (sec) 40 50 60
Figure 10.12 Groundspeed – Run 1a
F15: Run 1b
329
328
Groundspeed (kts)
327
326
Radar GPS
325
324
323 0 10 20 30 Elapsed time (sec) 40 50 60
Figure 10.13 Groundspeed – Run 1b
130
F15: Run 1c
373 372 371 Ground speed (kts) 370 369 368 367 366 365 0 10 20 30 Elapsed time (sec) 40 50 60
Radar GPS
Figure 10.14 Groundspeed – Run 1c
For the first run (number 1a), Figure 10.15 illustrates a comparison of true airspeed. The pacer aircraft has a direct output of corrected true airspeed. This is compared to a computation of true airspeed from GPS groundspeed plus the computed windspeed.
F15 Run1a True Airspeed
370 369 368 True Airspeed (kts) 367 366 365 364 363 362 0 10 20 30 Elapsed Time (sec) 40 50 60
GPS plus Wind Corrected Aircraft PitotStatic
Figure 10.15 True Airspeed An interesting observation is that as long as the error in airspeed is the same on each leg, the computed value of wind will be identical. That means one could use this technique to “measure” winds; “measure” since one would actually compute the winds rather than measure them.
131
From the start of the first pass (1a) to the completion of the last pass (3c) was 37 minutes. This was an excessive amount of time for these tests. It seems clear that something considerably less than a full minute of data on each pass would be quite adequate. A 10second average would suffice. Then, by relaxing the requirement to maintain the test airspeed exactly, an additional amount of test time could be saved. Without the need for radar, tracking it becomes unnecessary to coordinate with the radar tracking team and that saves even more time. It seems reasonable that a factor of two or more savings in flight time could be achieved. Thus, not counting the time required to climb to the test altitude, each set of three passes could be concluded in about 5 minutes or less.
10.11 Mathematics of the Cloverleaf Method
The basic vector equation that one will solve for the cloverleaf method is nothing more than true airspeed equals the vector sum of groundspeed and windspeed.
! ! ! Vt = Vg + Vw VtN = VgN + VwN VtE = VgE + VwE Vt = Vti + ∆Vt
(10.13) (10.14) (10.15) (10.16)
The north and east components of groundspeed are either direct outputs of the GPS or are computed as follows:
VgN = Vg ⋅ cos(σ g ) VgE = Vg ⋅ sin(σ g )
(10.17) (10.18)
The aircraft track angle (or the direction of the groundspeed vector) is σ g . Writing down the relationship that true airspeed squared is equal to the sum of the squares of its components.
Vt = VtN + VtE
2 2
2
(10.20)
Substituting equations 10.14 through 10.16 into equation 10.20 yields equation 10.21.
(Vti + ∆Vt )2 = (VgN + VwN )2 + (VgE + VwE )2
Multiplying out equation 10.21 and collecting terms, one gets:
(10.21)
∆Vt ⋅ (2 ⋅ Vti + ∆Vt ) − VwN ⋅ (2 ⋅ VgN + VwN ) −VwE ⋅ (2 ⋅ VgE + VwE ) = (Vg − Vti )
2 2
(10.22)
132
Defining the following: a. b. c.
x = ∆Vt y = VwN z = VwE
2 2
d. C = Vg − Vti
A1 = 2 ⋅Vti + ∆Vt = 2 ⋅ Vti + x A2 = 2 ⋅ VgN + VwN = 2 ⋅ VgN + y A3 = 2 ⋅ VgE + VwE = 2 ⋅ VgE + z
(10.23) (10.24) (10.25)
Each pass produces an equation. As show in equation 10.26, subscript 1 is the first pass, 2 is the second, and 3 is the third. The unknowns x, y and z are presumed constant for all three runs. In matrix form, the equations are as follows:
A11 A1 2 A13
In matrix shorthand form:
− A21 − A2 2 − A23
− A31 x C1 − A32 ⋅ y = C2 − A33 z C3
(10.26)
[ A] ⋅{X } = {C}
The vector of unknowns the inverse of the [ A] matrix.
(10.26)
{X }
is solved by multiplying each side of equation 10.26 by
{X } = [ A] ⋅ {C}
−1
(10.27)
The unknowns x, y and z in the {X } are also contained in
[ A] .
So an iteration is
required. The initial estimates for the X values will be zero. Then, the matrix equation is used to compute a new set of X values. These values are inserted into [ A] , [ A] is inverted again, and equation 10.27 is used again. Repeat the process until convergence occurs. When the iteration is complete you have solved for the desired numbers, namely an error in true airspeed and two components of wind.
133
SECTION 10 REFERENCES
10.1. Albert G. DeAnda, AFFTC Standard Airspeed Calibration Procedures, AFFTCTIH815, Air Force Flight Test Center, Edwards AFB, California, June 1981. Olson, Wayne M. 1998. “PitotStatic Calibration Using a GPS MultiTrack Method.” Paper presented at the 29th Annual Symposium of the Society of Flight Test Engineers (SFTE), Reno, September 15. Wayne M. Olson, “True Airspeed Calibration Using Three Radar Passes,” Performance and Flying Qualities Branch Office Memo, Air Force Flight Test Center, Edwards AFB, California, August 1976. J.A.Lawford and K.R.Nipress, “Calibration of Air Data Systems and Flow Direction Sensors,” pages 1620, AGARD AG300Vol.1, September 1983. Gregory V. Lewis, “A Flight Test Technique Using GPS For Position Error Correction Testing,” National Test Pilot School, Mojave, California, July 1997.
10.2
10.3.
10.4.
10.5.
10.6.William Gracey, "Measurement of Aircraft Speed and Altitude,” John Wiley and Sons, 1981.
134
10 D = 0. e. d. Even for fighter aircraft. This is equivalent to roughly 0. Mach number. c. which add up to this 1 percent. We have errors in gross weight. A simple example will show this 1percent factor. it was a significant portion of the performance flight test program. which is roughly 1 percent in drag or fuel flow.10 L ≅ Wt D / Wt = 0. Especially for transport and bomber aircraft since most of the fuel consumed during a typical mission is during stabilized cruise. For accurate mission planning. b. a 1percent change in drag will equate to about a 1percent change in fuel flow. With modern instrumentation (as of the writing of this handbook). Range factor is specific range multiplied by gross weight. we have computed flight path acceleration ( N x ) to better than 0. the pilot made only small throttle changes between points and kept the throttle fixed at near constant flight conditions for several minutes so very long stabilization periods should not be required with modern engines.001⋅ Wt = = −0. pressure altitude.001 ⋅Wt ∆D −0. L / D = 10 or D / L = 0. and flight path acceleration. it is critical to be able to predict fuel consumption. With an INS. ambient temperature. The main sources of error are in the last two: fuel flow and flight path acceleration. we have been achieving at least 1percent uncertainty in fuel flow. Cruise testing was also the most time consuming test for transport and bomber aircraft.001 g.001 g in flight path acceleration.1 Introduction Cruise performance is usually considered the most important test performed during the performance testing phase. Specific range is nautical air miles per pound of fuel used.10 ⋅Wt For nonafterburner operation. we no longer have to spend 10 minutes to get the aircraft perfectly stabilized because we can accurately measure any small acceleration and make accurate corrections to the data.11. The primary parameters in cruise performance are specific range ( SR ) and range factor ( RF ). a typical lift to drag ratio is an even 10. The typical stabilization requirement is an airspeed change of 1 knot per minute. We strive for an accuracy of 1 percent in cruise performance.001 ∆D = 0. The other reason for 10minute speed power points was to get the engine perfectly stabilized. By using INS data. During a series of cruise points.0% D 0. This is usually required for engine and aircraft stabilization. as efforts are being made to reduce the amount of flight time spent collecting cruise performance data. There are many sources of error. The emphasis is on was.01 or − 1. A typical cruise data point can take up to 10 minutes to perform.0 CRUISE 11. fuel flow. 135 . For a transport category aircraft. a.10 ⋅Wt D = Fn − Fex N x = 0.
11.2 Cruise Tests
Cruise tests are done to determine aircraft range and endurance and to help in the development of drag, thrust, and fuel flow relationships. Cruise is a wings level, constant altitude, and constant speed maneuver. Testing is often accomplished by testing a matrix of constant aircraft gross weightpressure ratio ( Wt / δ ) points. The altitude is varied between points to yield an average Wt / δ to be a specified value. It is, however, an approximation that constant Wt / δ generalizes the data in any way. There are altitude effects on the data. The preferred method is to do constant altitude testing at varying gross weights to cover a range of Wt / δ and altitude. The data could be corrected to nominal Wt / δ values, but by correcting to weight and altitude it is easier to make flight manual comparisons. Table 11.1 represents B52G data. The G model has turbojet engines that were 1950's vintage. Table 11.1 B52G CRUISE DATA
Altitude Weight Specific Range Range Factor (ft) (lbs) (nm/lb) (nm) 35,000 400,017 0.0242 9,680 50,000 194,574 0.0437 8,503 Note: The cruise condition was 1.7 million pounds Wt / δ and Mach number = 0.76.
The average degradation in range factor for the B52G is 0.81 percent per 1,000 feet of altitude increase. In the case of the B52H model, the average degradation in range factor is 0.56 percent per 1,000 feet of altitude increase. Another data point is early F16A data that indicated about a 0.50 percent per thousandfoot degradation factor. The F16A is not a longrange aircraft and as such had a much smaller fuel fraction. Fuel fraction is the ratio of total fuel weight at engine start to empty gross weight.
Points are flown by stabilizing as nearly as possible to aim airspeed and altitude, typically ±0.01 Mach number and ±100 feet of altitude. The usual stabilization criterion is 1 knot per minute in airspeed and 50 feet per minute in altitude. With an INS to compute aircraft acceleration, the stabilization criterion could be relaxed somewhat. Typically, it takes up to 10 minutes to get the aircraft stabilized followed by 30 seconds to 1 minute of recorded data. Cruise testing is very time consuming with this method. By relaxing the stabilization criterion, considerable savings in time could be achieved. In addition, a realtime display of computed flight path acceleration could be useful in reducing the time required to stabilize.
11.3 Range
The computation of range ( R ) during cruise is the integration of true airspeed as follows:
R = ∫ Vt ⋅ dt
(11.1)
136
where:
dt = time increment (hours), and R = range (nam [nautical air miles]), 6,076.115 feet = 1 nm (1,852 meters, exactly).
We could put the range equation in different forms by making some substitutions. First, we want to put Mach number ( M ) into the equation by using the Mach number equation as detailed in the airspeed section of this handbook.
a.
M=
Vt
a
, and
b. a = aSL ⋅ θ = 661.48 ⋅ θ . Substituting into the range equation.
R = ∫ 661.48 ⋅ M ⋅ θ ⋅ dt
Defining fuel flow as the negative of the rate of change of weight:
dWt Wf = − dt
(
)
(11.2)
(11.3)
where:
Wf
= fuel flow (pounds/hour), and
dWt = incremental weight (pounds).
1 dt = − Wf
⋅ dWt
(11.4)
Substituting for equation 11.4 into equation 11.2:
661.48 ⋅ M ⋅ θ R = −∫ ⋅ dWt Wf
Making these substitutions:
(11.5)
Wf Wf = ⋅δ ⋅ θ δ ⋅ θ
(11.6)
δ = Wt / Wt δ
137
(11.7)
Wf Wf = δ ⋅ θ
Wt ⋅ W t δ
⋅ θ
(11.8)
The integration is from a start weight ( Wts ) to an end weight ( Wte ).
Wte 661.48 ⋅ M ⋅ R=−∫ Wts W f Wt ⋅ δ ⋅ θ Wt δ
θ ⋅ dt ⋅ θ
(11.9)
It’s not as bad as it looks. Canceling the θ terms and putting Wt under dt :
W 661.48 ⋅ M ⋅ t δ R=−∫ W Wts f δ⋅ θ
Wte
(
)
dt ⋅ Wt
(11.10)
If one were to fly constant Mach number and maintain constant Wt / δ , then the numerator term could be brought out of the integral. This would involve a slow cruise climb and we will show how much extra thrust that requires. At constant Wt / δ and M , the lift coefficient would be a constant. Then, ignoring the change in skin friction drag with altitude, the drag coefficient will be constant. Ignoring the thrust component, drag coefficient (as derived in the lift and drag section) is as follows:
CD = 0.000675 ⋅
( Fn / δ )
M2 ⋅S
(11.11)
Then Fn / δ will be constant, since we have assumed that Mach number and CD are constant. The corrected thrust specific fuel consumption relation is as follows:
W f δ⋅ Wf tsfc / θ = = Fn Fn ⋅ θ δ
(
θ
)
(11.12)
138
We have presumed the denominator ( Fn / δ ) to be a constant. The tsfc / θ is also considered to be approximately a constant at constant Mach number and Fn / δ . Now, we can pull these (approximately) constant terms out of the integral and integrate.
W 661.48 ⋅ M ⋅ t Wte δ dt R=− ∫ W Wts Wt f δ⋅ θ
(11.13)
(
)
The term in front of the integral is called range factor ( RF ).
R = − RF ⋅
Wte
Wts
∫W
dt
t
(11.14)
You may be more used to seeing RF in the following identical form:
RF =
where:
Vt ⋅ Wt = SR ⋅Wt (nautical air miles) Wf
(11.15)
SR = specific range (nautical air miles per pound of fuel).
From a table of integrals and natural logarithm relationships:
∫
a
b
dx = ln b − ln a = ln b = − ln a a b x
( )
( )
where:
ln = natural logarithm. W R = RF ⋅ ln ts Wte
(11.16)
The above equation is convenient to get a quick estimate of range given only the average range factor and the start and end cruise weight. Note that this is the range during the cruise segment and does not include taxi, takeoff, climb, and descent.
11.4 Computing Range from Range Factor
Using the previous tabulated B52G data, we will compute range and show the magnitude of the climb factor. We will assume that the two points at 35,000 and 50,000 feet are the beginning and end of the cruise segment of a mission. The cruise is at constant 0.77 Mach
139
number and a Wt / δ of 1,700,000 pounds. Using previously defined formulas for true airspeed, energy altitude, and pressure ratio we construct Table 11.2. We will linearly interpolate versus altitude for range factor. Table 11.2 RANGE FACTOR VERSUS ALTITUDE FOR B52G
True Energy Gross Net Range Altitude Airspeed Altitude Weight Thrust Factor (ft) (kts) (ft) (lbs) (lbs) (nm) 35,000 443.84 43,721 423,547 42,355 10,843 36,089 441.65 44,724 402,052 40,205 10,777 40,000 441.65 48,635 333,155 33,316 10,539 45,000 441.65 53,635 261,986 26,199 10,234 50,000 441.65 58,635 206,020 20,602 9,930 Note: Thrust was computed by assuming a lift to drag (L/D) ratio of 10. This is typical for a transport category aircraft.
We could get a first estimate of range by using an average range factor and the start and end conditions.
W (9, 680 + 8,503) 400, 016 R = RF ⋅ ln s = ⋅ ln = 6,552 nam 2 194,574 We
(11.17)
Since we assumed a linear variation of range factor with altitude, we will get the same result by integrating the individual segments. Range factor will not be a linear function of altitude, usually. The time for this mission computes to be 54,100 seconds (15.04 hours). From the table, the delta energy altitude is 14,914 feet. The average speed is 736.5 feet per second. Now, we can calculate the average longitudinal load factor necessary to produce enough excess thrust to sustain this cruise climb.
Nx =
! HE Vt
(14,914 ) (51, 000 ) = 0.2955 = 0.00040 =
736.5 745.6
(11.18)
At the average weight of 297,295 pounds, the average excess thrust calculates to 119 pounds. The average thrust is 29,730 pounds, therefore the ratio of excess thrust to net thrust is: a.
Fex 119 = = 0.0040 or 0.40% Fn 29, 730
By ignoring the excess thrust, we over estimated the range by 26 nam (0.40 percent of 6,552 nam). Quite small, but not negligible. On an actual mission, the mission profile would be step climbs. For this example, you would start the cruise segment at 35,000 feet and fly
140
mid. The altitudes are then rounded to the nearest 5. and finally doing an idle power descent. When flying in civilian airspace. Nevertheless.000 feet. progressively stepping up in the altitude during constant altitude/Mach number cruise segments. and descent.6 Range Mission Range missions are performed to gain confidence in the performance data collected during climb. This might be in increments of 4.000foot increments for standard weights. representing rounded values of heavy.constant altitude until it was decided to climb to a new altitude.000foot increments. Texas. A practical reality of the flight test programs was that it was difficult to justify devoting an entire sortie to only a range mission.5 Constant Altitude Method of Cruise Testing The recommended method of doing cruise testing is the constant altitude method. The altitudes are chosen by selecting six evenly spaced Wt / δ ’s from minimum to maximum with one at the predicted optimum.000 feet apart. this may mean flying up to six altitudes at up to three weights (heavy. The mission is performed by climbing to a given start cruise altitude. For an aircraft with a large weight fraction. and light). This could mean a maximum of 18 weight/altitude combinations. A calibration of the fuel quantity system is obtained during the aircraft empty weight and fuel calibration. Flying all three weights at the predicted optimum cruise Wt / δ is usually desirable. A comparison of fuel used was made with simulation results. During the B1B project. the primary measurement during a range mission is aircraft fuel quantity indications. 11. the test day mission performance could be estimated. with a reasonable amount of thrust/drag/fuel flow analysis. The F15 and F16 projects used constant altitude method. the altitudes are 4. The data came from constant airspeed/altitude segments of several hours in duration. The constant altitude method consists of choosing a range of weight and altitude conditions to cover the aircraft envelope and then flying each weight/altitude combination over a range of speeds. mid. The minimum is based upon minimum weight at a minimum altitude and the maximum is based upon the cruise ceiling defined as a climb capability of 300 feet per minute. 11. the data presented in reports are a specific range. and descent. this could be cut in half or more. though the points were flown using the constant weight/pressure ratio ( Wt / δ ) method. The differences were well within the oftenquoted 3percent accuracy for performance data.000 feet. fuel used data were acquired from several training sorties flown on production aircraft at Dyess AFB. The simulation estimates of fuel used were compared with measured fuel used during the mission. which allows for easy flight manual comparisons since flight manuals typically have cruise charts at even 5. 141 . cruise. The simulation thrust/drag/fuel relationships were previously determined using data from several maneuvers including climb. The B1B used constant altitude analysis method. This provided a valuable confirmation of the flight test results. Using a performance simulation. cruise. and light gross weight. Total fuel used is obtained from the fuel quantity system. A compromise was to obtain fuelused data during long cruise segments that often occurred during certain systems tests. Rather than relying on fuel flow measurements and thrust/drag analysis. or range factor versus Mach number at even 5. For ease of flight manual comparisons.
7 Slow AccelDecel A supplement. If the cruise tests are done with a relaxed stabilization criterion (±100 feet and ±2 knots in 20 seconds) with only 20 seconds of recorded data. SR = Vt Wf (11.85 Mach number. is to do slow accels and decels. Vt = 661.8 Effect of Wind on Range The typical high altitude cruise for both fighter and transport aircraft is about 0. We could estimate the zero excess thrust range factor from both the accel maneuver and the decel maneuver. a.65 degrees K.85 ⋅ 216. The true airspeed for standard day in the lower atmosphere (troposphere) and upper atmosphere (stratosphere) can be computed using formulas from the airspeed section. The data are used to build or verify a thrust versus fuel flow model. but not so small that the maneuver would take too long.19) We can compute a specific range with respect to the ground as follows: SRg = Vg Wf (11. thrust and fuel flow data over a range of conditions then the slow acceldecel is a good approach.65 = 487. or perhaps even an alternative to cruise testing. For standard day from 11 kilometers (36. The average of the accel and decel standardized range factors is a good estimate of zero excess thrust range factor since relatively small corrections are being made. This compared with 1percent corrections made to cruise data.11.20) Since groundspeed equals true airspeed minus wind and taking just the component parallel to the direction of flight (track angle): Vt = Vg + Vw (11. The maneuvers are flown sufficiently slowly to make the maximum correction to a range factor of about 10 percent. thereby losing the advantage over stabilized cruise.5 knots 288.15 The formula for specific range (nams per pound of fuel) is just true airspeed ( Vt ) over fuel flow ( W f ).617 feet). The maneuver is done at a rate of less than 1 knot per 3 seconds to yield an accel/decel rate of about 20 times the cruise stabilization criterion. then the dynamic cruise has an advantage over the slow acceldecel data. the data could be standardized to zero excess thrust. the temperature is 216. If it is desired to collect. A typical accel/decel maneuver takes about 6 to 12 minutes. In addition.089 feet) to 20 kilometers (65.21) 142 . 11. The throttle is moved in small increments during the run to keep the accel/decel rate small.48 ⋅ 0.
A headwind is a positive wind.5 − 40 ) 487.5 percent improvement In general.661 feet) averages about 40 knots above Edwards AFB. For this condition.5 = 0. the range degradation would be: a. In addition. SRg SR = ( 487.5 percent degradation One could just as easily be flying with that wind as a tailwind. A set of data collected for the cloverleaf paper (a portion of which is in the cloverleaf subsection of the air data system calibration section) had winds in excess of excess of 100 knots. 143 . This wind effect is only relevant in computing physical (ground) nautical miles with a given wind.22 divided by equation 11.23) As shown in Appendix A.19) is as follows: SRg SR = (Vt − Vw ) Vt (11. Flying directly into a 100knot wind would produce the following specific range degradation: a. the wind data shown Appendix A indicates a standard deviation of about 25 knots. an aircraft heading of 215 degrees would have a 40knot headwind for this average Edwards wind.2). SRg SR = ( 487.SRg = (Vt − Vw ) Wf (11.2 percent degradation This is for an average wind if one were heading directly into the wind. windspeed at an ambient pressure of 200 millibars (mb) (38.795 = 20. The average direction is about 215 degrees (SW).5 = 1.22) Finally. the ratio of specific range with respect to the ground to the specific range with respect to the moving air mass (equation 11.5 + 100 ) 487. When collecting cruise data. you would only be affected by the component of wind parallel to the flight direction.5 − 100 ) 487. you are flying with respect to the moving air mass. but was AFFTC data set number 2 in the referenced paper (Reference 10.5 = 0.11. Since wind direction is the direction from which the wind is blowing. SRg SR = ( 487.205 = 20.918 = 8. a. This data were not included in this handbook. Wind vector relationships are discussed in detail Section 10.
First. distance. and distance to accelerate. The fairing is the result of modeling thrust and drag. These are discussed in the turn section.12. then computing specific excess power from thrust and drag. Climb data can be used for this purpose also. The data points have been corrected to standard conditions. and fixed throttle at constant altitude. The altitude is maintained constant during the run. Once through the transonic jump. 144 . depending on whether it is desirable to achieve a minimum time to climb or minimum fuel for fixed range.1 Acceleration Accelerations are conducted for multiple purposes. Then the aircraft accelerates to a point where the acceleration rate is reduced to a small value (less than 1 knot per 10 seconds)..1 is a sample of some actual acceleration data. fuel used. it is necessary to maintain zero flight path angle usually by maintaining pitch attitude ( θ ). an indicated altitude could be used for the rest of the acceleration. Standard conditions consist of standard weight. Figure 12. Modern aircraft with a headup display (HUD) and INS have a velocity vector displayed on the HUD. to determine the obvious acceleration performance. a few continuous climbs need to be conducted to confirm that performance (time. 1g. i. to determine optimum climb schedules by observing the peak of specific excess power versus Mach number. Level flight through the transonic region is obtained by maintaining the velocity vector on the horizon. Indicated altitude will jump as the aircraft passes through the transonic speed regime. and fuel used) computed from accelerations yields the same result as that from climbs. The acceleration maneuver is performed wings level. By performing accelerations at various altitudes. Third. one obtains a range of speed (Mach number) at a given altitude. The actual optimum occurs to the right of the peak of specific excess power ( Ps ) versus M curves. With one relatively short maneuver. to determine drag/thrust/fuel flow models. The accelerations are conducted over a range of altitudes. and standard day atmospheric conditions. However.e. Second.0 ACCELERATION AND CLIMB 12. climb performance can be computed. however. Accelerations are also performed at elevated g levels. pressure altitude. accelerations are a more efficient method. time. Usually a climb or turn is done at the beginning of the run to get the engine thermally stabilized. Thus.
and standard normal load factor. which include accelerations.1 Specific Excess Power from Acceleration 12.250 2. The definitions are from the flight manual specification. there may be differences in performance accelerating through a Mach number/pressure altitude condition versus climbing through the same condition. For high performance aircraft. The normal climb is a constant calibrated airspeed climb to a break altitude above which a Mach number is maintained constant. time to climb.2 Climb The climb maneuver is performed primarily as a check of predicted climb performance derived from acceleration data.50 Standardized Data Fairing from Thrust/Drag Model 0.Specific Excees Power (ft/min) versus Mach Number 4.750 3.65 0. The data are reduced at constant altitude increments rather than constant time increments to yield a more even distribution of data. 145 .500 Specific Excess Power (ft/min) 3.70 Mach Number 0. standard day.250 3.1. The climb continues to a climb ceiling (300 feet per minute rate of climb defined as the cruise ceiling). Thrust and drag data are obtained during the climb.85 0.60 0. Climbs are usually terminated at the “cruise ceiling. standard weight.000 3. and distance traveled are plotted versus pressure altitude. gross weight. Usually climbs are conducted at flight manualpredicted best climb speeds.750 2.55 0. Data are standardized to the climb schedule. A standard day rate of climb.000 0.75 0.000 2. This effect needs to be taken into account. A flight manual comparison is accomplished with this data. Determination of actual best climb speeds requires an analysis using data from several sources.80 0. fuel used.” Climb ceiling definitions are given in Table 12.90 Figure 12.500 2. This is due to an engine fuel control system lag.
For each speed. A comparison was made for an average day above Edwards AFB in January.1 CLIMB CEILING DEFINITIONS Ceiling Combat Cruise Service Absolute Rate of Climb (ft/min) 500 300 100 0 12.605 to 23. one can expect to see large changes in windspeed and direction as a function of altitude.1).2) is shown Figure 12. or a 14percent difference in rate of climb. 146 . The difference in delta energy altitude flying directly into a headwind versus flying directly into a tailwind was 1. A sample of some actual flight test sawtooth climb data from an AC119G (Figure 12.000 feet pressure altitude. for instance.500 feet. How this would impact climb performance was discussed in the effect of wind gradient portion of the altitude section. climb data were attained using the sawtooth climb method. Data were obtained from FTCTR694. Depending upon the performance level of the aircraft. The sawtooth climb tests are a series of alternate heading climbs through a given altitude at a range of speeds. A zero acceleration rate of climb is the rate of change of energy altitude. the aim altitude might be 5. perhaps 4. the aircrew would establish a climb speed and climb power at that altitude and would collect data over an established data range. a start altitude would be determined.937 feet.000 feet and repeat the same airspeed point. For instance. This was one of the last AFFTC projects where sawtooth climbs were flown. Then. The thrust designation METO on Figure 12.Table 12.3 denotes Maximum Except for TakeOff. This was over a geometric altitude range from 14. AC119G Aircraft Limited Performance and Stability and Control Test (Reference 12. you would correct the data to zero acceleration.3 Sawtooth Climbs As seen in Appendix B. you would descend back to the initial altitude of 4.3. Using the acceleration factor. Then. a climb would be conducted through the aim altitude and airspeed and altitude data would be collected versus time. Before the advent of accelerometer and INS methods. The idea here is that the average of these two points would be a zero wind gradient condition. Then test points would be chosen over a range of speeds to bracket the expected best climb speed. but this time at an opposite heading angle (based upon magnetic compass).308 feet.500 to 5.
2 AC119G Aircraft Sawtooth Climbs: AC119G Cruise Configuration METO Power 1. and present them in a different manner.300 1. 66. ! Fex = N x ⋅Wt = h ⋅Wt Vt (12. Since we had two altitudes and two weights.000 lbs opposite heading 5.100 1. Then. 66. let us attempt to minimize the weight effect in the data by computing the excess thrust.000 lbs opposite heading 10.4. The data are presented in Figure 12.1) ! The h is the zero acceleration rate of climb in Figure 12.000 ft. 58.2).000 ft.Figure 12. without distinguishing opposite headings.000 900 800 700 600 500 400 300 200 90 100 110 120 130 140 150 160 170 Calibrated Airspeed (kts) 10. The specific algorithms used to standardize that data can be found in AF TR No.3 AC119G Sawtooth Climb Data We can take these data points.000 lbs opposite heading 5. 6273.000 lbs opposite heading Figure 12. 58. 147 .200 Zero Acceleration Rate of Climb (ft/min) 1. take the excess thrust and divide by the pressure ratio ( δ ) to minimize the altitude effect.000 ft.3. Flight Test Engineering Handbook (Reference 12.000 ft.
For high performance fighters (with installed thrusttoweight ratios greater than 1) the initial part of the climb may be done at a constant pitch attitude ( θ ) transitioning to a Mach number at a given altitude.000 lbs: 10.4 Continuous Climbs A climb could be done with any number of different climb schedules.000 4. but they do not vary during the climb.24 0. Alternatively. The most common type of climb is one that keeps calibrated airspeed ( VC ) constant until a given Mach number ( M ) is reached at which time Mach number is kept constant. a constant true airspeed climb.000 0. These types of climbs are required for high performance fighters when the aircraft has a longitudinal acceleration load factor greater than 1.000 lbs: 5.000 1.22 0.18 0.000 7.000 ft 6.28 0. The flight path angle for the constant θ climb is as follows: γ = −α + θ (12.000 Excess Thrust/delta (lbs) 5. In contrast.Sawtooth Climbs: AC119G: Fex/delta versus Mach Number 8.000 ft 66.000 lbs: 10.000 pounds gross weight at sea level the recommended schedule is 181knots calibrated airspeed while at 20.30 Mach Number Figure 12.000 feet the climb speed is down to 166 knots. the early part of the climb may be performed at less than maximum thrust. A variation on that schedule is one in which calibrated airspeed is a function of altitude.000 3.16 0. or a varying Mach number climb. The C130H climb schedule is an example of a varying calibrated airspeed climb. At 150. both calibrated airspeed and Mach number may have been a function of gross weight ( Wt ).000 58.4 AC119G Excess Thrust Data 12.26 0.20 0. 148 .2) Other types of climbs are variable climb schedules such as a varying airspeed schedule. Usually.00 and can accelerate flying straight up. most aircraft use a constant calibrated airspeed/Mach number climb schedule.000 ft 58.000 ft 66.000 lbs: 5. A climb schedule is a speed or attitude variation with altitude.000 2.
Figure 12. AF = ! HE ! H Most aircraft climbs are conducted by either holding calibrated airspeed ( VC ) or Mach ( M ) number constant. The longitudinal load factor ( N x ) is simply the Ax divided by the acceleration of gravity ( g 0 ). 149 . though there could have been a MIL power deceleration under certain conditions such as supersonic.Accelerations and climbs are both fixed throttle maneuvers. and AF = acceleration factor. g is assumed a constant at the reference gravity and given the value of 32.5 is a representation of acceleration factor for climb at constant calibrated airspeed. a. The true airspeed vector defines the flight path (or wind) axis. 12. The component of aircraft acceleration parallel to the flight path is the longitudinal acceleration ( Ax ).3) 12. Decelerations and descents are usually done in power settings such as IDLE.174 ft/sec². In reality. (12. the calibrated airspeed or Mach number is not exactly constant but let us make some calculations assuming that they are held exactly constant and that there is zero wind so that true airspeed ( Vt ) and inertial speeds ( Vg ) are identical. In conventional aircraft performance.5 Climb Parameters ! R/C = H V dV AF = 1 + t ⋅ t g0 dH where: R / C = rate of climb (ft/sec). They are usually done with power settings like MIL or MAX.6 Acceleration Factor (AF) The acceleration factor ( AF ) is used in climb performance as a simple conversion between a rate of change of tapeline or geopotential altitude and rate of change of energy altitude.
000 10.000 feet. The two cases will cover the range from a highspeed.000 30. 2. H = 30.5 at 36. using the equations found in the airspeed section of this text.Constant Calibrated Airspeed Acceleration Factor 6.0 1. The above chart is for a standard atmosphere.6.0 4. 12.000 50.48 kts Vc = 700 kts Figure 12.0 3.5 3.000 35.000 40. and M = 0.000 15.0 2.5 1. For standard conditions.000 45.0 0 5.000 20.1 Case 1 High speed. high altitude.1 Two Numerical Examples for AF To illustrate the importance of the concept of AF .5 Acceleration Factor – Constant Calibrated Airspeed The discontinuity in Figure 12. highaltitude fighter to a lowspeed. assume the following flight conditions: 1. The following additional significant figures are necessary to make the computations accurately: 150 .5 2. high performance typical of a fighter type aircraft: a.000 Pressure Altitude (ft) Acceleration Factor Vc = 100 kts Vc = 200 kts Vc = 300 kts Vc = 400 kts Vc = 500 kts Vc = 600 kts Vc = 661.6.5 5. let us illustrate AF by two numerical sample cases.1. Please note that we are listing the numbers to at least one more significant figure than our limits of flight test data accuracy.000 25. lowaltitude aircraft.5 4.900.089 feet is due to the transition from a temperature decreasing with altitude to a constant temperature. we could compute the values for calibrated and true airspeed. 12.0 5. For case 1.
19 ) 2 V ∆V AF = 1 + t ⋅ t = 1 + 32.).1 feet second. the rate of climb at a constant Mach number is 1. Below 36.31 + 895.31 − 895. The difference in rate of climb between holding constant Mach number versus constant calibrated airspeed is 52 percent.9 percent. 12.90 knots.19 ) ⋅ = 0. R/C = ! HE 200 = = 224. a constant calibrated airspeed climb would be accelerating in true airspeed and hence. b. 000 ) (891. and 2.6.000 feet holding a constant calibrated airspeed of 340 knots.09 knots = 891.8923 For a climb through 30. Vt = 530.3 feet per second.31 feet/sec (Note that the aircraft is decelerating while climbing at a constant Mach number.2 Case 2 The second case is what is a typical climb for a light aircraft.089 feet in the standard atmosphere in a constant Mach number climb. the AF computes to 1.089 feet in the standard atmosphere. and 2.1. rate of climb would be less than the specific excess power.1 AF 0. This is small. Above 36. 000 − 30.24 knots. the true airspeed would decrease with increasing altitude (Figure 12.8. At a Ps of 1. Now we could numerically calculate the AF : dVt / dH = ∆Vt ∆H (891.7 fpm and the rate of climb at constant calibrated airspeed is 984. VC = 338.174 g 0 ∆H For a Ps of 200 feet per second.8923 (31. Then.089 feet. but not small enough to ignore. Assume a 100knot calibrated airspeed climb through 5.1.3576 for a R / C of 147.000 fpm.000 feet.000 feet and 0. The percentage difference gets proportionately smaller at lower airspeeds.003. This illustrates how large an effect the acceleration factor could be and that it certainly needs to be taken into account. Vt = 528. The difference in rate of climb between a constant calibrated airspeed and a constant Mach number climb is now down to just 1. the R / C would be 224. At 31. VC = 346.6).39 knots = 895. when temperature is a 151 .19 feet/sec.900 Mach number: 1. below 36. Conversely.
The acceleration perpendicular to the flight path (ignoring gravity) is a centripetal acceleration. Figure 12.90 0. Hence. The first velocity vector ( Vt ) is at a flight path angle of γ 1 . with a radius R .92 0.00 at all Mach numbers.8 0. 152 . The magnitude of the change is exaggerated. Keep in mind that Figure 12.6 Acceleration Factor – Constant Mach Number 12.0 0. the acceleration factor would be 1.94 0. The second Vt is at γ 2 .089 1.7 Normal Load Factor During A Climb To derive the formula for the normal load factor in a climb. Accelertion Factor: Constant Mach Number: H<36.7 0.1 0.3 0.6 0. the true airspeed is a constant for a constant Mach number.2 0.5 Mach Number 0.0 Figure 12.98 0.4 0.86 0.84 0.6 is for standard day.7 illustrates the vectors during a pullup.96 Acceleration Factor 0.constant with altitude for the standard atmosphere.88 0.9 1.00 0. The aircraft rotates about a point C . consider the aircraft flying in a pullup maneuver. but consider the change infinitesimal.
C Figure 12.8) N z = cos γ + Vt ⋅ γ! g0 (12.5) (γ 2 − γ 1 ) (t2 − t1 ) = γ! (12. dividing by g 0 yields the load factor in the normal axis. ω = dγ dt = ∆γ ∆t = (12.7 Centripetal Acceleration Diagram The centripetal acceleration is as follows: V a= t R 2 (12.7) Adding in the component of gravity yields: a = g 0 ⋅ cos γ + Vt ⋅ γ! Finally.6) Solving for the radius R in equation 12.5 and substituting into the acceleration equation 12. (12.9) 153 .4: a= Vt = Vt ⋅ γ! Vt γ! 2 (12. Vt = ω ⋅ R The angular velocity ω is just the derivative of the flight path angle.4) The radius is related to the linear velocity through the angular velocity ( ω ).
and drag polar information to complete the performance model. Russel M. James K. California. 154 . FTCTR694. the difference between a tapeline rate of climb and pressure altitude rate of climb is taken into account.2 Herrington..9 Deceleration Decelerations are conducted to provide data to compute descent performance. Flight Test Engineering Handbook. The performance model is used to check mission performance. and the appropriate software to do the axis transformations. Usually. sensitive to the presence of any vertical winds. March 1969. Edward D. Edwards AFB. Edwards AFB. Due to inaccuracies in the inflight thrust deck. The idle power descent could be accomplished with speed brakes extended. 12. The data are used to generate descent performance. AC119G Aircraft Limited Performance and Stability and Control Tests. The INS data is. since the INS yields geometric rate of climb. so efforts are made to fly in areas where no vertical winds are expected. the best place to conduct performance tests is over the ocean. and McDowell.0 REFERENCES 12. then there is no need to be concerned about horizontal winds and wind gradients.8 Descent A typical descent schedule is a constant Mach number intersecting a constant calibrated airspeed. SECTION 12.1 Pape. AF TR 6273. if the aircraft has an INS. The same maneuver could be accomplished with the speed brakes extended. For Edwards AFB.The above equations are valid for constant winds. Therefore. 12. constant altitude deceleration. revised January 1966. What is desired are inertial accelerations in the wind (or flight path) axis. an idle thrust map. AFFTC. there could be a drag difference at idle thrust versus drag polar data acquired at higher power settings.. This maneuver gives us idle thrust versus speed. In addition. the load factors are computed from INS velocities and angles plus true airspeed to enable a transformation from the inertial axis to the flight path axis. 12. however. et al. Both the B1B and C17A aircraft conducted their entire cruise testing over the ocean. California. AFFTC. A deceleration is performed by accelerating to the Mach number limit then moving the throttle to idle and conducting a wings level.
With fuel controls scheduling on total temperature in the inlet. Usually about 30 seconds or 180 degrees of turn data are recorded at stabilized conditions. The thrustlimited turn is stabilized at a given Mach number and pressure altitude condition. however. in general. applying specified throttle. One value of thrustlimited turns is it produces thrust data that is stabilized while accelerations and decelerations are dynamic in thrust. and liftlimited.1 Introduction Turning performance is defined as flight at other than 1 g. and fixed throttle are referred to as accelerating or decelerating turns. may be of higher quality than data from dynamic maneuvers.3 ThrustLimited Turns A thrustlimited turn is a turn where the pilot attempts to maintain throttle setting. constant altitude.2 Accelerating or Decelerating Turns Accelerating or decelerating turns are performed at a fixed throttle. Turns at fixed g. usually in the horizontal plane. For accelerating turns. Not relying completely upon data obtained from dynamic performance maneuvers is important. It is still necessary to perform a limited number of turns as checks on the model. are used to quantify the turning performance capability of the aircraft and to help in the development of the drag and lift curves. 13. So. There may have been throttle effects on the drag polar due to inaccuracies in the inflight thrust computation. Mach number. There are four different types of turns: accelerating or decelerating. 155 . The data acquired could be used to generate energy maneuverability charts or to contribute to the aircraft drag. the maneuver is done by starting fast. The data are used to verify the thrust/drag model for sustained g and to assist in the development of the drag and lift curves.0 TURNING 13. fewer turns are conducted in flight test. the lag time constant for thrust could be estimated. With the advent of dynamic performance. Next. The data are collected at a stabilized g and as such. As with accelerating or decelerating turns. Nevertheless. Nevertheless. and pressure altitude while varying normal load factor. thrust. and constant altitude. Turns. Turns are used primarily to check the performance model created from 1g acceleration and dynamic performance maneuvers. 13. only a limited number of sustained or thrustlimited turns are performed because they are very fuel and timeconsuming tests compared with the more efficient dynamic maneuvers. reduce g level to the specified value and accelerate to either the specified Mach number or the maximum speed. It has been necessary on past projects to do significant numbers of turns because of disagreements between turn data and dynamic data on the drag polar. keep in mind that the thrustlimited turn is dynamic since it is at elevated g values (and large pitch rates) and may be at different power settings than the dynamic performance data. stabilized. thrustlimited. some turns are still necessary as confidence builders in the model and to demonstrate specification performance. and pulling into a turn to decelerate the aircraft. and fuel flow model. constant g.13. Developing correlation factors to adjust the drag polars to match the measured turn performance may be necessary. maintaining stabilized conditions is often difficult. there may be a different lag constant depending on whether the aircraft is climbing or accelerating through a point.
The throttle setting is usually MIL or MAX. Specs are usually written for standard day at a standard weight. pressure altitude. you would have needed some specialized software to perform the standardization or the turn could have been standardized assuming it is an accelerating turn at a given pressure altitude.Using an INS for flight path accelerations requires a 1g level run be accomplished before the turn to get a wind calibration. Besides getting limit performance. then the spec condition was met. Therefore. The check of angle of attack is performed with INS data. then the turn is a full aft stick maneuver. Mach number. The flight test objective is to determine if such conditions can be achieved in stabilized flight at something less than or equal to maximum throttle. as is the case on the F16. These winds are assumed to remain constant during the turn. These maneuvers are useful to obtain lift and drag data at specific points along the drag and lift curves and to check for specification compliance.5 LiftLimited Turns When it is desired to determine limit performance at the angleofattack ( α ) limit or the normal load factor ( N z ) limit then a liftlimited turn is performed. which makes this maneuver very difficult to fly without exceeding aircraft limits. The angleofattack limited portion of the maneuver is used to quantify the lift coefficient at the limit angle of attack and to check the angleofattack calibration at the limit. This maneuver produces data at the highest limits of the drag polar and the lift curve. the pilot must observe the flight manual limits. center of gravity. Another way to evaluate that spec point would be to do a thrustlimited turn at MAX thrust at the specified flight conditions and then determine whether the desired normal load factor in stabilized flight is achieved. 13. For the stabilized turn. etc. If the longitudinal flight path load factor ( N x ) was positive for the given spec conditions.4 Stabilized Turns Stabilized turns are turns where Mach number. 13. and normal load factor. Winds are computed from the wind calibration maneuver assuming zero sideslip. drag polar data at or near maximum lift coefficient are obtained. Otherwise. It is possible to attain lag time constants by comparing thrustlimited turn data to climb and acceleration data. You also obtain limited angleofattack data from a splits. If the aircraft has an α/g limiter. then determine the flight path acceleration for standard conditions. The splits maneuver is discussed in the dynamic performance section. but the maneuver may be done at any power setting. This defines the lift limit and g limit performance. This applies to all turning maneuvers. 156 . Liftlimit and glimit turns are accomplished by accelerating to limit speed then pulling into a maximum allowable g turn and allowing the aircraft to decelerate to the lift limit. you must correct the data to standard conditions to determine spec compliance since the spec may have been missed on the test day but the aircraft would have achieved the spec on a more favorable standard day. The thrust and fuel flow data obtained in climbs and acceleration is dynamic and subject to engine and instrumentation lag. and normal load factor are specified and throttle is varied to obtain a stabilized condition.
constant speed turn. Nz N zv N zh Figure 13. for the ideal case of exactly constant altitude. Two. the horizontal component is a centripetal acceleration.2) Nx Nz N xb N zb = flight path axis longitudinal load factor. and = body axis normal load factor.6 Turn Equations 13. One.1 shows these vectors. the normal load factor in the wind (flight path) axis system in terms of the turn rate can be derived in a similar manner as the formula for normal load factor in a climb.1) The additional sideslip transformation matrix is given in the Accelerometer Methods subsection of the Flight Path Accelerations section. Figure 13.13.3) 157 .1 Normal Load Factor The transformation equations for load factors from the body axis system to the flight path axis are as follows (ignoring sideslip): N x cos α = N z − sin α sin α N xb ⋅ cos α N zb (13. The inverse transformation from the flight path axis to the body axis is as follows: N xb cos α = N zb sin α where: − sin α N x ⋅ cos α N z (13. the vertical component is exactly 1.0. There are two components.6. = body axis longitudinal load factor.1 Normal Load Factor Vectors In a Turn N zh = Vt ! ⋅σ g g0 (13. For a constant altitude. = flight path axis normal load factor.
5) What both of the N z equations have in common is that they rely upon unrealistic idealizations of zero wind and exact constant altitude and speed. either accelerometer methods or INS methods are used to compute the actual flight path axis load factors. Nz = 1 cos φ (13.Nz = (N 2 zv + N zh 2 ) V 2 ! 2 = 1 + t ⋅ σ g g0 (13.2 Banked Turn Diagram Where: N zv = 1. 158 . Figure 13.4) Where σ g is the ground track angle and the assumption of zero wind is made. and N cos φ = zv = 1 .0 . In flight test. the normal load factor in terms of the bank angle can be determined as shown in Figure 13. Nz Nz Hence. constant speed. and zero wind.2. With the same idealized assumptions of constant altitude.
10) 159 . we can compute the turn rate. just the right amount to maintain exactly constant altitude for this idealized relationship. Azh = where: Vt (feet/sec2) R 2 (13. The relationship derives from the kinematics of constant speed rotation about a point.8 into equations 13.6 and solving for R : a. level turn the centripetal acceleration is the horizontal component of normal acceleration. (N 2 z −1 ) (13.6. From trigonometry: N zh = and.9).7 and equations 13.653 ⋅ 2 (N 2 z −1 ) (13.6878 32.174 ⋅ 2 (N 2 z −1 ) For R in feet and Vt in knots: R= Vt 1.13. and Azh = horizontal component of normal acceleration (ft/sec2). Vt = ω ⋅ R where: (13. Vt = true airspeed (ft/sec). The vertical component is 1g.7) N zh = Azh g0 (13.7 Turn Rate Once the turn radius is determined (equation 13. R= g0 ⋅ (N Vt 2 2 z −1 ) = Vt 32.6) R = turn radius (ft).2 Turn Radius In a steady.9) 13.174 ⋅ 2 (N 2 z −1 ) = Vt 91.8) Substituting equations 13.
it has been possible to compute accurate values of air data parameters in dynamic maneuvers such as turns. Then.1 m/sec (0. solving for turn rate: ! V a.11) " " " Vt = Vg + Vw Solving for the magnitude of the true airspeed vector: (13. For R in feet.19 knot).R = radius of turn. The INS gives you six parameters of interest for performance and flying qualities.2958 = 33. and the pitch attitude designated theta ( θ ). ! The symbology we previously used for turn rate was σ . These are three angles called Euler angles and three velocities in the north ( N ). The groundspeed components from an INS are VgN . we will proceed to develop the equations starting with the basic vector relationship of true airspeed. we have introduced a new level of accuracy into our data. east ( E ) and down ( D ) directions. Now. The Euler angles are the heading from true north designated psi (ψ ). σ g = t R ! The above equation is valid for units of R in feet. We didn’t! The typical drift rate of an INS was on the order of 1 nautical mile per hour. and VgD . (13. In addition. This combines the outputs of an INS with the velocities and position data from the GPS using a filter. an accurate value of groundspeed was available. Now (late 1990s) we have a new device designated as embedded GPS/INS (EGI). However.12) 160 .8 Winds Aloft Since the advent of the INS in the 1970s. and wind. the roll (or bank) angle designated phi ( φ ). Therefore. INS data had small drift errors in the groundspeeds.947 ⋅ Vt ! σg = R (R) 13. this required the use of wind calibration runs conducted in wingslevel 1g flight where the air data system errors were known from conventional tests. groundspeed. VgE . The mathematics and illustrating data for one such technique used in turning flight (that does not require the use of a wind calibration) will be presented. This small error does not drift with time. we had typical errors of about 1 knot in the horizontal groundspeeds at any one time. Vt in knots and σ g in degrees per second we get: Vt 1. and ω = turn rate. The problem is that we assumed we knew the groundspeeds accurately. The GPS specification accuracies for the horizontal speeds are 0. With the availability of the GPS in the 1990s.6878 ⋅ 57. Vt in feet per second and σ g in ! radians per second. the rate of change of ground track angle. Therefore.
we will simply strive to minimize the sum of the difference between the left and right side of the above equation.15) The 0. VwE . g and h : (13.14) From here on in the derivation. Vtx = VgN + VwN Vty = VgE + VwE Vtz = VgD Vt = Vti + ∆Vt Defining three more parameters: f . ∆Vt ).18) (13. b.17) (13. Defining a parameter we shall call F* (F – star).13) We will assume the vertical wind is zero.19) f = ∑ Fi ⋅ Vtx * i =1 N N (13.5 factor is just to eliminate ½ factors in the final formulation.2 2 2 Vti + ∆Vt = (VgN + VwN ) + (VgE + VwE ) + (VgD + VwD ) (13.21) h = ∑ Fi ⋅ Vt * i =1 N (13. and 161 .20) g = ∑ Fi ⋅ Vty * i =1 (13. The iteration is the method of Taylor’s series in three dimensions: F * = 0. we want to minimize the sum of this parameter simultaneously with respect to each of the three unknowns ( VwN .5 ⋅ Vtx + Vty + Vtz − Vt 2 2 2 ( 2 ) (13. y. y = VwE . Taking the square of both sides: (Vti 2 2 2 2 + ∆Vt ) = (VgN + VwN ) + (VgE + VwE ) + VgD (13. z unknowns are as follows: a.22) There are N data points and N must be at least three. The x. x = VwN .16) (13.
initialize f . ∂g / ∂x = ∂g / ∂y = ∂g / ∂z = 0 . x= y=z=0 In addition.30) 162 . z = ∆Vt . h and the partial derivatives to zero as follows: a. g . f = g =h=0. ∂f / ∂x = ∂f / ∂y = ∂f / ∂z = 0 .27) ∂g / ∂z = ∑ ( −Vt (i ) ) ⋅ (Vty (i ) ) N i =1 (13.24) ∂f / ∂z = ∑ ( −Vt (i ) ) ⋅ (Vtx (i ) ) i =1 N N (13. We will assume zero initial estimates for the unknowns.26) (13.29) ∂h / ∂y = ∑ (Vty (i ) ) ⋅ (Vt (i ) ) i =1 (13.25) ∂g / ∂x = ∑ (Vtx (i ) ) ⋅ (Vty (i ) ) i =1 2 ∂g / ∂y = ∑ (Vty (i ) ) + F * i =1 N (13. Summing from one to N : ∂f / ∂x = ∑ (Vtx ) + F * i =1 2 N (13.23) ∂f / ∂y = ∑ (Vty (i ) ) ⋅ (Vtx (i ) ) N i =1 (13. g and h . a.28) ∂h / ∂x = ∑ (Vtx (i ) ) ⋅ (Vt (i ) ) i =1 N N (13. and d. b.c. Next we will generate a matrix of partial derivatives of f . c. ∂h / ∂x = ∂h / ∂y = ∂h / ∂z = 0 .
simply return to the beginning of the algorithm and repeat the process until convergence occurs.∂h / ∂z = ∑ − (Vt (i ) ) + F * i =1 2 N (13. The parameter j is the iteration number.31) The following matrix formulation will solve for improved values for the unknowns: VwN VwN ∂f / ∂x ∂g / ∂x ∂h / ∂x f VwE = VwE − ∂f / ∂y ∂g / ∂y ∂h / ∂y ⋅ g ∆V t j +1 ∆Vt j ∂f / ∂z ∂g / ∂z ∂h / ∂z − h −1 (13. This will usually occur after just a few steps.32) With improved values for the unknowns. We now have the north and east components of wind and the previously unknown error in true airspeed. 163 .
The maneuver is also called a pushoverpullup because that is what is done. Then a pullup is performed back through N z of 1.0 to an N z of 1.0 or more on fighter aircraft. The maneuver begins with a pushover to a g level less than 1. After attaining maximum N z then a pushover is performed back to N z = 1. there is a net altitude loss during the maneuver and a net Mach number loss. The aircraft drag model was the same as for the takeoff simulation presented in the takeoff section.50 g per second. If N z is more than 2. or 2.01 and the altitude loss is less than 1. The throttle is kept constant during the maneuver. The rate of change of N z is between 0.90. 0.1 Introduction Dynamic performance typically involves the collection of lift and drag data at near constant Mach number with maneuvers that last less than 15 seconds.25 and 0. The slower rate would produce larger Mach number variations but would also produce smaller rate effects on the data. For a lift coefficient less than 0. splits. but corrections are made to the data to nominal Mach numbers.0 during the pullup.0. Nominal Mach numbers would typically be 0.1 represents the drag polar used. The maneuver begins with a stabilized trimmed point at an aim Mach number.5 on transport aircraft. Both Mach number and rate corrections are made to the data. This maximum α is usually (but not always) something less than the limit α .6 and low Mach numbers where compressibility is not substantial.5.0 and some have used an aim angle of attack ( α ) instead of a maximum load factor as the maximum point in the roller coaster. There are three dynamic performance maneuvers: roller coaster. This was for a pseudo F16 aircraft. then the Mach number loss could be more than 0. This is accomplished by varying normal load factor ( N z ) in a short time period.70. A simulation of a roller coaster maneuver was conducted. 0. .2 Roller Coaster The roller coaster is a smooth sinusoidal variation of load factor versus time.01.85. and windup turn. Figure 14. The Mach number loss is usually no more than 0. the maneuver will take an average of 8 seconds to perform. Some fighter projects used a maximum N z of more than 2. therefore. On fighter aircraft that is usually to an N z of 0. altitude ( H C ).000 feet. etc.0 DYNAMIC PERFORMANCE 14. and N z = 1. This is because a large maximum α would produce large Mach number losses during the maneuver because the aircraft is at a high drag condition at a positive flight path angle ( γ ) and is decelerating very rapidly. 0. but both are quite small.0.80. 14. Generally.0 and on transport aircraft that is usually to an N z of 0.0.14.
14. . On the positive side.30 0.20 0.028 0.05 0.2) A sinusoidal variation of normal load factor was chosen to produce a period of 4 seconds with amplitude of 1.020 0. and pressure altitude are shown in Figures 14.6 Mach number at 30.1 Drag Model The initial condition chosen to illustrate the roller coaster is 0.35 Lift Coefficient .022 0. there is a relatively small loss in altitude (80 feet) and gain in Mach number (0.4. α is small. As shown.024 0.CD Figure 14.45 0. However.15 0. the range of CL .1) N z = cos γ + (14.10 0.40 0.000 feet pressure altitude. due to the slow N z variation. the noise in the data is usually quite low.026 0.00 0.CL 0.Lift Coefficient versus Drag Coefficient 0.032 Drag Coefficient .3 and 14. for a fighter type aircraft. The first data point was at N z = 1.0 and then thrust was set equal to the drag at that point and kept constant during the remainder of the maneuver. Mach number. standard day.25 0.0 g.030 0.2. The N x and N z formulas used are those derived in earlier sections for nonbanked flight as follows: Nx = ! ! Vt H + g0 Vt Vt ⋅ γ! g0 (14.018 0.004). The time histories of normal load factor.
5 0.0 Figure 14.Roller Coaster Simulation: Normal Load Factor versus Time 2.5 Normal Load Factor (g's) 1.900 0 1 2 Elapsed Time (sec) 3 4 Figure 14.3 Roller Coaster Altitude Time History .0 0.0 0.980 29.940 29.960 Altitude (ft) 29.0 1.2 Roller Coaster Normal Load Factor Roller Coaster Simulation: Altitude versus Time 30.0 4.000 29.0 Elapsed Time (sec) 3.920 29.0 2.0 1.
No attempt is made .02. The total maneuver.000 feet. Since the maneuver only lasted a few seconds.603 0. 14. could produce an altitude loss of up to 10. A better maneuver to perform is a pure inverted pullup. which is a portion of a splits. This is ideal to collect dynamic performance data. There is also an altitude loss during the maneuver of up to 2. This usually takes no more than 8 seconds and is often as little as 3 seconds.4 Roller Coaster Mach Number Time History 14. until a limit condition on N z or α is reached.0 to the limit g of the aircraft.0 g per second.602 0. Then.3 Windup Turn The windup turn begins at wings level trimmed at an aim Mach number and altitude.4 SplitS The splits is a fighter tactics maneuver used to change direction and altitude very rapidly. the aircraft is gradually pulled into a turn. at a rate of up to 1. The aircraft is pointed downhill during the maneuver to minimize the Mach number loss during the highg maneuver as drag gets very high and the aircraft decelerates rapidly.604 Mach Number 0. The aircraft is trimmed at an aim Mach number and altitude.600 0 1 2 3 4 Elapsed Time (sec) Figure 14.605 0.0 g per second to the limit N z or α . including the recovery. A portion of the maneuver is an inverted pullup during which N z is varied from near 1.000 feet as the aircraft ends up pointed nearly straight down at the conclusion of the maneuver. This takes approximately 3 to 8 seconds. The throttle is kept constant during the maneuver to give an accurate thrust computation. The aircraft is rolled inverted (180 degrees roll angle) and an inverted pullup is performed at a rate of up to 1. The throttle is kept constant during the maneuver because most inflight thrust computer programs are ineffective at computing thrust accurately during throttle transients.601 0. even large deceleration rates would not vary the Mach number more than about 0.Roller Coaster Simulation: Mach Number versus Time 0. The aircraft is trading altitude for airspeed.
Then.to minimize the Mach number variation. This maneuver is better than the windup turn for data processing with an INS since there are only small bank angle ( φ ) variations from 180 degrees and terms in the INS equations involving φ are negligible.000 feet. The initial conditions chosen were 30. which is less than 8 seconds. We also did not have any significant roll rate effects. but the total maneuver including recovery could produce an altitude loss of up to 10. sin φ = 0 . from that used for the roller coaster. The simulation was ceased at a lift coefficient of 1.6 2 (14.5 plots the drag model used. however.6 ) ∆CD = 0 if CL < 0. We attempt to collect data from pitch attitudes ( θ ) of 0 to about 70 degrees to avoid getting data during the INS transition through 90 degrees of θ at which the heading (ψ ) changes by 180 degrees. cos φ = −1 . an altitude loss of up to 2.3) The N x formula is identical to the one used for the roller coaster. For the pure inverted case ( φ = 180 degrees): a. a simulation is shown.60.4) Figure 14.0 g per second. As with the windup turn.5 ⋅ (CL − 0. . This would often dictate the g onset rate since it is desired to achieve maximum g or α before the aircraft reaches about a negative 70 degrees pitch angle. V ⋅ γ! N z = − cos γ + t g0 (14. To illustrate the splits.000 feet during the data acquisition portion of the maneuver is typical.85. The transformation for N z involves sin φ and cos φ terms. The simulation was performed at a rate of 1. and b. The drag model was modified. but the Mach number usually decreases no more than 0. with the addition of a separation drag term as follows: ∆CD = 0. the N z formula is the negative of the roller coaster formula. This can be seen from the axis transformations in the excess thrust section.02 during the data portion of the maneuver.000 feet and a Mach number of 0.
0 0.CL 1.2 0.8 0.6 through 14.5 SplitS Drag Model The timehistory parameters of normal load factor.8 Figure 14.2 Lift Coefficient .5 0. Mach number.0 0.2 0. and pressure altitude follow in Figures 14.4 1.6 0.4 Drag Coefficient .7 0.1 0.Splits and Pullup Drag Model: CL versus CD 1.3 0.6 SplitS Normal Load Factor .6 0.0 0.4 0. SplitS Simulation: Nz versus Time 7 6 Normal Load Factor (Nz) 5 4 3 2 1 0 1 Elapsed Time (sec) 2 3 Figure 14.8.6 1.CD 0.
85 0.81 0. In addition.Splits Simulation: Mach Number versus Time 0.86 0. the .600 29. That is.0 g per second.000 29.400 0 1 Elapsed Time (sec) 2 3 Figure 14. A simulation for the pullup was conducted using the same drag model and initial conditions as for the splits. a pullup maneuver has been used in lieu of the splits to obtain high.82 0. They have found that the pullup maneuver has one big advantage over the splits.83 0.α data.800 Pressure Altitude (ft) 29.7 SplitS Mach Number Time History SplitS Simulation: Pressure Altitude versus Time 30.84 Mach Number 0.8 SplitS Altitude Time History 14.5 Pullup On the F15 projects. The pullup simulation was conducted at the same g onset rate of 1.80 0 1 Elapsed Time (sec) 2 3 Figure 14. there is no need to recover back to the original altitude.
200 Altitude (ft) 30.300 30.85 0.000 0 1 Elapsed Time (sec) 2 3 Figure 14.82 0.79 0.84 0.78 0 1 Elapsed Time (sec) 2 3 Figure 14.81 0. Pullup Simulation: Mach Vs Time 0.100 30.10.60 was the same.83 Mach Number 0.end condition of CL = 1.10 Pullup Altitude Time History Table 14.9 Pullup Mach Number Time History Pullup Simulation: Altitude Vs Time 30. . The Mach number and pressure altitude time histories are in Figures 14.80 0.9 and 14.1 compares the initial conditions and end conditions of the pullup and the splits.
. then it could be that there is a vertical wind. the throttle could be varied to maintain speed.5) (14.Table 14.219 29. Since you are not trying to get drag data.0 +226.450 6.1 Initial Pullup SplitS 1. especially a highg (high bank angle) turn. the full transformation equations are used.785 0.1 PULLUP AND SPLITS INITIAL AND END CONDITIONS Nz M 0. M ) In addition.850 0. one can use an INS method to calibrate angle of attack during turns.452 0.0 428. In practice. The vertical wind would not affect the turn data as much.6) The roller coaster maneuver. the one shortfall of the INS method is that vertical wind is assumed zero.6 Angle of Attack During the roller coaster. a.7 45. could be used to calibrate production angleofattack probes or vanes.2 As can be seen.936 500. The turn. particularly.800 Vt (kts) HC (ft) ! H (ft/sec) ! Vt (kts/sec) 0. keeping the bank angle ( φ ) near 90 degrees. You can detect vertical wind by comparing data on the lift curve.9 462. 14.0 58.000 30. However.8 30. the pullup does not end up with a very large vertical velocity. and splits maneuvers the computation of angle of attack from the INS is quite simple for bank angles near 0 or 180 degrees. The above equations are simplified for illustration purposes only. since the vertical wind is nearly perpendicular to the axis of the angle of attack. Only for very high angle of attack would you want to use the splits for calibration of production systems. As discussed in the flight path acceleration section. α = θ − γ (φ = 0) roller coaster and pullup α = −θ + γ (φ = 180) splits (14.3 472. One way to minimize the effect of vertical wind is to do a varying g maneuver during a stabilized highg turn. the splits has the advantage of not losing as much Mach number.7 Vertical Wind If there is an unexplained bias in your data. will be less sensitive to vertical wind since the vertical component of velocities in the angleofattack formula is proportional to the cosine of the bank angle. 14. The full equations involved bending and rate corrections and allowance for being off exactly φ = 0 or 180 degrees. pullup.00 6. α = f (CL .
1. the normal load factor ( N z ) would not be precisely 1.3048 (divide meters by 0. and b. The internationally agreed upon exact conversion factor between meters (or metres in Great Britain) is 0. As of the writing of this text. Under these conditions. and heading.005278892 ⋅ sin 2 ϕ + 0. .1) Where the symbology used by the International Union is as follows: a. some typical values of sea level gravity are shown in Table 15. We will consider them individually. we will take the liberty of changing the International Union’s sea level gravity symbology from γ 0 to g 0 .03185 ⋅ 1 + 0. First. the error is quite small (<1 percent). γ 0 = sea level gravity (cm/sec2). Given that and using equation 15.0 SPECIAL PERFORMANCE TOPICS 15. As will be seen.1. but not zero. ϕ = latitude (degrees).1 Effect of Gravity on Performance Below is the international gravity formula as adopted by the International Union of Geodesy and Geophysics as presented in Britannica Online. In this text.00. There are four variables: latitude.3048 to yield feet) and the number of centimeters (cm) in a meter is 100. we have used a rather simplified gravity model of g = constant = 32.000023462 ⋅ sin 4 ϕ cm/sec 2 (15. Consider only a 1g flight where the aircraft is unbanked and has zero vertical velocity and zero rate of change of vertical velocity. speed. that simplification is widely used in the conventional aircraft flight testing community. γ 0 = 978. This topic will address the magnitude of error that this simplification produces.15. altitude.174 ft/sec2.
Edwards is at 2.925.1854 0.1432 0.26 Northern Greenland 80. Point Barrow.643 feet) is also from the International Union of Geodesy and Geophysics and is a value for the equator. Alaska 62.00 9. Minnesota.00 9. The gravity varies with altitude. This compares to 20.00 9.1170 0. The above local g values are computed for sea level.20 Arctic Circle 66. represent the extreme latitudes of the continental United States.7973 32. The earth’s radius (20.92 9.7864 32. and Florida Keys.50 9.8192 32.2578 0.Table 15.27 Note: The local gravity at Edwards of 32.2202 0. Petersburg.S. Florida 24. The places in Table 15.8062 32. 49.05 Lake of the Woods.8306 32.7893 32.8155 32.30 9.8239 32. England 51.2306 0. Russia 60. AFFTC 34.2031 0.21 Costa Rica 10.7803 32.25 Equador (Equator) 0.1075 0. Standard Atmosphere. Table 15.33 9.7819 32. Alaska.1725 0.18 Anchorage.8207 32.0928 0. Alaska 71. Florida.7957 32.10 Baghdad 33.8118 32.553 feet from the 1976 U. Minn. Barrow.00 9.2397 0.8101 32.00 9. Lake of the Woods.8267 32. Using latitude of 35 degrees.00 9.11 Florida Keys.14 St.00 9. For instance.09 London.50 9.2526 0.136 ft/sec2.136 ft/sec2 has been measured and agrees with the model.00 Bldg.1 EFFECT OF LATITUDE ON GRAVITY AT SEA LEVEL Variation g from the g 32.00 9.04 45 deg latitude 45.300 feet geometric altitude and the gravity at that altitude is 32.24 Pt.8322 32.1911 0.80665 (ft/sec2) (pct) Place (deg) (m/sec2) Reference North Pole 90.2 illustrates this effect using the inverse square gravity law.13 Copenhagen 55.1 were chosen to represent either even latitudes or interesting places.00 9.1380 0.17405 Standard Latitude 9. is the highest latitude in the lower 48 states.2151 0. .00 9.0877 0. 2750.855.58 9.18 Mexico City 20.
000 30.8 Mach number. Using 40.113 32.000 feet than at 2. for conventional aircraft performance.9943 0.07 0. Russia.38 0.12 0.48 0.000 100.838 2 Percent from Surface 0.9972 0.8 2.9952 0.0 0.57 0. Table 15.000 20.8 2.95 1.000 g (ft/sec ) 32. Speed has an effect upon normal load factor due to coriolis terms in the gravity equations that are functions of the true heading. we have been ignoring these factors.0 0.000 60. 0.051 32.3 illustrates the speed and heading effect.36 0.10 0.021 31.000 90.000 70.0 0.960 31. Generally. 40.00 0.04 The last two variables are speed and heading which need to be considered together.9952 0.26 0.300 10.0 Normal Load Factor (g) 0.59 percent more lift is required heading west than heading east.9824 0.67 0. For a given mass of aircraft.45 0.55 0.02 0.0 0.2 EFFECT OF ALTITUDE ON GRAVITY Altitude (ft) 0 2.9896 0.36 percent less lift at 40.9952 0.76 0. . what is the significance of this? The normal load factor experienced by an aircraft varies with latitude over the earth.8 2.300 feet over Edwards AFB.Table 15.93 Percent from Standard 0.990 31.9952 0. We needed 0. than over Edwards AFB. At 0.8 2.9896 0.3 EFFECT OF HEADING AND SPEED ON NORMAL LOAD FACTOR Heading (deg) 0 0 0 90 90 90 180 180 180 270 270 270 Mach Number 0.000 feet.23 percent more lift over St. Table 15.082 32.869 31.000 80.29 0.0 0.000 feet and latitude of 35 degrees.83 0. how high and how fast the aircraft is flying and in what direction.000 40.899 31.143 32.64 0.0 0.19 0.9914 0.74 0.136 32.9968 So. Petersburg.17 0.9943 0.000 50.929 31.86 0. we needed to generate 0.0 0.
the data collected heading west would have shown about ½ of 1 percent more drag and fuel flow than the data collected heading east. What we are talking about is roughly up to a ½ of 1percent factor we had been ignoring. the average increase in fuel flow for the receiver aircraft was 25 percent. the variations in drag are less than 0. The tanker downwash creates a negative vertical wind that the receiver aircraft encounters.02641 0. Although N z varied more at M=2. Using a B52G drag polar at 0.02634 0. it was necessary to generate more lift and therefore. The significant comparison is between west and east being nearly ½ of 1 percent apart.8.72 Very similar percentage differences were obtained using an F15 drag polar.0 than at M=0. With an INS as a data source. 15. This is due to the much smaller amounts of drag due to lift at the higher speeds. In cruise performance. Relative to the wind axis. more drag due to lift was created. This does not produced a bias in our data (unless all our cruise data is collected heading east) but is rather a source of the scatter.00 0.4 was generated. Nevertheless.1 percentage.0000 0. The bias between the reference and the other data tended to fall out in flight test data as the drag polars generated are biased to compensate for this effect and there is not a ½ percent error in range data.8 Mach number. we can account for the variation in gravity.0 for the F15 aircraft.4 EFFECT OF HEADING ON DRAG COEFFICIENT Heading Reference 270 (west) 0 or 180 90 (east) Nz 1. The drag of the receiver aircraft is unchanged. a 1percent increase in drag is about a 1percent increase in fuel flow required to sustain stabilized flight. corresponding to an optimum cruise at 40.How did these variations in N z translate to performance? As N z increased.02622 Percent from Reference 0.000 feet. the receiver is climbing at a flight path angle exactly equal to the tanker downwash angle to maintain a constant altitude.02628 0. Table 15.26 0. The B1B behind a KC135 aircraft showed a 15percent increase.9914 CD 0. The thrust required of the receiver is increased due to the receiver climbing in the tanker downwash. Table 15. the effect on performance was actually much less.9972 0. The YC141B increase in fuel flow behind a KC135 was 20 percent. .2 Performance Degradation during Aerial Refueling A common misconception is that the drag of the receiver aircraft during aerial refueling is increased. During tests of the KC10 aircraft with 10 different types of receiver aircraft. the receiver aircraft requires additional thrust and a resultant increase in fuel flow. To sustain this climb. if the data were accurate enough to detect that small difference.49 0.9943 0. At Mach number 2.
While in the terrain following mode. the fuel flow increases dramatically. and ARt = aspect ratio of the tanker aircraft. the magnitude of the drag increase in the pullup is greater than the magnitude of the drag decrease in the pushover. you only need to know the theoretical downwash angle behind the tanker and then apply a downwash factor.2) CLt = lift coefficient of the tanker aircraft. the average increase in fuel flow can be substantial.To estimate the increase in thrust required for a receiver aircraft. . The net effect is there is a net increase in average thrust required and a resultant increase in average fuel flow.5. The thrust specific fuel consumption (tsfc ) will typically be less than 1. The downwash factor (K) is simply a multiplicative factor to account for the fact that the receiver aircraft is in a flow field that is a combination of the tanker flow field and the free stream. εt = CLt (π ⋅ ARt ) (15. The theoretical downwash angle ( ε 0 ) is exactly twice the ideal angle of attack. ∆Fn = Wt ⋅ sin(ε ) 15. In a pullup ( N z >1) the drag is increased over that for an N z =1 due to an increase in drag due to lift (or induced drag). For both the KE3A and the B1B aircraft.4) Flight while performing terrain following results in an increase in average fuel flow when compared to flight at the same average Mach number and altitude level. ( N z <1) the drag is reduced due to a decrease in the drag due to lift. Because of the parabolic nature of the drag polar.3) Then the increase in thrust of the receiver could be computed by the component of weight through the downwash angle. this K factor is about 0.5) to be approximately equal to the ideal angle of attack of the tanker. the receiver aircraft is climbing while behind a tanker in level flight.0 in afterburner.3 Performance Degradation during Terrain Following (15. the aircraft is constantly either pulling up or pushing over. With respect to the wind axis. ε0 = where: ( 2 ⋅ CLt ) (π ⋅ ARt ) (15. The actual downwash angle is found (with a K of 0. For the case of an aircraft with automatic terrain following and afterburner.0 in nonafterburner and >2. Every time afterburner is used. In a pushover.
VC = 280 knots (calibrated airspeed). and b.5. Some typical parameter uncertainties are shown in Table 15. True airspeed is: e. “how accurately do we measure certain performance flight test parameters.5 knots ±0. an altitude error of 25 feet produces a 0. .11percent error.5 knot produces a 0.8213. and a calibrated airspeed error of 0. Calculating the Mach number: d.000 feet (pressure altitude). H C = 35.4 Uncertainty in Performance Measurements There is no precise answer to the question. In some cases. these parameters are not direct instrumentation measurements.01 g ±0. In addition.” as each instrumentation system is different.5% ±0. On a standard day the ambient temperature is: c. Let us choose a typical high altitude cruise condition: a. a performance figure of merit might be the specific range at optimum speed and altitude.26percent error.04percent error.81 °K.5 PARAMETER UNCERTAINTIES Parameter Fuel Flow Calibrated Airspeed Gross Weight Longitudinal Load Factor Normal Load Factor Ambient Temperature Pressure Altitude Units lbs/hr kts lbs g g °K ft Symbol Uncertainty ±1% ±0. Vt = 473.44 knots.5 °K ±25 feet Wf VC Wt Nx Nz T HC 15. If the computed ambient temperature is in error on the high side by 0.001 g ±0.15. but rather the result of computations involving several measurements.98 knots for a 0. T = 218. our experience has given us some approximate uncertainties that we feel are obtainable and had been achieved.5 degree K then the true airspeed would be Vt = 473. Table 15. M = 0.5 Sample Uncertainty Analysis For a transport category aircraft. Nevertheless.
00 percent. since it would be blowing on some side of the structure – never away from the structure. Vt 0. The RMS of the four uncertainties computes to be 1.50 percent. We shall assume error in drag produces a 1. In about 100 BC an octagonal (eightsided) marble structure. 0.53 percent. 15. for range factor ( RF ). Please note that carrying out the speeds to five significant figures did not imply that we could measure speeds to that level of accuracy. with the advent of EGI even greater accuracies than those presented above may be achieved for airspeeds. d. . and flight path accelerations.11 percent due to T error. was constructed. 42 feet high and 26 feet in diameter. This direction is the direction from which the wind is blowing. Vt 0. etc). we have the following errors: a. Wf The root mean square (rms) of the three Vt uncertainties computes to be 0. and 1. the same as the compass heading of the Tower. c. Improper in the sense that defining the wind direction as from which it is blowing is opposite from the vector direction of wind.00 percent. Wt f. a structure called the Tower of the Winds is discussed briefly.6 Wind Direction Definition What may seem to be an improper definition of wind direction (from which the wind is blowing) may derive from ancient Greece. One could think of this Tower as either an aircraft control tower or an aircraft. altitudes.04 percent due to H C error. The wind would always be positive. In Britannica Online. this positive wind would have a direction of north (0 degrees).26 percent due to VC error. Vt 0. e. NE. Then. The eight sides face points of the compass (N. so to speak.285 percent. if the wind were blowing directly on the north side of the Tower of the Winds.001 g in longitudinal load factor yields a 1.0percent error in range factor. Therefore. E. an error of 0.At an L / D = 10.0percent error in drag.0 . It would seem logical that a wind blowing on the structure would be considered a positive wind. N x 1. At the time of this handbook. b.
1 Introduction For presentation and comparison purposes. fuel flow ( W f ). about a 10percent decrease in thrust at MIL or MAX. and drag ( D ) are as follows: ′ ′ Fns = Fnt + ( Fns − Fnt ) where: (16. and ′ W ft = test day predicted fuel flow (pounds/hour). cg (center of gravity).1) Fns Fnt ′ Fns ′ Fnt = standardized net thrust (pounds). The standardization is performed using an additive increment method. we incurred only a 1. The standard conditions are specified values of gross weight. but these could be large when temperature is substantially off standard day. At the AFFTC in midsummer. ′ ′ W fs = W ft + (W fs − W ft ) where: (16. on average. usually 18. the temperature at 30.16. small corrections to standard day conditions are made. = standard day predicted net thrust (pounds).400 Btu/pound. Standardization relies upon a predicted model of drag. 16. performance data are usually corrected to standard conditions. ′ W fs = standard day predicted fuel flow (pounds/hour).S. and fuel flow. and = test day predicted net thrust (pounds). thrust.2 Increment Method The general principle of standardization is an additive increment method.3) 180 .000 feet is.2) W fs = standardized fuel flow (pounds/hour). Standard ambient temperature is usually based on the 1976 U. typically. = test day net thrust (pounds). and Mach number. W ft = test day fuel flow (pounds/hour). 10 degrees C hotter than standard day.percent error in the standardized results. Standard Atmosphere. which produces. Usually. If there is a 10percent error in the predicted model and we made 10percent corrections to the data. pressure altitude. Fuel flow is first standardized to a minimum fuel lower heating value (LHV). LHVtest W ft = W ft ⋅ 18. 400 (16.0 STANDARDIZATION 16. The formulas used to standardize net thrust ( Fn ).
which amounts to a ½percent correction. Dt = test day drag (pounds).550 Btu/pound.2. The predicted thrust and fuel flows are determined from a prediction (or status) deck.2 Acceleration/Deceleration Excess thrust and fuel flow are standardized: a. almost all actual fuel will have an LHV greater than the spec. formulas presented in the lift and drag section could be used. N z is computed. The correction will generally increase fuel flow.5) (16.1 Climb/Descent Excess thrust and fuel flow are standardized: a. 16. Ds′ = predicted standard day drag (pounds).2. The standardization parameters for various maneuvers are discussed in the following text. The test net thrust is determined. These are described briefly in the thrust section. and Dt′ = predicted test day drag (pounds). usually.Typical test values of LHV are in the vicinity of 18.4) Ds = standardized drag (pounds). 16. In lieu of that. That is. Dt = Fnt − Fex t Fext = N x ⋅ Wt = test day measured excess thrust Then.7) The above equations illustrate the general principle. since the spec is a minimum. (16. The contractor drag model should include an accounting for skin friction drag.6) Fexs = Fext + ( Fns′ − Ds′ ) − ( Fnt′ − Dt′ ) (16. Ds = Dt + ( Ds′ − Dt′ ) where: (16. The predicted drags are obtained from a contractorprovided predicted drag model subroutine. 181 . Each maneuver involves a different parameter being adjusted to standard conditions but the basic method is the same incremental difference method. N z = 1. from an inflight thrust deck.0.
16. fuel flow. the above equation would suffice.8) (16.16. Excess thrust = 0.5 ThrustLimited Turn N z and fuel flow are standardized: a. standard day excess thrust ( Fexs ) would be: (16.4 Cruise Fuel flow is standardized: a.0. where standard excess thrust should be zero.2.0.9) (16.0g would be for standardizing data in an endurance turn.3 Accelerating/Decelerating Turn Excess thrust and fuel flow are standardized: a.2. The question that needs to be answered is “what is the difference in the magnitude of difference between the ratio and difference methods?” Take the case of the standardized 182 . 16.0 (usually) (Note: a rare exception to the 1. N z = 1.2.). an iteration is required. and b. The formulas for standard day net thrust. 16. N z is specified.11) For fixed throttle maneuvers (climb. turn. For cruise. Excess thrust = 0.3 Ratio Method An alternative to the increment method of standardization is a method based upon ratios. and drag would be as follows: F ′ Fns = Fnt ⋅ ns Fnt′ W ′ W fs = W ft ⋅ fs W ′ ft D ′ Ds = Dt ⋅ s Dt′ Then. and accel).10) Fexs = Fns − Ds (16.
a. Conversely. let us write out the full Fexs formula for both increment and ratio methods. Your cruise standardization algorithm should check to assure that standard day drag is less than the maximum available thrust.excess thrust in acceleration. 183 . then the ratio method will introduce the least error. but would be on the standard day. For cruise at high altitude. The test day temperature may have been substantially colder than standard day giving the engine much more thrust than would be available on the warmer standard day. Either way.12) Dt = Fnt − Fext for both methods Then. b.15) Then. For instance. If the prediction models are in error by approximately a constant percentage. the ratio method becomes: (16. The engine may be in some manner limited (turbine temperature or rpm limit) on the test day. Standardization is performed as a means of convenient data presentation. This is due to the errors canceling out when doing the subtraction. the engine may not be on this limit on the standard day.13) D ′ F ′ D ′ Fexs = Fext ⋅ s + Fnt ⋅ ns − Fnt ⋅ s ratio method Dt′ Fnt′ Dt′ Fexs = Fext + ( Fns′ − Fnt′ ) − ( Ds′ − Dt′ ) increment method (16. (16. That is due to having sufficient thrust on a test day. It represents an extrapolation of an actual test point. The following are two sources of error in standardization. One should recognize that a data point on a plot presented as standard conditions is a data point that was not flown. If this limiter is not accurately modeled in the status deck. then the correction to standard day will have errors. whichever method introduces the most error into the standardized excess thrust is a function of the errors in the prediction models. If there was zero error in both test day measured net thrust and in the thrust model. then the increment method will introduce the least error. the standard day conditions may be unachievable. but not on a standard day. then there would be zero error in the standardization for both ratio and increment methods. F ′ D ′ Fexs = Fnt ⋅ ns − Dt ⋅ s ratio method Fnt′ Dt′ However. yielding additional thrust. one is invariably introducing some errors (hopefully small) into your data by the very process of standardization. From the above equations. This is because the errors would cancel out when doing the division. it may not be on the limit on the test day. if the models are in error by approximately a constant magnitude. Conversely.14) (16.
One should not assume that their drag. We could think of drag as having only two components: minimum drag and drag due to lift. For the purposes of our model. thrust.17. 17. Drag has three components.80. Figure 17. From the subsonic condition to Mach number = 1. and drag due to lift.1 Minimum Drag Coefficient In order to illustrate the shape of performance parameters. or fuel flow models would be the same as.1 Introduction In this section. Drag due to lift is also called induced drag. we will construct a performance model. That is a typical minimum drag coefficient for a wide range of aircraft. That drag model is fiction. Our basic formula for drag coefficient is the AFFTC drag model formulation from the previous section. Some data points were assumed and a curve fit was applied. but approximates that of an F16 aircraft.0200 (200 drag counts) for Mach number < 0.0200 + ∆CD (17. or as simple as. we will construct a drag model. Then. we will make up numbers for standard day at 30.000 feet pressure altitude.1 is delta drag coefficient for the subsonic condition.2 Drag Model 17.0. The model will be highly idealized. We will start by assuming that CD min = 0. These are skin friction. profile drag.2.1) 184 . Profile drag is sometimes called form drag. such as specific excess power as a function of Mach number or altitude. those presented here.000 feet. Minimum drag is then the sum of profile drag and skin friction drag. the drag coefficient approximately doubles. The equation for minimum drag coefficient at any given Mach number is as follows: CD min = 0. our predicted skin friction drag formulas will be used to compute minimum drag at conditions other than standard day at 30.0 A SAMPLE PERFORMANCE MODEL 17. The purpose of this section is to illustrate some general concepts.
70 0.007 0.90 Mach Number 1.6633 0.00 1.0000 0.1998x 2 + 5.2 and 17.009 0.10 1. of course.9003x 0.1.0050 0.38.019 0.9433x2 + 139.0100 0. Each aircraft will have different characteristics.drag rise 0.7.Subsonic y = 2.90 0.25 Figure 17.5066x4 + 113.0250 3 .15 1.0150 delta Cd .10 Figure 17.1 Subsonic Drag Increment The drag coefficient in the transonic regime will peak out somewhere just past Mach number = 1.0 and then will sometimes decrease slightly with increasing Mach number.4193x3 .3 are for transonic and supersonic speeds.0200 0.0050 0.013 0.20 1.7543x .00 1. Figures 17.7038 0.95 1.021 0.9828x .005 0.2 Transonic Drag Increment 185 .011 0.80 0. Data values for minimum drag were assumed at various Mach numbers and curve fits were applied.Transonic y = 25. Delta Cdmin .188.017 delta Cd 0.delta Cd versus Mach Number .05 Mach Number 1.023 0.015 0.
the corresponding curve fits values.4193⋅X3 188.0.9828⋅X 1.018 0. Y = 25.1 contains the data points.08343 Table 17.011534x + 0.1 1.1998⋅X2 + 5. Supersonic 3.01153⋅X3 + 0.5 1.0185 0. Y = 0. and the errors in the curve fits. Y = 2.6633 b.083435 0. Summarizing the following curve fit formulas (where X = Mach number and Y = delta CD ): a.3 1. Subsonic 1.95 and 1.019 0.9003⋅X3 .delta Cd . Transonic 2.9433⋅X2 + 139.02 0.4 1.Supersonic y = 0.5066⋅X4 + 113.7038 c.8 1.7 1.0175 1 1.9 2 Mach Number Figure 17. For instance.7.3 Supersonic Drag Increment Notice that there were overlapping data points in each of the plots.109113x + 0.0205 delta Cd 0.2 1. 186 .06127⋅X2 0. 0.0215 0.0195 0.022 3 2 0.10911⋅X +0.0 Mach number appeared in both the subsonic and transonic plots.061267x .6 1.7543⋅X 38.021 0.
00031 1.0213 0.0216 0. Bold numbers are at Mach numbers where the curve fits equate. ∆CD = subsonic for 0. d. and e.1500 0.9995 .00005 2. The error numbers are carried to one extra digit.4 puts all three pieces of the minimum drag model together on a single plot.0208 0.0750 0. b.9995 0. ∆CD = sup ersonic for 1.0 .00030 0.0090 0.00011 1.00019 0.0037 0.00000 Notes: 1. ∆CD = 0 for M < 0.7993.6000 0.1 TABULATED DRAG RISE DATA Mach Number ∆CD Data ∆CD Fit Error = Data – Fit 0.1467 0.0040 0.7993 0. If this were actual flight test data. ∆CD = 0.00019 0. and 3 on page 187) with transition points at the following Mach numbers: a.0 ) is to keep the polynomial from giving very unreasonable results in event the model is used beyond the last Mach number.0215 0. The lowend constraint ( M < 0. 187 .00000 0.0184 0.0000 0.00002 0.0500 0. The first and last conditions are constraints applied to the model.0200 0.0214 0.00002 0.0023 0.0180 0.1467 .0190 0.0214 1. wind tunnel data could perhaps be utilized to extrapolate beyond where flight test data were obtained.8000 0.2000 0. The Mach number ranges for the above are not meant to imply any general definition of the terms subsonic.0000 0.00028 0.0191 0. However.0020 0.00010 1.1467 < M ≤ 2.000 .9000 0. ∆CD = transonic for 0.02144 0.0199 0.0185 0.7993 . we could not be certain what the behavior of the minimum drag might be beyond where actual test data were acquired.7993 < M < 0.0218 0.0180 for M > 2.0092 0.8750 0.0216 0. Figure 17. 2. transonic.4000 0.02148 1.9500 0.0200 for all Mach numbers less than 0.0180 0.00021 1.7993 ) is to keep the minimum drag at 0.1000 0.9995 ≤ M ≤ 1.Table 17. 2. The highend constraint ( M > 2. c.0210 0.0216 0. or supersonic.00004 1.01984 1.00021 1. The model for minimum drag is then the three equations (1.0000 0. They are simply where the curve fits for this particular arbitrary data set intersected.
Figure 17.6 0.2 1.4 Summary of Delta Drag Coefficient 17.000 ft 50.2 1. and presume a characteristic length of 10 feet.0025 0.7 1.4 0.0 1.8 1.5 is for standard day conditions.5 Skin Friction Drag Coefficient 188 . We will use the empirical skin friction flat plate turbulent boundary layer equations presented in the lift and drag section.1 1.000 0.000 ft 40. Skin Friction Drag Coefficient versus Mach Number 0.8 0.4 1.9 2.8 1 1.6 1.015 0.2 0.4 1.0010 0.delta Cd versus Mach Number 0.000 ft 30.005 0.025 0.000 ft 0.8 2 Mach Number Figure 17.9 1.3 1.0030 Sea Level 10.3 Skin Friction Drag Coefficient Skin friction drag coefficient varies with Reynolds number and Mach number.0 Mach Number Figure 17.020 delta Cd 0.0040 0.010 Data Points Subsonic Fit Transonic Fit Supersonic Fit 0.5 1.000 ft 20.0035 Skin Friction Drag Coefficient 0.0020 0.0015 0.6 1.
This is positive with increasing altitude. the total variation in drag coefficient due to skin friction (at the same lift coefficient) can be calculated as follows: ∆CD = S wet ∆C f ⋅ ⋅ ∆h = 4 ⋅ 0. one will be able to develop a much more detailed and accurate model.wing span.000 feet and 0. for instance. at 0. 189 . S = 300 ft2 . f. the slope of the C f curve is 0. c.000 feet.0028 (28 drag counts) S ∆h (17. 17.8 Mach number.MAC (characteristic length).000 pounds . that is.fuel capacity. The temperature slope is positive with increasing temperature.6 presents idealized drag due to lift slope data points and a secondorder polynomial curve fit of those points.200 ft2. Quite significant! For our fictional aircraft (modeled after an F16 aircraft). at the same condition.000 feet. With actual flight test data.4 Drag Due to Lift A drag due to lift (induced drag) model will be derived based upon the formulas presented in the lift and drag section of this handbook. Figure 17.000014 per 1. Fuel = 6. and g.2) That is a 28drag count number over the range of sea level to 50. WZf = 18.000014 ⋅ 50 = 0. e.zero fuel weight.8 Mach number.000 pounds . the C f is higher on a day that is hotter than standard. we will presume the following dimensional data: a. that is. we have mostly ignored the variation in the transonic Mach number range.0 ⋅ S = 1. As you can see. d. the higher altitude has the higher skin friction drag. b. These numbers will be used to illustrate performance parameters in other sections of this handbook.wing area. S wet = 4. b = 35 feet .083. l = 10 feet . on a standard day. AR = b 2 / S = 4.At 30. Compare that to the typical number of 200 for the minimum drag coefficient.0000018 per 1 degree K. This model (as with the minimum drag and skin friction drag) is developed only as a rough approximation of an actual airplane. Again. Therefore.000 feet. Those ∆C f might appear small until one considers that the typical ratio of wetted area to wing area is about 4 and the altitude range of a fighter aircraft is 50. the slope of the C f curve versus temperature is 0.
1 3 0 .Theo retic al D rag Due to Lift Slo pe y = 0. a general formula for drag coefficient that seems to match most flight test data quite well for a given Mach number.09 90 0 .0 7 0 .0 9 0 .1 1 0 . We will call this the nonlinear portion of the model. In our model. we will assume the following for a CL min .6 0 .4) 190 .8 1 .3) where: K 2 = 0 if CL < CLb .2 1 0 . In most textbooks.1 5 0 . In general.0 5 0 .1 7 2 C d l/Cl^2 0 .02 94x + 0.1 9 0 .05 ⋅ M (17. The curve is a parabolic fit of the data and ignores the variations in the transonic speed range.100 − 0.2 3 0 .4 0. The CL min (lift coefficient at minimum drag coefficient) is usually some small positive value due to positive camber on most wings and positive wing incidence.0 M a ch Nu m b e r 1 .6 1 . and longitudinal center of gravity position condition is as follows: CD = CD min + K1 ⋅ (CL − CL min ) + K 2 ⋅ (CL − CLb ) 2 2 (17.0 Figure 17.0 0 .4 1 . CL min = 0. As shown in the lift and drag section.2 1 .8 2 . The y parameter in the theoretical drag due to lift plot is equal to K1 .0 182x + 0. pressure altitude.2 0 .6 Drag Due to Lift Slope The above drag due to lift model is for the linear (or pure parabola) portion of the drag polar.2 5 0 . there will be a deviation from the linear model as flow separation develops. the CL min is ignored.
b. and 0. 2 (1.40 0.45 Drag coefficient .0. and K2 = 2 (CL − CLb ) a.4 0.5 − 0. consider that the drag coefficient will double over that predicted by the linear model by the time a CL of 1.5 Lift coefficient .5) At CL = 1.5 0. for this model the CL min is 0.8 Mach Number 191 .0 Linear Model NonLinear Model 0. For the break lift coefficient CLb .3 1.55 0.25 0.7 0. From our models at M = 0. 0. To get a rough number for K 2 .2140.50 0.3 0.60.80.10 at M = 0.2 1.30 0. in general.7 is for this model at M = 0.1 0.099 ⋅ (CL − 0.00 0.0 1. Drag Coefficient versus Lift Coefficient (Mach Number = 0. but for simplicity. We need to emphasize that this model is pure fiction.20 0. Solving for K 2 from equation 17.5: CD − (CD min + K1⋅ (CL − CL min )2 .00. Both K 2 and CLb will.10) 2 (17.7 Drag Model at 0.6 ) Figure 17.2642 . K 2 = [2 ⋅ 0.9 1.80) 0.Hence. CD = 0.8 0.00.2 0.60. but the trends do roughly approximate that of a real aircraft such as the F16.Cl Figure 17. we will give them constant values.6 0.35 0.10 0.05 at M = 1.4 1. be functions of Mach number.Cd 0.2140 − 0. CD = 0.05 0.15 0.1 1.0 and CL <0.00 at M = 2.0200 + 0.2140] = 0.50.50 is attained. we will assume a constant value of 0.
062 0.Figure 17.2 0.90 0.038 0.60 Mach=0.03 0.09 0.CL 0.Cd 0.80 Mach=0.10 Mach=1.20 0.Cl 0.10 Drag coefficient .0 Mach=0. Figure 17.13 0.050 0.06 0. To illustrate the general shape of the polar for CL < CLb .00 Mach=0.07 0.12 0.3 Lift coefficient .1 0.05 0.01 0.5 0.80 Mach=0.054 Drag coefficient .042 0.9 Drag Model – All Mach Numbers 192 .026 0.04 0.4 0.02 0.018 0.8 represents only the subsonic Mach numbers.10 0.CD 0.046 0.30 Lift coefficient . we will plot drag coefficient versus lift coefficient as a function of Mach number.066 0.6 Figure 17. Note to those who are accustomed to seeing drag coefficient on the xaxis: the plot axes are opposite of the usual convention.40 0.50 0.8 Subsonic Drag Model Drag Coefficient Versus Lift Coefficient {f(Mach)} 0.08 0.034 0.9 includes all Mach numbers. The vast majority of 1g flight occurs at lift coefficients below the point where significant flow separation begins.022 0.60 Figure 17.90 Mach=1.7 illustrates how dramatically the drag polar can deviate from the pure parabola. and Figure 17.60 0.030 0.40 Mach=1.60 Mach=0. Drag Coefficient versus Lift Coefficient {f(Mach Number)} 0.058 0.11 0.
7) θt 2 = Tt 2 288. We will begin with a fuel flow model for nonafterburner. The parameter referred net thrust is net thrust divided by total pressure ratio at the inlet. However. the total temperature would be measured in the engine inlet.6 Thrust Specific Fuel Consumption Thrust specific fuel consumption ( tsfc ) is simply the ratio of fuel flow to net thrust. we will use a Pitotstatic derived total pressure ratio. we will construct a set of equations to represent net thrust and fuel flow. Ram air temperature is total temperature. that parameter is difficult to measure and even more difficult to model so one usually (but not always) will use a ram air temperature measurement. We will use this to compute performance during cruise. and turn. This will be used in combination with a thrustfuel flow model to compute performance parameters. That means we have assumed zero inlet losses.5 Thrust and Fuel Flow Model As with the drag model.10 is a sample representation of thrust specific fuel consumption referred ( tsfcr ) versus referred net thrust ( Fn / δ t 2 ).11) 193 . There will be two separate models.0: Fn δt2 (17.15 (17. Fnr = For M < 1. tsfcr = tsfc θt 2 (17. In this case. 17. climb. Figure 17.9) Tt 2 = T ⋅ (1 + 0.8) (17.2 ⋅ M 2 ) Ideally.6) The parameter will sometimes generalize by dividing by the square root of the total temperature ratio. 17.5 (17.We now have all of the required components for a sample drag model.10) δ t 2 = δ ⋅ (1 + 0. One will be for nonafterburner engine operation and the other will be for maximum afterburner. tsfc = Wf Fn (17.2 ⋅ M 2 )3.
0 0. generally.000 feet RNI = 0.1 1. This is the ratio of Reynolds at the test condition to the Reynolds number at sea level.000 feet RNI = 0.12) TSFC/sqrt(thet2) Vs.7.792E05x + 1.000 3.000 5.000 feet we would have a 5percent degradation. Hence.2.9 0.8 1.9216 ⋅ M 7 ( 7 ⋅ M 2 − 1) (17.000 Fn/delt2 .3 4 .For M ≥ 1: 2.13) 194 .000 7.000 10.4010 − 0. 50.000 feet. a formula for a multiplicative factor on tsfcr would be as follows: Ftsfcr = 1 + (0.000 8. standard day.5 δ t 2 = δ ⋅ 166. an additional term will be added to our tsfcr model. for the same test day Mach number.2 TSFC/sqrt(thet2) 1.000 6. For standard day.000 4. A typical degradation in tsfcr is on the order of ¼ percent per 1. We will assume the above curve is valid up to a Reynolds number corresponding to a standard day at 30.046E08x 2 .000 2.606E16x 1.000 9.lbs Figure 17. degradation in the parameter with increasing altitude (or decreasing Reynolds number).000 feet of altitude. Therefore.4010. and b.1661) (17.1661. Fn/delt2 y = 1.000 11.10 Thrust Specific Fuel Consumption To better illustrate real effects.324E+00 1. for 20. There is.4010 − RNI ) ⋅ 0.05 (0. The parameter Reynolds number index ( RNI ) is introduced in the lift and drag section. 30. we have the following values for RNI : a.265E12x 3 + 1.
15) ηr = inlet temperature recovery factor. Then. That means that the thrust is constant to some value of inlet total temperature.0 assumption is made with actual data analysis.4010.4010 (17. Therefore. the model presented in Figure 17. we will construct a generalized form. we will presume that ηr = 1. the equation for our model is as follows: Fnr = 9.2129 Ftsfcr = 1. We will take a value of 9.15 degrees K).7 Military Thrust For maximum thrust without afterburner.01⋅ (Tt 2 − 288. Usually. After that point. For standard day.11 is a graphical representation of the above equations.000 pounds as the flat rated value of referred net thrust. we will assume a relationship of referred net thrust versus inlet total temperature ( Tt 2 ). First. For this model. 17.12 is for thrust versus Mach number as a function of altitude.15 ) if Tt 2 > 288. We must emphasize again. the thrust will decrease at some lapse rate.or: Ftsfcr = 1 + (0.4010 − RNI ) ⋅ 0.0. We shall presume the lapse rate to be 1 percent per 1. Tt 2 = ηr ⋅ Tt where: (17. 000 ⋅ 1 − 0. 000 if Tt 2 ≤ 288. It should be noted that this model is highly idealized. typically.15 Fnr = 9. For our model. the ηr = 1. An actual model will have altitude and Mach number effects. that the models presented here are very simplified and are presented to illustrate general trends only.15 (17. 195 . we have already introduced the parameter called referred net thrust.14) The above multiplicative factor is a number greater than one for Reynolds number indices less than 0. we have a simplified model for fuel flow for nonafterburning. With that term.17) Figure 17. A turbine engine is often said to be flat rated.0 if RNI > 0.16) (17. the recovery factor is difficult to measure and even more difficult to model anyhow. usually designated MIL power. We will presume that value to be standard day sea level temperature (288.0 degree K.
7 0.Referred Net Thrust versus Total Temperature: MIL Thrust 10.000 2.000 9.000 ft 20.000 8.5 0.000 4.lbs 220 240 260 280 300 320 340 360 380 Total Temperature .000 1.6 Mach Number 0.000 5.000 ft 40.000 7.000 8.000 3.3 0.000 5.000 3.000 2.12 Military Thrust 196 .2 Sea Level 10.000 Net Thrust (lbs) 6.deg K Figure 17.0 Figure 17.11 Military Referred Net Thrust Net Thrust versus Mach Number (Nonafterburning) Standard Day 10.9 1.000 9.000 ft 35.000 0.8 0.000 200 Referred Net Thrust Fn/delt2 .500 ft 0.000 7.000 ft 30.4 0.000 6.000 ft 42.000 1.000 4.
15 ) if Tt 2 > 288.0 percent per 1. 197 .000 20. The equations for referred thrust are as follows: Fnr = 20. ignoring Mach number and altitude effects.000 18.000 12.000 6. the formulas for the pressure ratio are presented for an assumption of a normal shock inlet.15 (17. A normal shock inlet is one where the recovery is across a normal shock. First. we will construct a similar model. For the maximum thrust with afterburner model.000 pounds. a lapse rate of ½ percent is used instead. So.15 Fnr = 20. Referred Net Thrust versus Total Temperature Maximum Afterburner Th t 22.14.0 degree K) but ran into the effect of thrust going to zero within the range of achievable total temperatures.005 ⋅ (Tt 2 − 288.18) (17. By comparison.000 Referred Net Thrust (Fn/delt2) lbs 16.deg K Figure 17.00 200 220 240 260 280 300 320 340 360 380 400 Total Temperature . We took a flat rated value for referred thrust of an even 20.17. we were going to use the same lapse rate (1.000 14.13.000 10.13 Referred Net Thrust for Maximum Thrust The maximum thrust model is presented as net thrust versus Mach number as a function of altitude for standard day in Figure 17.000 8.000 pounds.8 Maximum Thrust For maximum (MAX) thrust.19) A graphical representation of the model is shown in Figure 17. 000 ⋅ 1 − 0. the static sea level uninstalled thrust ratings in the F16 engines are (as of this writing) on the order of in excess of 25. 000 if Tt 2 ≤ 288. This is just what you have in a Pitot probe. This model is also highly idealized.
tsfcr = 2. arbitrarily.000 ft 25.21) 198 .000 20.4 0.6 0.9 Cruise Using the previously developed drag and fuel flow models. Experience has shown that thrust specific fuel consumption during maximum afterburner operation is at least 2.Net Thrust (with Afterburning) versus Mach Number (Standard Day) 30. RF = Vt ⋅ Wt (nam) Wf (17.8 Figure 17.000 ft 20.000 Net Thrust (lbs) 15.14 Maximum Thrust For fuel flow during maximum thrust operation.8 1. Let us.000 0 0. we can compute cruise parameters.000 10. assume a value of 2. The parameter range factor was developed in the cruise section and is repeated here.000 ft 50.0.0 Mach Number 1.4 1.20) An equivalent form of the equation is as follows: W 661. 17.5: a.000 ft 40.000 ft 30.2 0.000 5.6 1. we will assume a very simple model.48 ⋅ M ⋅ t δ RF = W f δ⋅ θ (17.50 .000 Sea Level 10.2 1.
22 yields: Wt 661. but will be used in this text.000675 ⋅ D M 2 ⋅S (17.000675 ⋅ t δ CL = M 2 ⋅S CD = 0. W f tsfc Fn ⋅ δ ⋅ θ = θ δ (17. we will use the convention of ‘corrected’ for ambient conditions and ‘referred’ for total conditions. we know that both drag and engine thrust specifics vary with Reynolds number. W 0.21 and corrected fuel flow in equation 17. 199 . That derived from the simplified forms of lift and drag coefficient for 1g flight and thrust equals drag.23) tsfcr = tsfc referred tsfc θt 2 This may not be a universal convention.The term in the denominator is called corrected fuel flow and can be expressed in another form. Hence. then drag would be constant.28) However. (17.24) Combining the range factor in equation 17.25) The concept behind the old constant weightoverdelta ( Wt / δ ) method of testing was that if one kept M and Wt / δ constant.22) In order to differentiate between dividing by total or ambient conditions.27) Fn = D δ δ ( ) (17.48 ⋅ M ⋅ δ RF = Fn tsfcc ⋅ δ (17.26) ( δ) (17. tsfcc = tsfc θ corrected tsfc (17.
9 1. Figure 17.000 2. CL max = 1.000 ft 35. 200 .000 4.500 ft Figure 17. The maximum lift coefficient is simply: a.000 5.15 is a plot of range factor for a series of altitudes.50 value for maximum lift coefficient for all the problems in this section.500 2.500 pounds gross weight.000 0.000 ft 40.50 . We will use the same 1. The topic of optimum flight profiles is a topic that will not be covered in this section.16) peak range factor versus weightoverdelta. The thrust available is deemed to be that determined from our military thrust model.500 lbs) 6.000 ft 42. we can compute the parameter range factor.15 Range Factor By picking off the peaks of the curves we can plot (Figure 17. attaining a maximum lift coefficient or thrust required exceeding the thrust available. Range Factor versus Mach Number (Weight=22. Either the minimum Mach number is dictated by the left scale of the plot.500 1.17.500 3.500 5.10 Range For our model aircraft on a standard day.4 0.3 0.5 0.8 0.000 Range Factor (nm) 3.6 0.000 ft 20.500 4.000 1. at 22. but suffice it to say that in a sense the closest distance between two points is not necessarily a straight line.000 ft 30.7 0.0 Mach Number Sea Level 10.
The weightpressure ratio ( Wt / δ ) is 125.000 40.500 2.000 5.500 4. Altitude Effect (W/delta=125.Constant Altitude Cruise: Weight=22.400 5.17 Range Factor – Altitude Effect 201 .90 0.000 120.88 0.000 140.000 lbs) 6.500 3.000 lbs 5.000 80. The engine blades are experiencing the same skin friction drag effects as the aircraft wing and other surfaces.000 5.500 5.72 0.000 Figure 17.76 0. This is due to skin friction effects on both aircraft drag and on the engine.92 0.800 0.lbs 100.NM 4.16 Maximum Range Factor Figure 17.86 0.800 Range Factor (nm) 5.500 lbs Weight=25.000 lbs Weight=22.000 60.000 Weight/delta .17 illustrates the effect of Reynolds number on cruise performance and demonstrates that you do not get the same range factor at a given Wt / δ and Mach number regardless of altitude (or temperature).600 5.80 0.74 0.000 2.78 0.000 pounds for all the data in the next two plots.000 4.000 Range Factor .000 20.500 lbs: Range Factor versus WeightOverDelta 6.94 Mach Number Figure 17.82 0.84 0.000 3.200 Weight=20.70 0.
Taking the midweight as the baseline.85 Mach number.000 RNI 0.800 Range Factor (nm) 5.000 22.500 lbs:Altitude=40. This number is comparable to the actual flight test derived values shown in the cruise section for three different aircraft.72 0.030 40.94 Mach Number Figure 17.400 T=+20 deg K T=Std Day T=20 deg K 5.At 0.86 0.76 0. Table 17.78 0.000foot change in altitude calculates to 0. Weight=22.3.000 5. we can also vary temperature and keep altitude and weight constant.85 Mach number and weightpressure ratio.580 38.200 5. This will achieve a variation in Reynolds number. the effect of temperature is shown in Table 17.2 RANGE FACTOR VARIATION WITH ALTITUDE Altitude (ft) 43.2 summarizes the numbers off the above plot. as shown in Figure 17.18.92 0.18 Range Factor – Variation with Temperature At the same 0.2322 0.39 percent.74 0.8 0.580 ft 6.88 0.800 0.388 Weight (lbs) 20.82 0.500 25. Table 17.7 5794.2612 0.2903 Range Factor (nm) 5736.84 0.9 0.7 The percentage change per 1.3 5849.000 4.600 5. 202 .
6 Std (216.000 0.7 0.000 1.8 0.836.500 1.9 1. we would need to find the Mach number for minimum fuel flow.2 and 17.19 is a plot of fuel flow versus Mach number for the same weight and altitudes considered for range.794. To compute Ps from our model the following computations are performed.000 ft 30.11 Endurance For the case where it is desired to maximize endurance.65) 0.3 0. 17.Endurance 17.3 +20 (236.19 Fuel Flow .000 2.5 0.500 5. The drag and thrust models are defined in previous parts of this section. The parameterspecific excess power ( Ps ) was defined in the axis systems and equations of motion section.3.000 ft 40.736. Figure 17. 203 .3 RANGE FACTOR VARIATION WITH TEMPERATURE Temperature Above Standard (deg K) Reynolds Number Index Range Factor (nm) 20 (196.000 ft 35.000 4.0 Figure 17.2 Sea Level 10.12 Acceleration Performance Acceleration performance will be computed using our model.65) 0.2612 5.Table 17.65) 0. Fuel Flow (Wt=22.6 Mach Number 0.4 0. it can be seen that the slope of range factor versus Reynolds number index is essentially identical between varying altitude and weight at constant weightpressure ratio and varying ambient temperature. Both will achieve a variation in Reynolds number index.500 Fuel Flow (lbs/hr) 4.500 lbs) 6.500 2.2977 5.2312 5.500 ft 0.000 5.000 ft 42.8 By comparing the numbers Tables 17.000 3.000 ft 20.500 3.
000675 (17.32) (17.CD = f (CL . 204 .500 feet. M .000 feet was chosen for Figure 17. the model computes a just barely positive Ps .36) Vt = 1116.34) (17. our model and the above calculations produce Figure 17.2 ⋅ M 2 ) Fnr = f (Tt 2 ) Fn = Fnr ⋅ δ t 2 θ =T 288.33) (17.15 (17.29) (17.20 for standard day.45 ⋅ M ⋅ θ (ft/sec) Fex = Fn − D Nx = Fex Wt Ps = N x ⋅ Vt 17.21. The above altitudes and weights were chosen to be the same as for the cruise. where Ps could be considered the rate of climb achievable for constant true airspeed.30) Tt 2 = T ⋅ (1 + 0.13 Military Thrust Acceleration For military thrust (maximum without afterburner). To illustrate the effect of temperature on acceleration performance. RNI ) (δ ⋅ M D= 2 ⋅ S ⋅ CD ) 0.35) (17. At 42.31) (17. an altitude of 10.
8 Sea Level 10.9 1.000 ft 42.MIL Thrust Specific Excess Power (Wt=22.3 0.0 Figure 17.20 Military Thrust Specific Excess Power Ps versus Mach Number (Weight=22. Temperature Effect 205 .4 0.6 Mach Number 0.7 0.000 ft 35.5 0.500 lbs) 200 180 Specific Excess Power .000 ft 20.Ps (ft/sec) 200 T=20 deg K T= Std Day T=+20 deg K 150 100 50 0 0.6 Mach Number 0.000 ft 30.3 0.000 ft 40.21 Military Thrust – Specific Excess Power.0 Figure 17.8 0.500 ft 0.500 lbs) 250 Specific Excess Power .2 0.Ps (ft/sec) 160 140 120 100 80 60 40 20 0 0.4 0.9 1.5 0.7 0.2 0.
There is. Fnr = 9.5 0. At the point of minimum drag.000 Thrust and Drag (lbs) 6.000 0 0.16 and 17.9 1. Mach number is 0. There is a small drag difference due to skin friction as illustrated in Figure 17.8 0. 000 ⋅ (1 − 0.The above difference in acceleration (and hence. This produces net thrust versus Mach number for 10.22.17: a.01 ⋅ [Tt 2 − 288.6 0.15 . b.4. climb) performance as a function of temperature is due primarily to thrust. we have the following points from the model.22 Military Thrust – Thrust and Drag at 10. 000 for Tt 2 < 288.000 ft.000 Feet 206 .000 Thrust: T= Std Day Thrust: T=+20 deg K Drag (Std Day) 8.000 Thrust: T=20 deg K 10. and Fnr = 9. however.2 0. Wt=22.23. Drag is also plotted for standard day.3 0.42 in Table 17. To repeat the thrust model presented in equations 17.4 0.15 .15]) for Tt 2 ≥ 288.7 0.000 4. a small increase in drag at the higher temperatures due to skin friction.0 Mach Number Figure 17.000 feet pressure altitude as shown in Figure 17.500 lbs) 12.000 2. Thrust and Drag (10.
55 Mach Number Figure 17.35 0.Drag versus Mach Number (Weight = 22.5pound difference in drag.3) 1.000 ft) 2.000 pounds for total temperature below 288. is quite small for purposes of acceleration performance.200 2.40 0.500 lbs.23 Drag at 10.3) 1.4 DRAG VARIATION WITH TEMPERATURE Temperature (deg K) Drag (lbs) 20 (248. In addition. between ±20 degrees K of standard day at 10. we get to travel through the transonic region where some interesting drag effects may occur.900 1. Altitude=10.850 1.5 Now.3) 1. It’s just that the numbers are larger.800 0.825. However.5 +20 (288. The sea level rating for F16 engines are somewhat larger than that number. Be aware. By installing an engine in the aircraft.24.841.000 Feet – Temperature Variation Table 17.0 Std (268. those 16. if the aircraft were doing endurance tests. The thrust model presented earlier had a referred net thrust of 20.000 1. this 16.050 Drag (lbs) T= Std Day T= +20 deg K 2.15 (standard day sea level). 207 . however.45 0.100 T= 20 deg K 2. that a rating is uninstalled. we present the standard day Ps plot in Figure 17.50 0.150 2.000 feet.5 pounds would be almost a full 1 percent.950 1.833. you will incur substantial inlet and other losses. First. 17.14 Maximum Thrust Acceleration The analysis of data for maximum thrust is identical to that for military thrust.
Ps .1 1.4 1.2 1. Note that the temperature deltas this time are only 10 degrees K.4 1.(ft/sec) 700 600 500 400 300 200 100 0 0.5 1.000 ft) 400 350 Specific Excess Power .4 0.2 1. 30.Ps versus Mach Number (Weight=22.6 Mach Number T= 20 deg K T= Std Day T= + 20 deg K Figure 17.24 Maximum Thrust Specific Excess Power As we did with military thrust.000 feet to conduct a comparison.500 lbs.(ft/sec) 300 250 200 150 100 50 0 0.000 ft 30.2 0.000 ft Figure 17.000 ft 50.6 0.6 1.0 1.8 2.25.0 1. This time we will choose 30.000 ft 20. we shall examine the effect of temperature on acceleration performance. versus 20 degrees K for the military thrust case.500 lbs) 900 800 Specific Excess Power . Maximum Thrust Effect of Temperature (Wt=22.Ps . This Ps comparison is shown in Figure 17. In addition.8 1.000 ft 40.000 Feet 208 .0 Mach Number Sea Level 10.25 Maximum Thrust Specific Excess Power Temperature Effect at 30.3 1. the thrust model chosen had only a ½ percent per degree K slope.9 1.
There were other rules: the time would be computed for a standard day and with the weight held constant at a midcombat weight.6 Mach number at 30. Nx = ( Fn − D ) Wt = ! ! Vt h Ps + = Vt g 0 Vt (17.We chose to plot only between 0. A specification compliance condition was the time to accelerate from 0.9 to 1.000 feet above Edwards AFB are all greater than standard day.39) (17. The prototype F16 (YF16) was involved in a flying competition with an aircraft designated the YF17 (later evolved into the Navy F18) in 1974.15 degrees K. We were never able to accelerate the YF17 aircraft to 1.26.42) t = 30.000 feet.915 ⋅ 32.65 ⋅ ∆M ∆M = 30. standard day ambient temperature is –44. which is a 25.174 ⋅ N x Nx (17.65 ⋅ M (17. Also shown is a second thrust model.60 Mach number for a specific reason.71 ∆t = Finally.38) (17.60 Mach number at 30. The competition (YF16) had no problem getting to 1. 209 .000pound model with the same ½percent lapse rate beginning at 288.000 feet.40) At 30. Vt = 1116. A little historical footnote here to illustrate the criticality of getting data at as cold a test day ambient air temperature as possible at 30. 288. The YF17 performance tests were conducted in late summer and early autumn. In ! addition.45 ⋅ M ⋅ 228.90 to 1.37) We also had zero wind. To compute time is a simple numerical integration. since we are accelerating at constant altitude.915 ⋅ M =1.15 = 994.60 1 M = 0.174 ⋅ N x ∆Vt = 32. because the above equation is only valid for zero wind. One of the performance specification points was the time to accelerate from 0. the h term is zero.000 feet on a standard day.60 Mach number even on days hotter than standard.9 and 1.174 ⋅ N x (17.43) The results of the time integration as a function of ambient temperature are shown in Figure 17.9 N x ∑ ⋅ ∆M (17.60 Mach number on a test day. ! Vt = g0 ⋅ N x = 32. In Appendix A note that the average temperatures at 30.174 ⋅ N x ∆t ∆t = ∆Vt 32.44 degrees C (easy number to remember) = 228.71 degrees K.41) 994.000 feet.
an iteration scheme is required.44) Fn = D (17. In order to achieve that condition. From drag coefficient. we will ignore the angleofattack component and simplify to: (17. compute drag.Time to Accelerate: 30. then solve for N z . then lift. thrust must equal drag.15 Sustained Turn A sustained (or stabilized) turn is a constant altitude. Fn = Fg ⋅ cos(α + it ) − Fe = D For this example.9 to 1. L = N z ⋅Wt (17.46) Knowing thrust.000 lbs Thrust= 25. Here are the basics of what was used in this example.6 Mach Number (sec) 200 Thrust= 20.000 ft: Weight=22.500 lbs 250 Time to accelerate 0.000 lbs 150 100 50 0 20 15 10 5 0 5 10 15 20 Delta temp above standard (deg K) Figure 17. We know drag coefficient from the following: 210 . then drag coefficient. find lift coefficient. constant speed turn.45) We will make a similar simplification in the normal axis (perpendicular to the velocity vector).26 Acceleration Time – Variation with Thrust 17. Since we do not usually have lift coefficient as a function of drag coefficient.
The results for maximum thrust are shown in Figure 17.0 1. just simply repeat the process a few times until the change in CL is sufficiently small (say < 0. ∆CD ∆CL 2 = 0.000 ft 40.27 Maximum Thrust – Sustained Turn Normal Load Factor 211 .48) For the first iteration.10.4 0. Nz versus Mach Number (Wt= 22.7 0.Nz 7.27.001) between steps.000 ft 1.01)) − CD ( f (CL )) 2 (CL + 0.CD = 0.01.3 Sea Level 10.000 ft 30. ∆CD ∆CL 2 = CD ( f (CL + 0.0 2. Solve for CLnew from the above equation.1 1.6 1.000675 ⋅ Fn δ ⋅M2 ⋅S (17.0 Normal Load Factor . Now that you know lift coefficient.47) Begin at 1g. but use some positive drag polar slope for the first iteration.2 1.0 1. After the first iteration. such as 0. then just compute N z .0 3.6 0.8 Mach Number Figure 17. computed from the available net thrust.01) 2 − CL (17.1 = (CDnew − CDold ) (CL new − CLold ) 2 2 (17.4 1. We always know the new CD . This is necessary since the slope of the drag polar at 1g may be zero or even negative.0 6.3 0.5 1.49) Then. It is the one above.9 1.5 0.500 lbs) 9.0 8. the old values of CL and CD are the 1g values.7 1. we used 0.8 0. compute values for the slope numerically by choosing some small change in lift coefficient and computing the slope.000 ft 20.0 4.0 5.0 0. For instance.
a. and N z max = 9. d. CL max = 1. 212 . b. c. CL < CL max .The constraints imposed on this turn problem were the following.50 . N z < N z max .0 .
1).1 C17A Aircraft Wt RF = 661. W F CL = 0.1). Wt W f = 661. Pressure ratio ( δ ) formulas used are found in the altitude section.1) were corrected to a reference Reynolds number.18.3) Since the data presented in the report (Reference 18.1) ( ) Solving for corrected fuel flow. an estimate of drag at test and reference conditions was computed.1) testing report titled.000675 ⋅ t − g ⋅ sin(α ) δ δ (18.0 CRUISE FUEL FLOW MODELING This section had contained a regression analysis model of fuel flow and thrust extracted from the AFFTC C17A (Figure 18. Instead 213 . but since this handbook is intended for public viewing. Figure 18.2) The lift coefficient was computed using the curve fits for angle of attack ( α ) and gross thrust ( Fg ) provided in the report (Reference 18. “C17 Cruise Configuration Performance Evaluation” (Reference 18. it was necessary to delete the scales on the data plots shown in this section.48 ⋅ M ⋅ W f δ θ δ (18.48 ⋅ M ⋅ δ W fC = RF δ θ ( ) (18.
6) The assumption was made that the characteristic length used was the mean aerodynamic chord ( MAC ).15 θ 2 RN = 7. To perform a curve fit of the fuel flow data.101 ⋅106 ⋅ M ⋅ RNI ⋅ l (18.455 / log10 (1.9) 214 .5) (18.144 ⋅ M 2 )0.of the usual ‘standardization’ we are essentially ‘unstandardizing’ the drag data. ft2.4) (18.000 per foot.00238 From a formula defined in the lift and drag section. That value is as follows: l = MAC = 25. S = 3. Skin friction drag relationships are as follows: C f = 0. Inserting these numbers into equation 18.800. S wet = 19. 000 ⋅ 25.800.58 C fC = C f /(1 + 0. The test day drag coefficient ( CDt ) was computed as follows: CDt = CDs + (CDft − CDfs ) (18.800.455 / log10 ( RN )2.8) (T + 110 ) δ RNI = ⋅ 398. We are going from a reference condition to a standard condition. The formulas used are those presented in the lift and drag section.65 CD f = S wet ⋅ C fC S (18. and b.58 = 0.4: C fs = 0. The reference wing area ( S ) and the wetted area ( S wet ) are as follows: a.794 feet. (18. The standard day drag coefficient ( CDs ) was computed from the drag polar curve fit formulas in the report.075. we will remove the skin friction drag correction from the thrust data.7) The standard (or reference) skin friction drag coefficient is based upon the standard Reynolds number per foot and the characteristic length. ft2. The drag coefficient formula in the report was referenced to a Reynolds number of 1.794)2.
H3> 30.10) 18.81 H3: M 0.2) is a plot of the 141 data points being analyzed.000 feet] 0.11) The following (Figure 18. The maximum value of the dependent variable ( tsfc / θ ) is 11.0 percent.000.000.77 to 0. H2 20. We will take out the correction to a reference Reynolds number.64 0.60 H2: M=0.76 H3: M 0. however. since the data points are still at standard day temperatures. it will however give you some interesting information. This is a large variation.70 to 0.60 H2: M 0.825 TSFC/sqrt(theta) (1/hr) 0. so we proceed to curve fit the data using multiple regression.000675 (18.76 H2: M 0.77 to 0. it should be noted that range factor had a 14.1 Thrust Specific Fuel Consumption Next. The use of these ‘generalizing’ parameters is a good first step in modeling your data.50 to 0.2 percent greater than the mean and the minimum value if 17.56 0. the test values of corrected thrust are computed.70 to 0.50 to 0.46 H1: M 0.9 percent less than the mean.Finally.000 to 30. TSFC/sqrt(theta) versus Fn/delta [H1<10. [ Fn / δ ]t CDt ⋅ M 2 ⋅ S = 0.3percent variation about its mean (more than twice as much – percentage wise).2 Thrust Specific Fuel Consumption 215 .60 0.68 H1: M <= 0.72 0.52 20000 40000 60000 80000 Fn/delta (lbs) 100000 120000 140000 Figure 18.76 0. That is analogous to drag where we use lift and drag coefficients to aid in modeling. The 1sigma about the mean is 7. We still wish to reduce this variation. Even though the plot has no scales. Note a distinction between test values and test day.65 H2: M 0. we compute the thrust specific fuel consumption corrected as follows: W f / δ ⋅ θ t TSFCC = TSFC / θ = [ Fn / δ ]t ( ) (18.
we will make the following variable name changes: a. X 2 = M . X 3) (18.i i ∑ X m . the following is the formulation for multiple regression. and X 3 = RNI .i " 1. d. we will strive to develop an equation that fits the data presented in Figure 18.i ∑X ∑X ⋅X ∑X 2.i ∑X ∑X ∑X ⋅X 1. Excel has matrix operators. The general case for linear multiple regression: Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 2 + " + am ⋅ X m The coefficients are solved by the following: a0 a 1 a2 = # am N (18. the formula will be as follows: TSFC / θ = f (( F n / δ ) . For those who do not have a multiple regression program available. In general. b. X 2.2 Multiple Regression Now.i # 2 X m .13) The author used MS Excel to evaluate the data.i i 2. The simplest possible equation is a constant.12) For ease of representation.i 2.i # ⋅ X 1.2. X 1 = Fn / δ . We will use Reynolds number index ( RNI ) as an altitude parameter.i −1 ∑Y X ⋅Y ∑ ∑ X ⋅ Y (18.i 2 2. however it was necessary to develop a multiple regression method for use with Excel. equivalently: Y = f ( X 1.i 2.i " " ∑X ∑X ⋅X ∑X ⋅X m .14) ∑X ∑X # 1. M .15) i 1. The above general curve fit formula was developed by minimizing the sum of the squares of the residual errors ( SS ). The formula for SS is as follows: 216 .i 2.i ⋅ 2.i ∑X m .i 2 1.i # m .i ⋅ Yi # where: N = number of data points.i ∑X ∑X m .i ⋅ X 2. c. Y = TSFC / θ . Then. RNI ) (18.i " ∑ 1.18.i 1.i m .i m .
and ˆ c.3.0 percent to 5.30% .16% . So far. we have only reduced the 1sigma about the mean from 7.16 percent.19) ˆ The Y used will be from the last curve fit (equation 18. Here are the results of the curve fits: ˆ a. ˆ b.00% . At this point. 217 . we should pause to examine the residual errors rather than just blindly adding additional terms to the equation. (18. Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 12 %σ = 5.17) %σ = (σ / Y ) ⋅100 where: (18. we suspect there may be a substantial altitude and Mach number effect.18) Y = the mean value of the independent variable. σ = SS /( N − 1) A percentage standard deviation will be calculated. There are a number of ways to evaluate the quality of a curve fit.18).16) ˆ Y = the curve fit equation. we can see some apparent additional Mach number and Reynolds number effects. Y = a0 %σ = 7. The parameter we will plot is the percentage error as follows: ˆ Y % Error = Y − Y ⋅ 100 ( ) (18. This is a disappointing result. however. We will look at the standard deviation.ˆ SS = ∑ Yi − Yi where: ( ) 2 (18. From Figure 18. Y = a0 + a1 ⋅ X 1 %σ = 5.
ˆ b.000 Feet Altitude > 30.% Error in TSFC/sqrt(theta) Versus Mach Number 15 10 % Error in TSFC/sqrt(theta) 5 0 5 10 Altitude < 20.55 0.3 Percentage Error in Thrust Specific Fuel Consumption We can now proceed to add additional terms to our model.65 0. Throttle settings above and below that required for stabilized cruise should be included in any fuel flow model. To quote from the report (Reference 18. There is no reason for that with the advent of INS and GPS measurements to give 218 .000 to 30.” They required “not less than 2. This particular data set was not a very good one to develop a complete fuel flow model.224% . ˆ c.80 0.50 0. Actually.35 0. At this point. Enormous quantities of flight time were expended to collect these relatively few cruise data points.230% .75 0.85 Mach Number Figure 18.30 0.40 0. There were no data collected below 6.60 0.000 Feet Altitude 20.229% . we did not make significant gains past equation (a) but proceeded just to illustrate what additional gains were made.237% . and ˆ d. a.000 feet pressure altitude. The C17A project (Reference 18.1). for instance. Only stabilized cruise data points were used.1) illustrates that too much time was expended collecting cruise data. The stabilization criterion was much too stringent. Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 12 + a3 ⋅ X 2 + a4 ⋅ X 3 + a5 ⋅ X 22 %σ = 1. no significant additional gains are evident.45 0. “it was not uncommon for a single cruise point to take 20 minutes to complete.70 0. Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 12 + a3 ⋅ X 2 %σ = 1.000 Feet 15 20 0.5 minutes of stabilized data” on each data point. Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 12 + a3 ⋅ X 2 + a4 ⋅ X 3 + a5 ⋅ X 22 + a6 ⋅ X 32 %σ = 1. Y = a0 + a1 ⋅ X 1 + a2 ⋅ X 12 + a3 ⋅ X 2 + a4 ⋅ X 3 %σ = 1.
a much more complete fuel flow model could have been obtained at a much lower cost in terms of flight time. a few seconds of data is all that is required.100. and Chester Gong.000 25.800.000 40. This is more than a factor of two less than the degradation factor of older generation aircraft such as B52 aircraft.W/delta=1.W/delta=1.0 REFERENCE 18. AFFTCTR9323. To present just a few of the data points we choose to present those that illustrate an altitude effect.6.1).800. 219 . The data points are all from the aforementioned C17 Cruise Performance report (Reference 18.000 Data:M=0.000 Range Factor (nm) 11000 10000 1.4 Range Factor Variation with Altitude The degradation factor of range factor with altitude was 0.000 pounds Wt / δ and 0. SECTION 18.000 Figure 18.26 percent per 1.20 percent per 1.000 feet at 1. California.800.000 Pressure Altitude (FT) 30. Edwards AFB.000 35. Range factor variation with altitude is shown in Figure 18.W/delta=1.4. Once some reasonable stabilization is achieved.1 Weisenseel.000 Model:M=0.W/delta=1.000 9000 10.instantaneous acceleration data.000 15. AFFTC. December 1993.60. Charles W. With the addition of a series of accelerations and decelerations at partial thrust.000 pounds Wt / δ .000 20. Range Factor versus Altitude 12000 Model:M=0.000 Data: M=0. C17 Cruise Configuration Performance Evaluation.77.100.100.77.000 feet at 1.
19.19.1 Equations V Acceleration factor AF = 1 + t g0 dVt ⋅ dH ! HE = ! H Aircraft geometric height (Edwards flyby tower) ∆htower = 31.0 EQUATIONS AND CONSTANTS This section is a summary of the primary equations and constants that were derived and used in this handbook.422 ⋅ (grid reading) T Aircraft pressure altitude (flyby tower data) H C a / c = H p tower + ∆htower ⋅ std T cos α [α ] = 0 − sin α 0 sin α 1 0 0 cos α Alpha transformation body to flight path Angle of attack α = tan −1 (Vbz Vbx ) Angle of attack (zero bank) α = θ − γ Angle of sideslip β = sin −1 Vby Vt Aspect ratio AR = b 2 ( ) S sin β cos β 0 0 0 1 cos β Beta transformation body to flight path [ β ] = − sin β 0 Vbx VtN T T T Body axis airspeeds Vby = [φ ] ⋅ [θ ] ⋅ [ψ ] ⋅ VtE V V bz tD ! ! Body axis pitch rate q = θ ⋅ cos φ + ψ ⋅ cos θ ⋅ sin φ ! ! Body axis roll rate p = φ −ψ ⋅ sin θ ! ! Body axis yaw rate r = ψ ⋅ cosθ ⋅ cos φ − θ ⋅ sin φ 220 .
Calibrated airspeed (VC < aSL ) qC 2 VC = 1 + 0.5 −1 2.5 −1 (1 3.9216 ⋅ (VC aSL ) 7 ⋅ (VC aSL )2 − 1 7 2.5 q 1 Calibrated airspeed (VC ≥ aSL ) VC = aSL ⋅ 0.2559) / 6.5) q C Calibrated airspeed (VC < aSL ) VC = aSL ⋅ 5 ⋅ + 1 − 1 PSL Calibrated airspeed (VC ≥ aSL ) qC PSL = 166.87559 E − 6 Density altitude H d = 1 − θ ( θ ) Density ratio σ = δ θ Drag (test day) Dt = Fnt − Fex t 221 .5 Compressible dynamic pressure ( M ≥ 1) qC 7 = 166.5 − 1 P (7 ⋅ M 2 − 1) Corrected net thrust Fn / δ W f δ⋅ Wf Corrected thrust specific fuel consumption tsfc / θ = = Fn Fn ⋅ θ δ 1 δ ( 4.2 ⋅ PSL aSL 3.9216 ⋅ M 2.2 ⋅ M 2 ) − 1 3.881285 ⋅ C + 1 ⋅ 1 − 2 VC PSL 7 ⋅ aSL Cloverleaf method solves this equation (Vti + ∆Vt )2 = (VgN + VwN )2 + (VgE + VwE )2 Compressible dynamic pressure ( M < 1) qc P = (1 + 0.
00067506 ⋅ D δ ⋅ M 2 ⋅ S ( ) S Drag coefficient due to skin friction CD = C f ⋅ wet S Drag Model (given M ) CD = CD min + K1 ⋅ (CL − CL min ) + K 2 ⋅ (CL − CLb ) 2 2 K 2 = 0 when CL < CLb VwN Vt VgN Earth axis winds VwE = [ψ ] ⋅ [θ ] ⋅ [φ ] ⋅ [α ] ⋅ [ β ] 0 − VgE V 0 V wD gD Elliptic Wing Theory (M <<1) CL = 2 ⋅π ⋅α 2 1 + AR CDL = CL π ⋅ AR 2 Energy altitude H E = H + Vt 2 ( 2 ⋅ g0 ) V 2 PE KE + =H + t Wt Wt ( 2 ⋅ g 0 ) Energy per unit weight E / Wt = Equivalent airspeed Ve = σ ⋅ Vt Excess thrust Fex = N x ⋅ Wt Excess thrust Fex = [ Fg ⋅ cos(α + it ) − Fe ] − D Excess thrust test Fext = N x ⋅ Wt Ax cos β Flight path accelerations Ay = − sin β A 0 z sin β cos β 0 0 cos α 0 ⋅ 0 1 − sin α 0 sin α Abx 1 0 ⋅ Aby 0 cos α Abz 222 .Drag coefficient CD = D / ( q ⋅ S ) Drag coefficient CD = 0.
400 Ideal gas equation of state P = ρ ⋅ R ⋅ T − sinψ cosψ 0 0 0 1 223 . Ax T T T T T Flight path accelerations Ay = [ β ] ⋅ [α ] ⋅ [φ ] ⋅ [θ ] ⋅ [ψ ] A z ! Flight path angle γ = sin −1 h Vt N x Axf g0 Flight path load factors N y = Ayf g 0 N − A g z zf 0 AN ⋅ AE A D (VgN + VwN ) Vt Flight path to earth axis transform (VgE + VwE ) = [ψ ] ⋅ [θ ] ⋅ [φ ] ⋅ [α ] ⋅ [ β ] ⋅ 0 0 (VgD + VwD ) dWt Fuel flow W f = − dt Geopotential altitude g ⋅ dh = g 0 ⋅ dH r0 Geopotential vs. geometric altitude H = ⋅h ( r0 + h ) ! Gross thrust Fg = Wa + W f ⋅ Vexit + Aexit ⋅ ( Pexit − P ) Groundspeed east VgE = Vg ⋅ sin(σ g ) Groundspeed north VgN = Vg ⋅ cos(σ g ) ( ) cosψ Heading matrix (rotate about the z axis (or yaw)) [ψ ] = sin ψ 0 LHVtest Heating value corrected fuel flow W ft = W ft ⋅ 18.
328 RN Lift coefficient CL = 0.5] q Mach number ( M < 1) M = 5 ⋅ C + 1 − 1 P Ve Mach number from equivalent airspeed M = aSL ⋅ δ ( ) Normal load factor N z = − Az / g 0 Normal load factor in climb N z = cos γ + Vt ⋅ γ! g0 224 .881285 ⋅ P 7 ⋅ M 2 [1 3.5 ⋅ t ⋅ Vt g 0 2 Laminar skin friction empirical formula C f = Lateral load factor N y = Ay / g 0 Lift coefficient CL = L / ( q ⋅ S ) 1.5 ⋅ ρ SL ⋅ Ve 2 2 r0 Inverse square gravity law g = g 0 ⋅ ( r0 + h ) 2 W Kinetic energy KE = 0.5 1 qC + 1 ⋅ 1 − Mach number ( M ≥ 1) M = 0.Incompressible dynamic pressure q = 0.5 ⋅ ρ ⋅ Vt = 0.00067506 ⋅ L δ ⋅ M 2 ⋅ S ( ) ! ! Longitudinal load factor N x = H Vt + Vt g 0 Longitudinal load factor N x = Ax / g 0 Mach number M = Vt a 2.
2559) H= (6.87559 E − 6 ) Pressure ratio δ = P PSL {−[4. zero wind) N z = 1 + t ⋅ σ g0 Normal load factor in turn (constant altitude.089 feet 1 − (δ )(1 5.089 feet δ = 0.2559 Pressure ratio above 36.22336 ) Pressure altitude below 36.24)} 5.48 ⋅ M ⋅ t Wte δ dt Range for constant altitude (approximate) R = − ∫ W Wts Wt f δ⋅ θ ( ) 225 .84 ⋅ ln δ ( 0.089 feet δ = (1 − 6.V ! Normal load factor in turn (constant altitude.806343 E −5]⋅( H C −36089.22336 ⋅ e Pressure ratio below 36.87559 E − 6 ⋅ H ) ! Ram drag Fr = Wa ⋅ Vt W Range (approximate) R = RF ⋅ ln ts Wte Range factor RF = Vt ⋅ Wt = SR ⋅Wt Wf W 661.24 − 20805.089 feet H C = 36089. zero wind) N z = 1 Normal load factor times weight N z ⋅ Wt = L + Fg ⋅ sin(α + it ) 2 cos φ cosθ Pitch matrix (rotate about yaxis) [θ ] = 0 − sin θ Potential energy PE = Wt ⋅ H 0 sin θ 1 0 0 cosθ Pressure altitude above 36.
48 ⋅ θ 226 .15 θ 0 1 0 cos φ Roll matrix (rotate about xaxis) [φ ] = 0 sin φ cos β Sideslip matrix [ β ] = sin β 0 − sin β cos β 0 0 0 1 CDL = 2 ⋅ CL π ⋅ AR 2 0 − sin φ cos φ Slender Body Theory ( M ≈ 1) CL = π ⋅ AR ⋅ α 2 ! ! V ! Specific excess power Ps = H E = H + t ⋅ Vt = N x ⋅ Vt g 0 ( ) Specific range SR = Speed of sound a = Vt Wf (γ ⋅ R ⋅ T ) = 661.101E + 6) ⋅ M ⋅ l ⋅ RNI (T + 110 ) δ Reynolds number index RNI = ⋅ 2 398.Range for constant altitude (approximate) R = − RF ⋅ Wte Wts ∫W dt t W 661.48 ⋅ M ⋅ t δ Range for cruise at constant altitude R = − ∫ W Wts f δ⋅ θ Wte ( ) dt ⋅ Wt Range for cruise at constant altitude R = ∫ Vt ⋅ dt Reynolds number RN = ρ ⋅Vt ⋅ l µ Reynolds number RN = (7.
15 − 1.2559 = (1 − 6.65 °K Standard temperature below 36.089 feet T0 = 216.9812 E − 3 ⋅ H C Standardized drag Ds = Dt + ( Ds′ − Dt′ ) Standardized excess thrust Fexs = Fext + ( Fns′ − Ds′ ) − ( Fnt′ − Dt′ ) ′ ′ Standardized fuel flow W fs = W ft + W fs − W ft ( ) ′ ′ Standardized net thrust Fns = Fnt + ( Fns − Fnt ) Takeoff excess thrust Fex + µ ⋅ (Wt ⋅ cos(θ rw ) − L ) = Fn − D − Wt ⋅ sin(θ rw ) Temperature correction to pressure altitude change ∆h = T T ⋅ ∆H C STD Temperature ratio θ = T T = TSL 288.Standard day density ratio σ = δ 4.87559 E − 6 ⋅ H C ) θ Standard temperature above 36.089 feet T = 288.15 Theoretical tanker downwash angle ε 0 = ( 2 ⋅ CLt ) (π ⋅ ARt ) CDL = α ⋅ CL = M 2 −1 2 ⋅ CL 4 Thin Wing Theory (M > 1) CL = 4 ⋅α M 2 −1 Thrust horsepower THP = Fn ⋅ Vt (where Vt has units of feet/sec) 550 Thrust horsepower (user provided η and n ) THP = η ⋅ σ n ⋅ BHP Total energy E = KE + PE Total temperature Tt = T ⋅ 1 + 0.2 ⋅ M 2 ( ) ( ) 227 .
58 Vt 2 2 z Turn radius (constant altitude. zero wind) σ g = Vbx Vbxi 0 Velocity rate corrections Vby = Vbyi + −r V V q bz bzi Weight Wt = m ⋅ g 0 R r − q l x 0 p ⋅ l y − p 0 l z 228 .True airspeed Vt = (V 2 bx + Vby + Vbz 2 2 ) True airspeed down VtD = VgD + VwD True airspeed east VtE = VgE + VwE True airspeed magnitude Vt = (V 2 tN + VtE + VtD 2 2 ) True airspeed north VtN = VgN + VwN " " " True airspeed vector Vt = Vg + Vw Vt T T T T T True airspeed vector 0 = [ β ] ⋅ [α ] ⋅ [φ ] ⋅ [θ ] ⋅ [ψ ] 0 Turbulent skin friction empirical formula C f = VtN ⋅ VtE V tD 0.455 (log10 RN ) 2. zero wind) R = g0 ⋅ Vt (N −1 ) ! Turn radius (constant altitude.
17405 feet/sec² Reference radius of the earth ( r0 ) (from the 1976 U.855.553 feet Sea level standard temperature ( TSL ) = 288.4788 knots Standard sea level pressure ( PSL ) = 101.3048 (exactly) Conversion knots to feet/sec = multiply knots by 1.8136 feet²/(sec²°K) Viscosity at sea level ( µ SL ) = 3.852 meters = 6.15 °K Speed of sound at sea level standard day ( aSL ) = 1.7373⋅107 slugs/(feet sec) 229 .116.325 pascals (newtons/m2) = 2.2166 pounds/feet² Temperature in second segment of standard atmosphere ( T0 ) = 216.2 Constants Conversion feet to meters = multiply feet by 0.089. Standard Atmosphere) = 20.4505 feet/sec = 661.1155 feet Reference gravity ( g 0 ) = 32.19.S.65 °K Universal gas constant ( R ) 3.076.116.68781 Nautical mile ( NM ) = 1.
This page intentionally left blank. 230 .
APPENDIX A AVERAGE WINDS AND TEMPERATURES FOR THE AIR FORCE FLIGHT TEST CENTER
231
This page intentionally left blank.
232
AVERAGE WINDS AND TEMPERATURES FOR THE AIR FORCE FLIGHT TEST CENTER
The following average wind and temperature data were provided courtesy of the Edwards AFB weather squadron. The data represents average values obtained on a daily basis over a period of more than 30 years (1950s through 1980s). Figures A1 through A5 represent average temperature deviation data versus month for 10, 20, 30, 40, and 50,000 feet pressure altitude, respectively.
T e m p e ra ture from S tand a r d : P re s su r e Altitu d e = 1 0,0 00 F e et; A F F TC A ve ra g e D a ta; T e m p er atu re S ta n d a rd = 2 6 8 .3 4 d e g K
16 14 12
D elta T em pe rature fro m Stand ard (D eg K)
10 8 6
4 2 0
Jan F eb M ar Apr M ay Ju n Ju l Aug S ep O ct N ov D ec
M o n th
Figure A1 Delta Temperature at 10,000 Feet
T e m p e ra tu re fro m S tan d a r d : P re s su r e A ltitu d e = 2 0,0 00 F e et; A ve ra g e A F F T C D a ta; S ta n d a rd T em p e ra tu re = 2 4 8 .5 3 d e g K
14
12
D elta T em pe rature (Deg K )
10
8
6
4
2
0 J an Fe b M ar Apr Ma y Ju n Ju l Aug Sep O ct N ov D ec
M o nth
Figure A2 Delta Temperature at 20,000 Feet
233
T e m p e ra tu r e F ro m Sta n d ar d : P re s s u re Altitu d e = 3 0 ,0 0 0 Fe e t; Av er ag e AF F T C Da ta ; T e m p e ra tu r e S ta n d a rd = 2 28 .71 D eg K
10
8
D elta T em pe rature (Deg K )
6
4
2
0
2 J an Fe b M ar Apr Ma y Ju n Ju l Aug Sep O ct N ov D ec
M on th
Figure A3 Delta Temperature at 30,000 Feet
T e m p e ratu re fro m S tan d ar d : P re s su r e Altitu d e = 4 0,0 00 F e et: A F F T C a v er ag e d a ta; S ta n d a rd T e m p e ra tu re = 2 16 .6 5 d e g K
4
3
2
D elta T em perature (D eg K)
1
0
1
2
3
4 Ja n Feb M ar Ap r Ma y Jun Jul M o n th Aug S ep O ct Nov D ec
Figure A4 Delta Temperature at 40,000 Feet
234
T em p eratu r e fro m S tan d a rd : Pr es su re Altitu d e = 50 ,00 0 F eet : AF F T C A ve ra g e d ata ; S ta n d ard T e m p er a tu re = 216 .65 d eg K
M o n th Ja n 1 2 F eb M ar Apr Ma y Ju n Ju l Aug Se p O ct No v De c
D elta T em perature from Standa rd (D eg K)
1 0
8
6
4
2
Figure A5 Delta Temperature at 50,000 Feet
Figures A6 and A7 present average windspeed and direction versus month. They are presented at three different ambient pressure levels. These are in terms of pressures in millibar (mb). The following are the corresponding pressure altitudes: 1. 200 mb = 38,661 feet, 2. 400 mb = 23,574 feet, and 3. 600 mb = 13,801 feet.
W ind D ire c tio n ve rs us M o nth
3 20
3 00
W ind Direction (Deg rees from T rue No rth)
2 80
2 60
P = 200 m b P = 400 m b P = 600 m b
2 40
2 20
2 00
1 80 J an Fe b M ar A pr May J un J ul A ug Sep Oc t N ov D ec
M o nth
Figure A6 Wind Direction
235
the geometric height will not be equal to the pressure altitude.Pressu re Altitude (Fe et) 160 0 140 0 120 0 P = 200 m b 100 0 80 0 60 0 40 0 20 0 0 Ja n F eb Ma r Ap r M ay J un J ul Au g S ep Oct N ov D ec P = 400 m b P = 600 m b M o nth Figure A8 Geometric Height minus Pressure Altitude 236 . the geometric height (on average) is always greater than the pressure altitude. As can be seen. This is due to the fact (again on average) that the atmospheric temperature is greater than standard day for all months of the year through 30.000 feet. Figure A8 illustrates this difference for an average day above Edwards AFB.Windspeed versus Month 70 60 50 Wind Speed (kts) 40 P = 200 mb P = 400 mb P = 600 mb 30 20 10 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Figure A7 Windspeed On a given day.Pr essure Altitude versus M onth 200 0 180 0 G eo metric H eight . G eom etric H eight .
APPENDIX B WEATHER TIME HISTORIES 237 .
This page intentionally left blank. 238 .
Oc t 1 6O c t F L = 4 00 F L = 3 00 F L = 2 00 F L = 1 00 2 0O c t 24. 239 . respectively.WEATHER TIME HISTORIES The following charts represent time histories of data for September through October 1998. Figure B1 shows the variation of delta temperature above standard versus date.Se p 2 6S e p 3 0S e p 4 O ct 8 O ct Da t e ( 19 98 ) 1 2. the terminology flight level (FL) is used. On the charts. Flight level is pressure altitude in feet divided by 100. 200. D elta Tem per ature versus Date 16 D elta Temp eratu re Ab ove S tand ard D ay (Deg rees K) 14 12 10 8 6 4 2 0 2 4 2 2. 300 and 400.Oc t Note: FL = H C /100 Figure B1 Delta Temperature Time History Figures B2 and B3 illustrate the variation in windspeed and direction versus date at flight levels of 100.
Se p 4 O c t 8Oct Date (1998) 12Oct 16Oct 2 0 O c t 2 4 O c t Figure B3 Windspeed Time History 240 .W ind Dir ection ver sus Date 36 0 33 0 Wind Direction (D egrees from T rue North) 30 0 27 0 24 0 21 0 18 0 F L = 40 0 F L = 30 0 15 0 F L = 20 0 F L = 10 0 12 0 90 22 .Sep 26 S ep 30 .Sep 4O c t 8O c t D ate (1 99 8) 12 O ct 16 O ct 20 O ct 24 O ct Figure B2 Wind Direction Time History W in d Sp eed versu s D ate 120 100 FL = 400 FL = 300 FL = 200 W ind Sp eed (K nots) 80 FL = 100 60 40 20 0 2 2 S e p 2 6 S e p 3 0 .
APPENDIX C AVERAGE SURFACE WEATHER FOR THE AIR FORCE FLIGHT TEST CENTER 241 .
This page intentionally left blank. 242 .
AVERAGE SURFACE WEATHER FOR THE AIR FORCE FLIGHT TEST CENTER Figure C1 shows the average surface temperature for the Air Force Flight Test Center. Average Su rface Tem peratures 1 00 90 80 T emp erature (de g F) 70 M axim u m M inim u m 60 50 40 30 20 Jan Feb M ar A pr M ay J un Jul Aug S ep Oc t N ov De c M o n th Figure C1 Average Maximum and Minimum Surface Temperatures 243 .
244 .This page intentionally left blank.
3. 7. October 1976. Performance and Flying Qualities UFTAS Reference Manual. California. Theory of the Measurement and Standardization of InFlight Performance of Aircraft. 12. Ohio. (Out of print). U. Introduction to Flight. NASA TM 73. MILM7700D. D. Hans W.. Flight Test Measurement of Ground Effect for Powered Lift STOL Airplanes. April 1976. 14 February 1990. Tom.BIBLIOGRAPHY 1. Anderson.” Paper presented at the 16th Annual Symposium Proceedings 1985. 13. University of Dayton. John Wiley and Sons. July 1973. Jeff V. and David Nesst. 8. Air Force Flight Test Center. California. Russel M. Tippey. Air Force Flight Test Center. NASA TMX62. National Oceanic and Atmospheric Administration. Flight. Chapra. Olson. Herrington. and Milton Porter. Canale. Third Edition. 245 . Inc. Olhausen. Society of Flight Test Engineers. McGrawHill. 1976. Inc. McGrawHill. 333. No. Elements of Gasdynamics. 14. Military Specification. Dunlap. James N. Air Force Flight Test Center. Fundamentals of Aircraft Design. John D. 1985. Aircraft Performance. Standard Atmosphere. “The INS Wind Calibration in Climb Algorithm. 1985. FTCTD711. Moffett Field. NOAAS/T 761562. 256. 4. 15. Moffett Field. Numerical Methods for Engineers. New York.. February 1965. February 1988. 5. California. and Anatol Roshko. Ames Research Center. and Thomas Galloway. 11. Edwards AFB. AF TR 6273. Inc. 1989. Sweeney. Everett W.. 9. New York. 1975. Digital Performance Simulation. California. Kurt. Edwards AFB. Wayne M. California.. Bowles. New York. “Use of a Navigation Platform for Performance Instrumentation on the YF16.” Journal of Aircraft. December 1977. Flight Test Engineering Handbook. 4. Seattle. April 1971.August 2. 13. and Raymond P. Performance and Flying Qualities UFTAS Link13 Users Guide.. Edwin K. July 29 .. New York. California. Nicolai. California. Ames Research Center. 2.S.. Edwards AFB.. 10. Parks. Computer Programs for Estimating Takeoff and Landing Performance. et al. Leland M. revised January 1966. Manuals. January 1986.. Air Force Flight Test Center. October 1984. 6. Edwards AFB. Liepmann. Steven C. Vol. 16. Edwards AFB. Dayton. California. Air Force Flight Test Center. USAF Test Pilot School. Edwards AFB.
Inc.” Paper presented at the Society of Flight Test Engineers Symposium. GGD Publication 92013. Cheney. Engineering Aerodynamics. W. 1982. Aviator’s Guide To GPS. Dynamics of Flight. Aircraft Performance. “Measuring an Airplane’s True Speed in Flight Testing. Jones. Second Edition. “Airspeed Calibration Using GPS. John Wiley and Sons. 1923. 1990.G. McGrawHill Book Company. 1936. TN 135. DeAnda. Introduction to Avionics. Fox. July 1992. 1995. 1955.G. Diehl. Lawrence. 1997.” NACA Rep. 1976. 20. April 1964. June 1981. 18. 3rd Edition. Cambridge University Press. Second Edition. AFFTC Standard Airspeed Calibration Procedures.BIBLIOGRAPHY (Continued) 17. H. Society of Flight Test Engineers. John Wiley and Sons. David. Wind Tunnel Testing. Alan.” Paper presented at the 14th Annual Symposium Proceedings. August 1957. Brown. Technical Aerodynamics. 34. 23.” KITPLANES Magazine (February). TAB Books. Government Printing Office. August 1519. Bernard. Edwards AFB. Bill. Climatic Extremes for Military Equipment. 1994. Air Force Flight Test Center. Clark. Paper presented at the 6th Biennial Flight Test Conference. “Is Your Speed True. Birdsall. Kansas.P.C. 21. Harold. Karl D. Flight Dynamics of Rigid and Elastic Airplanes. 33. The Illustrated Guide to Aerodynamics. U.S.” AIAA 924090. 1992. 30. Austyn. Mair. (Out of print). California. Roskam. and David L. Wheeler. Kimberlin. Chapman & Hall. Princeton University Press. “Use of GPS for an Altitude Reference Source for Air Data Testing. 1992. Ralph and Joseph Sims. Edwards AFB. 32. MILSTD210A. Newport Beach. Roskam Aviation and Engineering Corporation. Richard. Walter. 1996. Albert. Defense Mapping Agency. Smith. “Takeoff Performance Data Using Onboard Instrumentation. Wing Theory. Second Edition. John and James Lackey. http://www. 1997. Dwenger. Pope. 26. Tab Books. Geodesy and Geophysics Department.uscg. August 2426. California. Jan. 19. Robert T. W. Collinson. 25. 1983. 29.mil/gps 35. Wood. 28. 24.navcen. 22. Etkin. 31. 246 . 27. R.
BIBLIOGRAPHY (Concluded)
36. NASA Allstar, www.allstar.fiu.edu/aero/research.htm 37. AIAA, www.aiaa.com 38. NASA Dryden, www.dfrc.nasa.gov 39. Denker, John S., See How It Fly’s, www.monmouth.com/~jsd/fly/how 40. Ojha, S.J., Flight Performance of Aircraft, AIAA Education Series, 1995. 41. Twaites, Bryan, ed. Incompressible Aerodynamics: An Account of the Steady Flow of Incompressible Fluid past Aerofoils, Wings and Other Bodies, Dover Publications. 42. Anderson, John D. 1998. A History of Aerodynamics. Cambridge University Press. 43. Chanute, Octave. 1897. Progress in Flying Machines. The American Engineer & Railroad Journal. 44. Lowry, John T., Performance of Light Aircraft, AIAA Education Series, 1999.
247
This page intentionally left blank.
248
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS
Abbreviation ADC AF AFB AFFTC AGL AIAA AOA AOSS air data computer acceleration factor Air Force Base Air Force Flight Test Center above ground level American Institute of Aeronautics and Astronautics angle of attack angle of sideslip acceleration acceleration factor aspect ratio aspect ratio of tanker acceleration in the down direction acceleration in the east direction acceleration in the north direction X axis body acceleration Yaxis body acceleration Zaxis body acceleration flight path longitudinal acceleration longitudinal acceleration flight path lateral acceleration lateral acceleration flight path normal acceleration normal acceleration (positive down) acceleration speed of sound Definition Unit ft deg deg ft/sec² dimensionless dimensionless ft/sec2 ft/sec2 ft/sec2 ft/sec2 ft/sec2 ft/sec2 ft/sec2 ft/sec² ft/sec2 ft/sec2 ft/sec2 ft/sec2 ft/sec2 kts
A AF AR ARt AD AE AN Abx Aby Abz Ax Ax Ay Ay Az Az a a
Note: 1. Velocity units in knots or feet per second. 2. Time in units of seconds or hours.
249
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS (Continued)
Abbreviation Definition temperature gradient mean (average) acceleration speed of sound standard day sea level angle of attack angle of attack from the aircraft system angle of attack computed from INS data body axis accelerometer British thermal unit brake horsepower wing span Celsius drag coefficient minimum drag coefficient lift coefficient break lift coefficient lift coefficient at the minimum drag coefficient tanker lift coefficient compressible skin friction drag coefficient incompressible skin friction drag coefficient center of gravity center of gravity centimeters differential GPS down drag drag of the aircraft body and wind Unit °K/1,000 ft ft/sec2 1116.45 ft/sec; 661.48 kts deg deg deg HP ft deg dimensionless dimensionless dimensionless dimensionless dimensionless dimensionless dimensionless pct MAC pct MAC lbs lbs
a a aSL
α α A/C α INS
BAA Btu
BHP
b
C
CD CD min CL CLb CL min CLt C fc C fi cg
cg cm DGPS
D D Dbw
250
LIST OF ABBREVIATIONS, ACRONYMS, AND SYMBOLS (Continued)
Abbreviation Definition standard day drag drag of the aircraft tail test day computed drag standard day predicted drag test day predicted drag distance change in true airspeed weight increment change in altitude time increment decibels degrees (either temperature or angle) east embedded GPS/INS east energy Fahrenheit flight level flight path accelerometer Fahrenheit summation parameter to be minimized propulsive drag excess thrust gross thrust net thrust referred net thrust Unit lbs lbs lbs lbs lbs ft lbs ft
sec
Ds Dt Dt Ds′ Dt′
d
dVt dWt
dh dt
dB deg E EGI
ftlbs deg (ft/100) deg lbs lbs lbs lbs lbs
E E
F FL FPA
F F* Fe Fex Fg Fn Fnr
251
AND SYMBOLS (Continued) Abbreviation Definition net thrust at zero speed corrected net thrust referred net thrust referred (inlet) net thrust standard day net thrust standard day predicted net thrust slope of thrust versus Mach test day net thrust test day predicted net thrust ram drag runway resistance force degradation factor for tsfcr nose gear load main gear load foot Global Positioning System acceleration of gravity reference acceleration due to gravity headup display mercury Hertz geopotential altitude rate of change of geopotential height pressure altitude energy altitude Unit lbs lbs lbs lbs lbs lbs lbs lbs lbs lbs lbs lbs lbs ft/sec2 32. ACRONYMS.LIST OF ABBREVIATIONS.17405 ft/sec² cycles per second ft ft/sec ft ft Fn 0 Fn / δ Fn / δ t 0 Fn / δ t 2 Fns ′ Fns Fnslope Fnt ′ Fnt Fr Frw Ftsfcr F1 F2 ft GPS g g0 HUD Hg Hz H ! H HC HE 252 .
ACRONYMS. AND SYMBOLS (Continued) Abbreviation Definition density altitude base geopotential altitude tapeline (or geometric) altitude rate of change of geometric height height above ground level height of wing above ground International Civil Aviation Organization inertial navigation system inches indicated horsepower point number thrust incidence angle iteration number kelvin thousand ft Kelvin kinetic energy parabolic coefficient of the drag polar nonlinear coefficient of the drag polar kilogram kilometers knot(s) lower heating value lift lift of the wing lift of the tail characteristic formula) length (in Reynolds number Unit ft ft ft ft/sec ft ft HP deg 1.LIST OF ABBREVIATIONS.000 ft deg K ftlbs dimensionless dimensionless Btu lbs lbs lbs ft Hd H0 h ! h hAGL hw ICAO INS In IHP i it j K K ft K KE K1 K2 kg km kt LHV L L1 L2 l 253 .
AND SYMBOLS (Continued) Abbreviation Definition Unit lx ly I yy lz MAC MAX METO MIL longitudinal (x) distance from cg lateral (y) distance from cg moment of inertia about the ybody axis normal (z) distance from cg mean aerodynamic chord maximum rated thrust maximum except for takeoff Military rated thrust Mach number moment mass meter millibar north not applicable National Advisory Committee for Aeronautics National Aeronautics and Space Administration Nose Boom Instrumentation Unit National Test Pilot School nondimensional nautical air miles nautical mile north number of points in multiple regression longitudinal load factor lateral load factor normal load factor (positive up) propeller efficiency ft ft ftlbs/sec ft dimensionless ftlb slugs g’s g’s g’s dimensionless M M m m mbar N N/A NACA NASA NBIU NTPS n/d nam nm N N Nx Ny Nz η 254 .LIST OF ABBREVIATIONS. ACRONYMS.
ACRONYMS.116.553 ft  q qC R RMS R R R R/C RF RN RNI r r0 S 255 .089.8136 ft²/sec² °K nam ft/sec nm dimensionless dimensionless deg/sec 20. AND SYMBOLS (Continued) Abbreviation Definition Unit η ηt P PE PSL Pa Ps Pt Pt′ p pph q temperature probe recovery factor inlet pressure recovery factor ambient (static) pressure potential energy ambient pressure sea level ambient pressure specific excess power total pressure total pressure behind a shock roll rate pounds per hour pitch rate incompressible dynamic pressure compressible dynamic pressure radius of a pullup root mean square radius of turn or pullup universal gas constant for air range rate of change of pressure altitude range factor Reynolds number Reynolds number index yaw rate reference radius of the earth south dimensionless dimensionless lbs/ft2 ftlbs 2.2166 lbs/ft² lbs/ft2 ft/sec lbs/ft2 lbs/ft² deg/sec deg/sec lbs/ft² lbs/ft² ft ft 3.LIST OF ABBREVIATIONS.855.
ACRONYMS.LIST OF ABBREVIATIONS.15 °K °K °K S SR SS δt0 δt2 δt2 S wet sec TPS T THP TSFC TSL Ta Tas Tt T0 t tsfc tsfcc tsfcr USAF °K ºK sec lb/hr/lb dimensionless lb/hr/lb kts  U cg VSTOL 256 . AND SYMBOLS (Continued) Abbreviation Definition Unit SFTE STOL Society of Flight Test Engineers short takeoff and landing reference wing area specific range sum of squares referred pressure ratio referred inlet pressure ratio total pressure ratio wetted area seconds Test Pilot School temperature thrust horsepower thrust specific fuel consumption sea level standard temperature ambient temperature ( T = interchangeable symbology) ambient temperature total temperature base temperature time thrust specific fuel consumption corrected thrust specific fuel consumption referred thrust specific fuel consumption United States Air Force Xbody axis true airspeed vertical or short takeoff and landing ft² nm/lbs dimensionless dimensionless dimensionless ft2 °K HP lb/hr/lb 288.
AND SYMBOLS (Continued) Abbreviation Definition Unit ! V VC VD VE VN Vbx Vby Vbz Vcg Ve Vg " Vg ∆Vt ! Vt Vt VtD rate of change of inertial velocity calibrated airspeed down (z) inertial speed east (y) inertial (ground) speed north (x) inertial speed longitudinal (xbody) axis airspeed lateral (ybody) axis airspeed vertical (zbody) axis airspeed Ybody axis true airspeed equivalent airspeed groundspeed (usually horizontal component of vector) groundspeed vector correction to be added to true airspeed rate of change of true airspeed true airspeed true airspeed down true airspeed east true airspeed north true airspeed vector indicated true airspeed vertical component of groundspeed vector windspeed windspeed vector (ft/sec)/sec kts kts kts kts kts kts kts kts kts kts kts kts ft/sec2 kts kts kts kts kts kts kts ft/sec kts VtE VtN " Vt Vti Vv Vw " Vw 257 .LIST OF ABBREVIATIONS. ACRONYMS.
LIST OF ABBREVIATIONS. ACRONYMS. AND SYMBOLS (Continued) Abbreviation Definition Unit VwD VwE VwN W down (z) windspeed east (y) windspeed north (x) windspeed west weight of an element of air zero fuel weight airflow Zbody axis true airspeed fuel flow kts kts kts lbs lbs lbs/sec ft/sec2 lbs/hr lbs/hr lbs/hr lbs/hr lbs/hr lbs lbs lbs lbs ft ft ft ft W WZf ! Wa Wcg Wf Wf / δ ⋅ θ ( ) corrected fuel flow standard day fuel flow standard day predicted fuel flow test day predicted fuel flow weight weight over pressure ratio end gross weight start gross weight with respect to independent variable distance from cg to wing center of lift distance from cg to tail center of lift distance main gear to thrust vector ground effect factor distance from nose gear to cg W fs ′ W fs ′ W ft Wt Wt / δ Wte Wts wrt X XL1 XL2 X Fn X GE X1 258 .
AND SYMBOLS (Continued) Abbreviation Definition Unit X2 x Y distance from main gear to cg the x unknown = Vwx dependent variable curve fit equation the y unknown = Vwy height of the body axis above ground height of the tail center of lift and drag above body axis the z unknown = ∆Vt ft kts kts ft ft kts ˆ Y y Z1 Z2 z Symbol σ σ β ∂ ambient density ratio standard deviation sideslip angle partial derivative symbol pitch attitude ambient temperature ratio thrust vector angle runway slope total temperature ratio ambient pressure ratio viscosity runway coefficient of friction coefficient of friction viscosity at sea level angular rate of a pullup flight path angle ratio of specific heats dimensionless deg deg dimensionless deg deg dimensionless dimensionless slugs/ft sec dimensionless dimensionless slugs/ft sec deg/sec deg dimensionless θ θ θV θ rw θt 2 δ µ µ µ µ SL ϖ γ γ 259 .LIST OF ABBREVIATIONS. ACRONYMS.
AND SYMBOLS (Concluded) Symbol γ0 φ º gravity at sea level (function of latitude) bank angle degrees engine losses factor heading angle (degrees from true north) increment integral latitude roll attitude summation theoretical downwash angle thrust increase time constant track angle cm/sec2 deg temperature or angle deg deg deg deg sec deg from true north λ ψ ∆ ∫ ϕ φ ∑ ε0 τ σg 260 .LIST OF ABBREVIATIONS. ACRONYMS.
164. 26 Descent. 168. 114. 15. 178. 113. 202. 47. 3. 121. 129. 26. 136. 30. 125. 25. 75. 94. 4. 67 Atmosphere. 164 A Accelerating or decelerating turns. 206. 97. 251 Gravity. 47. 15. 26. 116. 83. 146. 95. 1 Accelerometer accelerometer noise. 115. 191. 218. 57. 140 Equivalent airspeed. 113. 246. 114. 190. 174. Energy altitude. 100. 30. 2. 66. 144. 22. 158. 98. 116. 58. 12. 111. 136 L Landing. 42. 194. 102. 113. 245 E EGI. 154. braking forces. 181. 4. 254 C Calibrated airspeed. 112. 146. 113 Airspeed.INDEX 1976 U. 66. 106. 145. 28. 168. 246 Altitude Constant altitude. 124. 167. 131. 155 acceleration. 178. 236. 106. 97. 170. 73. 188. 109. 184 Dynamic performance. 2. 46. 45. 2. 103. 13. 60. 195. 165. 41. 101. 160 Excess thrust. 218. 250. 174 Lift. 52. 154. 135. 186. 60. 43. 181 Density. 111. 40. 173 Groundspeed. 26 Density altitude. 30.S. 96. 219. 160. 30. 254 261 . 107. 35. 13. 40. 36. 154. 142. 245 Latitude. 44. 136. 156. 82 Angle of attack. 57. 4. 43. 245. 3. 182 F Fuel flow. 200. 84. 197. 207 Drag coefficient. 203. 17. 172. 104. 246. 122. 41. 30. 108. 33. 160. 245 Cruise tests. 1. 108. 18. 82. 141. 189. 251 Instrumentation. 44. 134. 32. 114. 81. 104. 134. 72 Aerobraking. 191. 171. 113 Butterworth filter Fourpole Butterworth filter. 176. 187. 172. 80. 112. 192. 152. 61. 16. 201. 217. 45. 26. 250. 41. 206. 31. 37. 80. 134. 42. 58. 58. 160. 246. 177. 76. 3. 23. 129. 245. 251 Ambient pressure. 165. 144. 39. 150. 189. 55. 135. 166. 81. 149. 129 B Braking braking coefficient. 169. 106. 175. 32. 180 Drag due to lift. 13 Geopotential altitude. 140. 33 Drag. accelerometer rate corrections. 83. 81 M Mach number. 184. 188. 30. 82 D Deceleration. 136. 103. 145. 128. 3. potential energy. 126. 186. 2. 122. Standard Atmosphere. 121 Differential pressure. 23. 185. 40. 24. 202. 178. 185. 71. 83 Climb. 114. 181. 17. 63 I INS. 15 GPS. 151. 141. 180. 3. 179 Energy kinetic energy. 112. 116. 35. 87. 132. 156. 38. 37 Euler angles. 181 Differential GPS. 4. 190 Lift coefficient. 40. 140. 148. 120. 140. 3. 189. 192. 13. 209. 147. 108. 144. 5. 155. 111. 113. 182 G Geometric altitude. 184. 26. 152.
59 Noise. 102. 178 Turns. 199. 134 Ram drag. 172 T Takeoff. 203. 1 Total temperature. 3. 197. 51. 245. 101. 172 Standard atmosphere. 33 Pressure altitude. 205. 208. 25 Standardization. 211. 213. 41. 218 Thrust runs. 80. 239. 141. 219 Range factor. 88. 201. 202. 246 Thrust. 208 Minimum drag coefficient. 210. 78. 135. 30. 203. 155. 253. 216. 73 Refueling. 216. 92. 140. 136. 103 Reynolds number index. 102. 53. 204. 176 Reynolds number. 215. 50 Range. 193 Total pressure. 113. 54 Military thrust. 183. 21 Pressure ratio. 193. 188. 81 Thrust spe