AD-A248 270

m g** m ny%

ADVISORY GROUP FOR AEROSPACE RESEARCH & DEVELOPMENT

7 RUE ANCELIE 92200 NEUILLY SUR SEINE ' FRANCE

AGARS ADVISORY REPORT 370 Hutd Dy.uimk;* PumI Working Group 13

Air Intakes for ,

High Speed Vehicles

(Prises d’Air pour Vehicules a Grande Vitesse)

DTIC

tElECTF 0 .MAR 18 1992 a

I s I

D&tmBonoN stItemsht a

. - ■ - — 1 ■ ■■■■*■

Apfjo'rcd fci public

Dtetributtc’A Unlimited

This Advisory Report was prepared at the request of the Fluid Dynamics Panel of AC< ARD.

92-06897

NORTH ATLANTIC TREATY ORGANIZATION

92 3 1.7 019

* ote* * V

Pub'lshed September 1991 Distribution and Availability on Bach Cover

Best

Available

Copy

ADVISORY GROUP FOR AEROSPACE RESEARCH & DEVELOPMENT

7 RUE ANCELLE 922C0 NEUILLY SUR SEINE FRANCE

AGLARD ADVISORY REPORT 270 Fluid Dynamics Pans) Working Group 13

Air Intakes for High Speed Vehicles

(Prises d'Atr pour Vehicules a Grande Vitesse)

This \dvisory Report - prepared at the request of the Fluid Dynamics Panel of AGARD.

North Atlantic Treaty Organization Organisation du Traite de I'Atlantique Nord

The Mission of AGARD

According to its (Tenter, the mission of AGARD is to being together the leading personalties of the NATO nations in the fields of science and technology relating to aerospace for fix following purposes:

— Recommending effector- vs ess for the member nations to use their research and development capabilities for the common benefit of the NATO community-.

— Pro nding scientific and technical advice and assistance to the Military Committee in the field of aerospace research and development 'with particular regard to its military application);

— Continuously stimtihung advances in the aerospace sciences re erven t to strengthening the common oefcnce posture.

— Improving the co-opcration among member nations in aerospace research and development;

— Exchange of scientific and technical information;

— Providing assistance to member nations for the purpose of increasing their scientific and technical potential;

— Rendering scientific and technical assistance, as requested, to other NATO bodies and to member nations in connection ssith research and development problems in the aerospace field.

The mghest authority nothin AGARD is the National Delegatee Board consisting of officially appointed senior representatises from cacti member nation. The mission ot AGARD is carried out through the Panels which arc composed of experts appointed by the National Delegates, the Consultant and Exchange Programme and the Aerospace Applications Studies Programme. The results of AGARD wort «rc reported to the member nations anJ the NATO Authorities through the AGARD series of publications of which this is one.

participation in AGARD activiuc. is by invitation only and is normally limited to citizens oi the NATO nations.

The content of this publication has been reproduced directly from material supplied by AGARD or the authors.

Published oeptember 1991

Copyright C AGARD 1991 All Rights Reserved

ISBN 92-835-0637-5

Printed by Specialised Printing Services Limited 40 Chigwell Lane, Loughton, Essex 1G10JTZ

Recent Publications of the Fluid Dynamics Panel

AGARDOGRAPHS (AG)

Experimental Techniques In the Field of Low Density Aerodynamics

AGARD AG-3 1 8 (E), April 1991

Techniques Experimentales Liees a I’Aerodvnamique a Basse Density

AGARD AG-3 1 8 (FR), April 1 990

A Surrey of Measurements and Measuring Techniques in Rapidly Distorted Compressible Turbulent Boundary Layers AGARD AG-3 1 5. May 1 989

Reynolds Number Effects in Transonic Flows

AGARD AG-303, December 1988

Three Dimensional Grid Generation for Complex Configurations — Recent Progress

AGARD AG-309, March 1988

REPORTS (R)

Aircraft Dynamics at High Angles of Attack: Experiments and Modelling

AGARD R-776, Special Course Notes, March 1991

Inverse Methods in Airfoil Design for Aeronautical and Turbo machinery Applications

AGARD R-780, Special Course Notes, November 1990

Aerodynamics of Rotorcraft

AGARD R-781, Special Course Notes, November 1990

Three-Dimensional Supersonic/Hypersonic Flews Including Separation

AGARD R-764, Special Course Notes, January 1 990

Advances in Cryogenic Wind Tunnel Technology

AGARD R-774 Special Course Notes. November 1989

ADVISORY REPORTS (AR)

Appraisal of the Suitability of Turbulence Models in Flow Calculations

AGARD AR-29' , Technical Status Review. July 1991 '

Rotary-Balance Testing for Aircraft Dynamics .

AGARD AR-265. Report of WG 1 1 . December 1 990 \ •• ; . . }

Calculation of 3D Separated Turbulent Flows in Boundary Layer Limit

AGARD AR-255, Report of WG10, May 1990

Adaptive Wind Tunnel Walls: Technology and Applications

AGARD AR-269, Report ofWGl 2, April 1990

Drag Prediction and Analysis from Computational Fluid Dynamics: State of the Art

AGARD AR-256, Technical Status Review, June 1 989

CONFERENCE PROCEEDINGS (CP)

Vortex Flow Aerodynamics AGARD CP-494, July 1991

Missile Aerodynamics AGARD CP-493, October 1 990

Aerodynamics of Combat Aire mft Controls and of Ground Effects AGARD CP-465 .April 1990

Computational Methods for Aerodynamic Design (Inverse) and Optimization AGARD CP-463, March 1 990

Aoo63Sion For RTK iiF.A&I DT1C TAB □

Unannounced □

Ju-tlf '.cot ler -

By — - -

_Distritnit.*rn/ _ ^

Availability Codas Avail and/or Dlst I £.«>olel

r

Applications of Mesh Generation to Cooiplex 3-D Configurations AGARD CP-464, March 1 990

Fluid Dynamics of Three-Dimensional Turbulent Shear Flows and Transition

AGARD CP-438, April 1989

Validation of Computational Fluid Dynamics

AGARD CP-437, December 1988

Aerodynamic Data Accuracy’ and Quality; Requirements and Capabilities in Wind Tunnel Testing AGARD CP-429, July 1988

Aerodynamics of Hypersonic Lifting Vehicles

AGARD CP-428, November 1987

Aerodynamic and Related Hydrodynamic Studies Using Water Facilities AGARD CP-413, June 1987

Applications of Computational Fluid Dynamics in Aeronautics

AGARD CP-41 2, November 1 986

Store Airframe Aerodynamics

AGARD CP-389, August 1986

Unsteady Aerodynamics — Fundamentals and Applications to Aircraft Dynamics AGARD CP-386, November 1985

Aerodynamics and Acoustics of Propellers

AGARD CP-366, February 1985

Improvement of Aerodynamic Performance through Boundary Layer Control and High Lift Systems

AGARD CP-365, August 1984

Wind Tunnels and Testing Techniques

AGARD CP-348, February 1984

Aerodynamics of Vortical Type Flows in Three Dimensions AGARD CP-342, July 1983

Missile Aerodynamics

AGARD CP-336, February 1983

Prediction of Aerodynamic Loads on Rotorcraft

AGARD CP-334, September 1982

Wall Interference in Wind Tunnels

AGARD CP-335, September 1982

Fluid Dynamics of Jets with Applications to V/STOL AGARD CP -308, January 1982

Aerodynamics of Power Plant Installation

AGARD CP-30 1 , September 1 98 1

Computation of Viscous-Inviscid Interactions AGARD CP-29 1 , February 1981

Subsonic/Transonic Configuration Atrodynamics

AGARD CP-285, September 1980

Turbulent Boundary Layers Experiments, Theory and Modelling AGARD CP-271, January 1980

Aerodynamic Characteristics of Controls AGARD CP-262, September 1979

High Angle of Attack Aerodynamics

AGARD CP-247, January 1979

Dynamic Stability Parameters

AGARD CP-235, November 1978

Unsteady Aerodynamics

AGARD CP 227, February 1 978

Laminar-Turbulent Transition

AGARD CP-224, Octoiw 1977

IV

Preface

Future fighter concepts require air intakes with not only good performance characteristics over an even wider operating range, but also require inlet designs constrained by low signature requirements. For the engineers who have to deal with the problems of intake design, there exists the need to evaluate design tools and experimental capabilities for providing the innovative design concepts needed to meet the ever demanding challenges for engine inlets of advanced air vehicle configurations.

In recognition of tl 'he AGARD Fluid Dynamics Panel established Working Group 1 3 to report on the state-of-the-art in the field and compare different computational tools on the basis of available test cases. The test case results can be used for further comparisons and are meant as a first step to improve computational tools. A comparison has been made of testing techniques used in different wind tunnels for the measurement o? intake dynamic distortion using one common intake model. Design guidelines and rules have been reviewed and summarized.

The report presents the results of the Working Group study and its conclusions and recommendations.

Preface

Les specifications de conception des futurs avion* de combat exigent des prises d an qui soient ii la fois de faible signature el performantes sur une grande plage d'utilisation. Pour les ingenieurs d eludes qui xont confrontcs aux problemes de la conception des prises d'air, il y a lieu d’evaluer les outils dc conception et lev installations experimentales afm d'apprecier leur capacitc a fournir ies concepts d'etude innovateurs nccessaircs pour repondre aux specifications de plus en plus rigoureuses des prises d'air adaptees aux configurations avancees des vehicules aeriens.

En consequence, le Panel AGARD de la Dynamique des Fluides a erse !e groupe de travail No. 1 3 pour rendre comptc dc Petal de Part dans ce domains et pour comparer les different* outils de ealeui a partir des cas d cssai disponibies. Les resuhais des cas d'essais peuvent servir pour des comparaisons ulterieures et doivent etre consideres comme un premier pas pour ('amelioration des outils de calcul, L.es techniques d'essai employees dans differentes soufflerics pour la mesure de la distortion dynamique ii l'aide d'un scui modelc de prise d'air ont deja ete comparees. Des directives et des regies de conception onl etc examinees et resumees.

Ce rapport presente les resultats obtenus par le groupe de travail No. 13, ainsi que ses recommandations et ses conclusions.

Wolfgang Schmidt Chairm . V "• •

Richard G. Bradley Deputy Chairman, WG 1 3

Abstract

This report presents the results of a study by Working Group 13 of the AGARD Fluid Dynamics Panel which was formed to investigate the state-of-the-art of methodologies for aerodynamic design of engine intakes for high speed vehicles. The scope of the investigation included intake aerodynamics, intake/ engine compatibility, and intake/airframe integration for both aircraft and missiles.

The present capability of Computational Fluid Dynamics (CFD) methods was assessed through a comparative analysis of both CFD predictions and experimental data. This analysis was conducted for eight different flow field test cases designed to produce critical features of air-intake flow fields. Flow field results and comparisons are presented both in the report and in a microfiche appendix.

Air-inlet wind tunnel testing techniques and limitations were also inve itigated and reported. Results from measurements of inlet performance from three European wind tunnels using a common axisymmetric pitot intake are also presented.

The participants in Working Group 13 represented Belgium, France, Germany, Italy, the United Kingdom, and the United States.

VI

Contents

RECENT PUBLICATIONS OF THE FLUID DYNAMICS PANEL

PREFACE

ABSTRACT

Page

i i i

v

vi

1 OBJECTIVES AND SCOPE OF THE WORKING GROUP 13 1

2 INTAKE DESIGN AND PERFORMANCE 4

2.1 Introduction

5

Definition

of Intake Performance Parameters £

Description of Intake Flows

7

2.2.1

Internal Flow

7

2. 2. 1.1

Efficiency of Ram Compression

7

2. 2. 1.2

Flow Distortion at the Engine Face

8

2. 2. 1.2.1

Steady State Flow Distortion

8

2. 2. 1.2. 2

Dynamic Distortion

10

2. 2. 1.3

Flow Angularity

11

2. 2. 1.4

Flow Stability

11

2. 2. 1.4.1

Buzz

11

2. 2. 1.4. 2

Multi-intake Stability

12

2. 2. 1.5

Flow Quantity

12

2. 2. 1.6

Matching of Intake and Engine Airflow

14

2.2.2

External Flow

14

2. 2. 2.1

Pre-entry and Cowl Forces for a Pitot Intake

14

2. 2. 2. 2

Pre-entry and cowl forces for an intake with

a compression surface

16

2.2.2. 3

Spillage Drag

17

References

18

The Initial Design Process

19

2.3.1

Introduction

19

2.3.2

Generalized Inlet Model s Calculation Procedures

20

2.3.3

Specific Experience with the Level II Inlet

Installation Program (IIP)

20

2.3.4

Conclusions s Recommendations

22

References

22

Intake Design A Performance for Supersonic Cruise/Hypersonic

Operation

23

2.4.1

Finding a Mission for High Speed

23

2.4.2

Intakes for Mach Number 2 to 3+ Cruise

24

2. 4. 2.1

Characteristics of Intake Design

24

2. 4. 2. 2

Intake Technology in Current Mach 2-3+ Aircraft

24

2.4.3

Mach Number 4-6 Intakes

27

2. 4. 3.1

Requisite Technologies

27

2. 4. 3. 2

Specific Applications

28

2.4.4

lia'.h Number 6+ to 8 Air Intakes for First Stage Accelerators

29

2.4.5

Air Intakes for Scramjet Propulsion Mach Number 8

to 25+

30

nefe rences

31

Intake Design s Performance for Agile Tactical Fiqhters

33

2.5.1

Introduction

33

2.5.2

Isolated Intakes

34

2. 5. 2.1

Internal Flow

(a) Flow in the Subsonic Diffuser

34

(b) Combination of Intake and Subsonic Diffusers

36

2. 5. 2. 2

External Flow

38

2. 5. 2. 3 Intakes with Compression Surfaces at Subsonic and

Supersonic Speeds 39

2. 5. 2. 3.1 At Subsonic Speeds 39

2. 5. 2. 3. 2 At Supersonic Speeds 40

2.5.3 Intake-Airframe Integration 43

2. 5. 3.1 Fuselage Flow Fields for Side Mounted Inlets 44

2. 5. 3. 2 Performance of a Rectangular Compression Surface Intake on

the side of a Fuselage 45

2. 5. 3. 3 Performance of Axisymmetric Half Cone Intakes

on the Side of a Fuselago 46

2. 5. 3. 4 Performance of a Pitot Intake on the Side of a Fuselage 47

2. 5. 3. 5 Fuselage/Wing Flow Fields for Shielded Intake Installations 48

2. 5. 3. 6 Performance of Shielded Compression Surface Intakes

(a) Rectangular Intakes 49

(b) Half Axisymmetric Intakes 50

i

vii

Page

2.5. 3.7 Comparison of Performance of Shielded and Unshielded

Rectangular and Ha’f Axisymmetric Inlets 50

2.5. 3.8 Performance of Shi< led pitot Intakes 51

2.5.4 Technology Implementation in Current Aircraft 52

2.5.5 Conducing Remarks 56

References 56

2.6 VSTOL Aircraft Intakes, Design t Performance 57

2.6.1 Introduction 57

2.6.2 General Concepts, Specific Examples of Flight

Tested vehicles and Future Possibilities 57

2. 6. 2.1 Fixed, Direct Lift Engine Intakes 57

2. 6. 2. 2 Rotating Engine Intakes 57

2. 6. 2. 2.1 State-of-the-Art Experience 57

2. 6. 2. 2. 2 Future Possibilities 57

2. 6. 2. 3 Fixed Horizontal Engine with Flow Diverters 58

2. 6. 2. 3.1 State-of-the-Art Experience 58

2. 6. 2. 3. 2 Future Possibilities 58

2.6.3 Performance of VSTOL Intakes in Static and

Transient Conditions 59

2.6. 3.1 The Vertical Axis Inlet 59

2. 6. 3. 1.1 Inlet Physics 59

2.6. 3.1.2 The Lift Engines Intakes of the Do 31 VSTOL

Transport Aircraft 59

2. 6. 3. 2 The Fixed Horizontal Axis Intake - Harrier Type 6C

2. 6. 3. 2.1 Design Problems 60

2. 6. 3. 2. 2 Total pressure Recovery 61

2. 6. 3. 2. 3 Spillage Drag 61

2. 6. 3. 2. 4 Velocity Distribution 62

2. 6. 3. 2. 5 Low-Drag Cowl Section 62

2. 6. 3. 3 The Rotating Axis Intake - Vj-101 62

References 63

2.7 Intake Design and Performance for Missiles with Airbreathing Propulsion 65

2.7.1 Introduction 65

2. 7. 1.1 The relevance of Air Breathing Engines for

Missiles 65

2. 7. 1.2 Types of Air Breathing Propulsion 65

2. 7. 1.3 Specific Problems of Missile Intakes 66

2.7.2 Configuration Evolution and Constraints 66

2. 7. 2.1 Early Configurations 66

2. 7. 2. 2 Operational Constraints 67

2. 7. 2. 3 More recent Developments 67

2.7.3 Isolated Intakes 68

2. 7. 3.1 Intakes for Subsonic or Low Supersonic Speed

Missiles 68

2. 7. 3. 1.1 Pitot Intakes 68

2. 7. 3. 1.2 Flush Intakes 68

2. 7. 3. 2 Supersonic Intakes 68

2. 7. 3. 2.1 Axisymmetric and Derived intakes 69

2. 7. 3. 2. 2 Rectangular and Derived Intakes 70

2. 7. 3. 2. 3 Design Mach Number 71

2. 7. 3. 2. 4 Intake Size 71

2. 7. 3. 2. 5 Three Dimensional Intakes 72

2.7.4 Fuselage Flow Field 73

2. 7. 4.1 Fuselage with Circular Cross Sections 73

2. 7. 4. 1.1 Zero Incidence 73

2. 7. 4. 1.2 Non Zero Incidence 74

2.7.5 Missile Configurations 76

2. 7. 5.1 Electromagnetic Detection 75

2. 7. 5. 2 Types of Steering 75

2. 7. 5. 2.1 Skid to Turn Steeling 75

2. 7. 5. 2. 2 Bank to Turn Steering 76

2. 7. 5. 3 Possible Configurations with Circular

Fuselages <6

2.7.5. 3.1 One Intake 76

2.7.5. 3.2 Single Intake at Subsonic Speeds

2. 7. 5. 3. 3 Twin Intakes 78

2. 7. 5. 3. 4 Three Intakes 79

2. 7. 5. 3. 5 Four intakes 79

2. 7. 5. 3. 6 More than four intakes 83

2. 7. 5. 4 Fuselages with Wings or Strakes 63

2. 7. 5. 4.1 Wings 83

2. 7. 5. 4. 2 Strakes 84

2. 7. 5. 5 Fuselages with Non circular Cross Sections 85

2.7.6 Performance Prediction 86

2. 7. 6.1 Isolated Intakes 86

2. 7. 6. 2 Flow field around the Fuselage 86

viii

2. 7. 6. 3 Mounted intakes 87

2. 7. 6. 3.1 One Intake 87

2. 7. 6. 3. 2 Several Intakes 87

2.7.7 Air Breathing Missile Design 87

2.7.8 Conclusions 88

References 38

3 NUMERICAL SIMULATION OF AIR INTAKES 91

3.1 Introduction 91

3.2 CFD methods for inlets 92

3.2.1 Introduction 92

3.2.2 Grid Generation 92

3. 2. 2.1 Surface Grids 92

3. 2. 2. 2 Field Grids 93

3.2.3 Flow Solvers 94

3.2. 3.1 Flow Equations 94

3. 2. 3. 2 Solutions Algorithms 96

3. 2. 3. 3 Turbulence Models 96

3. 2. 3. 4 Current Status 97

3.2.4 Processing and Post-processing 98

3.2.5 References 99

3.3 Analysis of test cases 101

3.3.1 Test case 1 - Transonic norma shock/turbulent

boundary layer interactions 101

3. 3. 1.1 Introduction 101

3. 3. 1.2 Problem description 101

3. 3. 1.3 CFD techniques 101

3. 3. 1.4 Results 101

3. 3. 1.4.1 Test case 1.1 101

3. 3. 1.4. 2 Test case 1.2 102

3. 3. 1.4. 3 Test case 1.3 103

3. 3. 1.5 Conclusions 104

3. 3. 1.6 References 104

Appendix 3.3.1: Supplementary figures for test case 1 117

Microfiche 3.3.1: Contributions to test case 1

3.3.2 Test case 2 - Glancing shock/boundary layer interaction 128

3. 3. 2.1 Introduction 128

3. 3. 2. 2 Problem description 128

3. 3. 2. 3 CFD techniques 129

3. 3. 2. 4 Results 129

3. 3. 2. 5 Conclusions 132

3. 3. 2. 6 References 133

Appendix 3.3.2: Full comparison of CFD and Experiment 134

Microfiche 3.3.2: Contributions to test case 2

3.3.3 Test case 3 - Subsonic/transonic circular intake 139

3. 3. 3.1 Introduction 139

3. 3. 3. 2 Problem description 139

3. 3. 3. 3 CFD techniques 139

3. 3. 3. 4 Results 140

3. 3. 3. 4.1 Test case 3.1 140

3. 3. 3. 4. 2 Test case 3.2 142

3. 3. 3. 5 Conclusions 142

Appendix 3.3.3: Supplementary figures for test case 3 156

Microfiche 3. 3. 3/3. 3. 4: Contributions to test case 3 and 4

3.3.4 Test case 4 - Subsonic/transonic semi-circular intake 163

3. 3. 4.1 Introduction 163

3. 3. 4. 2 Problem description 163

3. 3. 4. 3 CFD techniques 163

3. 3. 4. 4 Results 163

3. 3. 4. 4.1 Test case 4.1 163

3. 3. 4. 4. 2 Test case 4.2 164

3. 3. 4. 5 Conclusions 164

Appendix 3.3.4: Supplementary figures for test case 4 167

Microfiche 3. 3. 3/3. 3. 4: Contributions to test case 3 and 4

3.3.5 Test case 5 - Supersonic circular pitot intake 171

3. 3. 5.1 Introduction 171

3. 3. 5. 2 Problem description 171

3. 3. 5. 3 CFD techniques 171

3. 3. 5. 4 Results 171

3. 3. 5. 4.1 Test case 5.1 172

3. 3. 5. 4. 2 Test case 5.2 172

Page

3. 3. 5. 4. 3

Test case 5.3

172

3. 3. 5. 4. 4

Intake pressure Recovery

173

3.3 5.5

Conclusions

173

Microfiche

3.3.5: Contributions to test case 5

3.3,

.6 Test case 6 - 2D hypersonic intake

183

3. 3. 6.1

Introduction

183

3. 3. 6. 2

Problem description

183

3. 3. 6. 3

CFD techniques

184

3. 3. 6. 4

Results

184

3. 3. 6. 5

Conclusions

189

3. 3. 6. 6

References

190

Appendix 3

.3.6: Full comparison of CFD and Experiment

192

Microfiche

3.3.6: Contributions to test case 6

3.3,

.7 Test case 7 - Axisymmetric nixed compression inlet

202

3. 3. 7.1

Introduction

202

3. 3. 7. 2

Problem description

202

3. 3. 7. 3

CFD techniques

202

3. 3.7. 4

Results

203

3. 3. 7. 5

Conclusions

205

3. 3. 7. 6

References

206

Appendix 3

.3.7: Full comparison of CFD and Experiment

207

Microfiche

3.3.7: Contributions to test case 7

3.3,

.8 Test case 8 - Intake/airframe integration

212

3. 3. 8.1

Introduction

212

3. 3. 8. 2

Problem description

212

3. 3. 8. 3

CFD techniques

212

3. 3. 8. 4

Results

212

3. 3. 8. 5

Conclusions/Recommendations

215

3. 3.8.6

References

215

Microfiche

3.3.8: Contribution to Test Case 8

216

3.4

CONCLUSIONS AND RECOMMENDATIONS

4

AIR INTAKE

TESTING

.717

4.1

Scope and

Purpose of Air Intake Tests

217

4.1.1

validation of Air Intake Test Runs without and

Engine

217

4.1.2

Test of the Air Intakes with Engine

217

4.1.3

Air Intake Interaction with the Aircraft

217

4.1.4

Similitude Parameters

218

4.2

Tests of Subsonic Transport Aircraft Intakes

218

4.2.1

General Aspects

218

4.2.2

Test of Air Intakes in High Subsonic Flow

218

4. 2. 2.1

Principle of the Test Setup

218

4. 2. 2. 2

Engine Face Equipment and Measurement of

Pressure Recovery P„

219

4. 2. 2. 3

Flow Rate Measurement

219

4. 2. 2. 4

Drag Measurement with External probes

220

4. 2. 2. 5

Other Drag Measurement Methods

220

4.2.3

Low /elocity Air Intake Tests

221

4. 2. 3.1

General Conditions

221

4. 2. 3. 2

Test Setups

221

4. 2. 3. 3

Engine Face Instrumentation and Measurement of

the Distortion

222

4. 2. 3. 4

Flow Rate Measurement

222

4. 2. 3. 5

Drag Measurement

222

4. 2. 3.6

Effect of the Aircraft Aerodynamic Field

222

4.2.4

Nacelle Installation Test Using Turbine Power

Simulator (T.P.S)

223

4.3

Supersonic

Air Intake Tests

224

4.3.1

General aspects

224

4.3.2

Study of the Internal Flow

224

4. 3. 2.1

Test Setups

224

4. 3. 2. 2

Engine Face Instrumentation

225

4. 3. 2. 3

Flow Rate Measurement

226

4.3.3

Drag, Lift and Moment Measurements

227

4.4

Transonic ,

and Subsonic Tests of Fighter Plane Air Intakes

228

x

i

Page

4.5 Special

Test Devices

229

4.5.1

Unsteady Flow Distortion Acquisition System

229

4. 5. 1.1

Complete Measurement

229

4. 5. 1.2

RMS Analysis

230

4.5.2

Intake Flow Dynamic Study

231

4.5.3

Complete Internal Flew Prohing

232

4.6 Measurements in Three European Wind Tunnels at Subsonic

and Supersonic Speeds o£ Dynamic Distortion and

Steady State Performance of an Axisymmetric Pitot Intake

232

4.6.1

Introduction

233

4.6.2

Model Geometries

233

4.6.3

Instrumentation

234

RAE

235

ONERA

235

MBB/DLR

235

4.6.4

Test Programmes and Test Conditions

235

RAE

236

ONERA

236

DLR

236

4.6.5

Data Reduction and Presentation

236

4.6.6

Comparison of Measurements

236

4. 6. 6.1

At Subsonic Speeds

236

(a) Effect of Incidence Variation

(b) Effect of Variation of Free Stream Mach

236

Number (a « 15° only) (Fig. 48)

238

(c) Effect of Reynolds number

238

4. 6. 6. 2

At Supersonic Speeds (a)Effect of Incidence Variation

239

(Fig. 50( a)-(d) )

239

4.6.7

Results Unique to ONERA and DLR Tests

240

4.6.8

Repeatability Study at RAE

242

4.7 CONCLUSIONS AND RECOMMENDATIONS

242

References

243

5 CONCLUDING

REMARKS

245

Microfiche Appendix

The Microfiche Appendix which accompanies ihis Publicat.on contains 'Contributions to Tost Cases' relating to Section 3.3 'Analysis of Test Cases' as shown below and in the main contents listing.

Sub-Section Page No. Fiche No.

3.3.1 Contributions to Test Case I A1 1

— Transonic normal shock/turboient

3.3.2 Contributions to Test Case 2 A33 1

— Glancing shock/boundary layer interaction

3.3.3/3.3.4 Contributions to Test Case 3 and 4 A 103 2

— Case 3 Subsonic/transonic circular intake — Case 4 Subsonic/transonic semi-circular intake

3.3.5

Contributions to Test Case 5 — Supersonic circular pitot intake

A184

3.3.6

Contributions to Test Case 6 — 2D hypersonic intake

A209

3.3.7

Contributions to Test Case 7 — Axisymmetric mixed compression inlet

A3 39

3.3.8

Contributions to Test Case 8 — Intake/airframe integration

\ .46

Note: Fiche No.5 contains tlk rnr reproductions referred to throughout Fiche No's I to 4.

2

3

4

4

xi

m

I

CHAPTER X

OBJECTIVES ADD SCOPE OP THE WORKING GROUP 13

Intakes foe ait breathing engines represent a major and very important component of high speed air vehicles. Intake efficiency contributes significantly to the performance and handling characteristics of modern aircraft. Intake design and integration exhibit signif rant interactions with the air vehicle configuration. Flow field structures are essentially very complex.

Over the last two decades there has been a continuous evolution and improvement of airf rame-in'eake integration and Intake design, mainly based on a wide set of wind tunnel tests. Problems which were detected in a number of cases only after prototype flying, i.e. damage of the intake structure during engine surges, distortion-induced intake/engine incompatibility etc. required modifications at a very late stage of an aircraft program. Problems that arose with highly integrated intake positions and complex supersonic intake solutions led to a comprehensive experimental intake/airframe integratior study in the USA (project Tailor Kate) in the late 1960's. An AGARD lecture series (LS53) was held in 1972 to review the state-of-the-art of airframe/engine integration at that time.

Since then computational methods as well as windtunnel testing techniques have impro- ed and deeper physical understanding has led to innovations such as the intake-airframe integration on the Rafale aircraft and to the intake design on the European Fighter Aircra ;t.

Ho'e recent achievements in Euler and Navier-Stokes methods along with new mesh generation capabilities are providing tools for detailed flow field analysis and intake optimization . E n unsteady phenomena such as buzz have been analysed analytically.

A large amount of experimental data has been collected at the various airframe- and engine companies and development centres in the USA and in Europe. Besides collecting data for the assessment of intake performance, increasing effort was placed on the subject of intake/engine compatibility. This effort has progressed from simple measurement of steady state total pressure distortion to the measuring of instantaneous distortion with digital, analog and hybrid data screening and the measurement of flow swirl at the engine face.

Efforts have also been made during the last few years with some success to reduce the amount of complex and expensive dynamic intake flow measurements by synthesizing dynamic distortion parameters based on a limited number of turbulence measurements. One problem however is that a universally applicable distorlion parameter is not available. More effort is required to achieve a better understanding of the complex Interaction of intake- and engine compressor flow.

Current goals in the intake design field include improved design concepts and rules, development and application of computational methods in the whole speed regime (subsonic to hypersonic), prediction of intake duct flow, understanding of the interaction of the intake- and engine compressor flow, development of techniques to reduce the complexity of intake distortion measurement and improvement of intake test techniques.

Future air vehicle concepts require air intakes with not only good perfo'mance characteristics over an even wider operat'ng range, but also require inlet designs with low observable characteristics. The thrust of this AGARD study is to evaluate existing dejign tools and experimental capabilities for providing the innovative design concents that meet the ever demanding challenges for engine inlets of advanced vehicle configurations.

The present Working Group 13 was formed to investigate the subjects of intake aerodynamics, intake/ engine compatibility and intake/airframe integration using results from bo":- experimental and computational techniques. Prediction of intake performance (pressure recovery, distortion and swirl) and care free handling of engines are of utmost importance to future military and civil aircraft projects.

After conducting a review of the state-of-the-art on intake design and performance for both aircraft and missiles (with air breathing engines) the group has compared critically results from the presently available methods for computing both external and internal intake flow fields. The experimental methods used to measure intake internal performance, drag and compatibility testing have also been critically compared and proposals for their improvement have been made.

Emphasis has been placed on the testing techniques for the assessment of intake flow distortion, the evaluation of relevant distortion parameters including swirl and the interaction of the intake flow with the engine compressor. Due to the high degree of this interaction, cooperation between the FDP and the PEP was considered essential and a group with PEP representation was formed.

\

2

The Working Group was composed of the following members. PDP PANEL MEMBERS

Dr. W. Schmidt (Chairman) Messerschmitt-Bblkow-Blohm GmbH, UP Postfach 80 11 60 D-8000 Miinchen 80 - Germany

Dr. R.c. Bradley (Deputy Chairman) General Dynamics GD/FW-MZ 2888 P. 0. Box 748

Port worth, tx 76101 - USA Nr. J. Leynaert

Direction Grands Noyen d'Essais

ONERA

B.P. 72

92322 Chatillon - France

NON-PANEL MEMBERS

Mr. T.J. Benson

NASA Lewis Research Center

Internal Fluid Mechanics Division

Mail Stop 5-7

Cleveland, OH 44135 USA

Dr. N.C. Bissinger Messerschmitt-BSlkow-Blohm, UF Postfach 80 11 60 D-8000 Nilnchen 80 - Germany

Mr. E. Farinazzo

internal Ae. -'ynamics - Propulsion AERITALIA-n:..", 10 ,'SLIVOLI DA COMBATTINENTO C. MAKCHE TURIN - ITALiT

Mr. E.t. Goldsmith ARA

Manton Lane Bedford MK41 7PF - GB

PEP PANEL MEMBER

Prof. Ch. Hirsch vrrje Universiteit Brussel Dienst Stromingsmechanica Plainlaan 2

1050 Brussel - Belgique

Mr. G. Laruelie

ONERA

B.P. 72

92322 Ciatillon - France

Dr. P.A. Mackrodt

DLR HA-KK

Bunsenstr. 10

3400 Gdttingen - Germany

Mr. L.E. Surber

Airframe Propulsion Integration Flight Dynamics Laboratory AFWAL/FIMM

Wright-Patterson AFB , OH 45433 - USA

Mr. D. Welte Denier Luftfahrt GmbH Postfach 1420

7990 Friedrichshafen - Germany

The term of the Working Group was three years, Sept. 1987 until Sept. 1990. Much of the preparatory effort was however organized and carried out between the formal meetings by the following three subcommittees for consideration and approval by the Working Group at large:

Committee A (Chairman: E.L. Goldsmith) Intake Design and Performance

Committee B (Chairman; N.C. Bissinger) Numerical Simulation of Air Intakes

Committee c (Chairman: J. Leynaert)

Air Intakes Testing Methods

t

i

The chapters were written by the authors noted in parenthesis and reflect the consensus of the Working Group:

CHAPTER 2 - INTAKE DESIGN AND Pt-K FORMING S

(E.L. Goldsaith, L.E- Surber, D. Kelte, G. Latuellei

CHAPTER 3 - NUMERICAL SIMULATION OT INTAKES

(T.J. Benson, N.C. Bissinger, S.G. Bradley)

CHAPTER t. - AIR INTAKE TESTING

(J. Leynaert, E.L. Goldssith)

The Working Group is grateful to all individual contributors and their organizations for contributing and supporting the Working Group and this report.

Finally we would like to thank NASA Langley, RAE Bedford. ONERA Modane and DLR Gottingen for hosting Working Group neetings and presenting their test facilities.

CHAMI1 2

r-

cross sectional area

a

angle of incidence

CD

drag coefficient [ ^ |

Ar.f'

3

* * sides 11 p/yaw

f N 1

7

ratio of specific heats

normal for** coefficient 1 - _ ____ |

1 <*o VeW

*n

compression surface angles

CT

thrust coefficient f- ^ |

1

boundary layer thickness

^ >Jo Ar.r

«

mass flow ratio

Ck

intake lip contraction ratio

*i

cowl internal & external angles

0

drag

y.4

diameter

1

compression efficiency

total pressure distortion

V

intake total pressure loss factor

H.

.'eight

h

he'ght of diverter

Su££ttii

**-oO spec’. fic distortion lnd»it*a

t>

bleed

K

proport ion of maximum allowable engine

BL

boundary layer

—

Kl IUi 1

face distort Ion

c.cap.

capture plane

L. f

length

D

design or SOL

i.

lift

ex

exit piara

Mach number

F

centre of pressure

P

pressure

f . 2

engine face

PR. PR

pressure recovery (- P. /P. )

■* o

1

intake entry plane

i.r

» ad i us

j

jet

£>

g-s constant

l

loca 1

R?

Reynolds number

N

OX It tiOZ7.lt*

S

surface area

n

behind normal shock wave

T

t cmpcrature

o

free stream

t hrust

0

at datum or reference condition

T

velocity

s

surface

V

IV

% i dth

SOL

shock-on- l . p

W.fn

mats flow

Sff

» i dewa 1 1

spill

at spillage

x ,Xi

y.v

i.'/l

axes

th, 1

throat plane

& associated lengths

w

behind oblique shock wave

*

at sonic condi t Ions

5

:.i MMsaaim

Th# iBportanr* of th* intake and the exhaust nozzle on the perform nee of the total propulsion unit for any type of aircraft has been e^ihasised on several occasions. Examples of this

l^ortance for too vastly different aircraft operating at either end of the speed range MqO-2.0 are shovn In Figs. 1.1 and 1.2. The first illustrates the distribution of thrust forces throughout the nacelle of the Concorde aircraft and the second shoes the laportance of high intake efficiency at static conditions for the Harrier aircraft when taking off vertically or hovering.

DISTRIBUTION OF THRUST FORCES ON CONCORDE NACELLE AT M.*?

9% *% r%

0 im emu ■ wm mju mu ■

FIG 1.1 SOME ASPECTS OF THE IMPORTANCE OF THE POWERPLANT NACELLE ON CONCORDE CRUISE FLIGHT EFFICIENCY

* mux ■ vto imx maam acorn . *» maun a mi Miar

FIG 1.2 SENSITIVITY OF HARRIER INTAKE EFFICIENCY UNDER STATIC CONDITIONS

Perhaps even more important than performance aspects are thoae of intake and engine compatibility, if a wrong choice is made in the geometry of the Intake (e.g. choice of lip bluntness, auxiliary intake area, or perhaps, subsonic diffuser shape) then the aircraft's engine may surge even before taxiing to the end of the runway for take off. For a ramjet powered missile, for similar reasons of poor flow distribution at the entry to the engine, ramjet operation may never occur and flight will ba terminatad at burn-out vf the rocket boost motor.

It Is appropriate et this time to review experience of intake design and performance gathered since the lest major AGARD publications that attempted anything similar - Airframe end Engine Integration (ACARD LS53 May 1972) and Inlets and nozzles for Aerospace Engines (AGARD CP91 Sept. 1971), Since that time soma aspects of intake operation end performance heve been included tn:-

Aerodynamlc Interference (ACARD CP 71-1 Sopi . I97G)

Airplane/Propulsion Interference (ACARD CP-150 Sept. 1974)

Distortion Induced Engine Instability (ACARD LS 72 Nov. 1970

Advanced Control Syst.es for Aircraft pswerplants (ACARD CP. 274 Oct. 1979)

Aerodynanlca of power Plant Installation (ACARD CP 301 May 19S!)

Improvement of aerodynamic performance through Boundary Layer Control and High lift Systems (ACARD CP 363 May 19*4).

Ramjets and Rocket Propula Ion Systems for Missiles (ACARD LS136 Sept. 19*4)

Engine Response to Distorted Inflow Conditions (ACARD LSI36 Sept . 1986)

Special course in Missile At-odynamics (ACARD R7S4 May 1987)

in addition over this time period one specialist lecture series on intakes has been held at VKI Brussels (Intake Aerodynamic. Eeb. 1988).

After three general Sections this review divides broadly into Sections based on classes of vehicle i.e. long range (supersonic) cruise aircraft, tsctlcal fighter aircraft. VSTOL aircraft and misailea with airbreathing propulsion. The most obvious omissions ore subsonic cruise transport aircraft and helicopters. The former haa been omitted because nearly all the propulsion aerodynamics Interest is ce* ' red on the problems of external InterUrence between

nacelle/pylon/wing/body and is therefore almost wholly e complete airframe subject. The latter primarily because of the limited range of the subject and the scarcity of available data.

Again because of the quantity of data available (which reflects the scale of the activity over the last fifteen or so years) the most substantial sections are numbers 5 6 7. It was decided not to re-review the Intense activity devoted to the Intakes of the B.70, Concorde and the Boeing SST and other SST designs of the 1960's. Thus Section 4 deals with the caiher more general aspects of design studies for a future SST. that have appeared more recently and whatever data is available for the SR-71 aircraft. However Interest In this class of vehicle together with the associated area of air breathing propulsion for hypersonic vehicles is beginning to grow rapidly at this time.

In this area wt return to the theme of Importance of the propulsion unit. For the high supersonic and hypersonic speed range the whole layout of the airframe Is dominated by the geouetry of the Intake and the exhaust nozzle as evidenced clearly In Fig. 1.3.

FIG 1.3 AIRFRAME INTEGRATED SUPERSONIC COMBUSTION RAMJETS

;

6

Fig 1.4 illustrates in nor* detail the possible geometries that emerge for a scramjet as a r«3ult of integration of the intake and the supersonic combustor combined with need wherever possible to keep all leading edges swept.

ovcS-uHDsa Tuascjrr/RAMjcT

MACH S

THE MTU HYPERCRISP COMBINATION TURBOFAN & RAMJET

FIG 1.4 SCRAMJET INTAKES

RAM BURNER

FIG 1.5 POSSIBLE COMBINATION POWERPLANTS

Thus whatever form the powerplant takes - and the choice is wide - ramjet, turbo-ramjet, lurborocket , ramrocket , variable cycle turbojet, scramjet and combinations of these - the intake will always be a vital element of the propulsion unit. Research and development of Intakes must therefore proceed, as In the past, at all speeds from zero to hypersonic velocities.

However for many applications flight from Mach number zero to hypersonic speeds is necessary and dual mode powerplants (usually a combination of turbojet and ramjet) are needed. Fig. 1.5 shows some possible combination powerplants that show the size and particularly the inevitable complication of variable geometry intakes and ducts.

; 'WT'jWcontwtor ;s- •" - *

T , . ^

i&M

7

OJL . Pf-flWlTIQW Of INTAKE PERFORMANCE 7 A^A.-4r~r £R£ & DESCRIPTION Of INTAKE FLOSS

CONTENTS

2.2.1

INTERNAL FLOW

2.2.1 .1

Efficiency of ram compression

2. 2. 1.2

Flow distortion at the engine face

2. 2. 1.2.1

Steady state flow distortion

2. 2. 1.2. 2

Dynamic distortion

2.2. 1.3

Flow angularity

2.2.1 .4

Flow stability

2.2.1 .4.1

Buzz

2. 2. 1.4. 2

Muit!~intakc stability

2.2.1 .5

Flow quant 1 ty

2.2.1 .6

Matching of intake and engine airfli

2.2.2

EXTERNAL FLOW

<<.2.2.1

Pre-en%ry and cowl forces for a plti

intake

22.2.2

Pre-entry and cowl forces for an

Intake

with a compression surface

2. 2.2.3

Spi 1 lage drag

REFERENCES

C cpti.-.’d

strtmtub* _ Extrrnal f I c -

FIG. 2.1 NACELLE FLOW STATIONS

This equation can be made independent of the engine face Mach number M2 by assuming that further compression to zero velocity is achieved {sent ropical ly so that - 0 and P2 - Pt and

which for incompressible flow reduces to:-

^oj “ Vrmm ” Ptj. “ P0

*0

This quantity has been used extensively to define performance of subsonic intakes at low forward speeds. However the definition cannot be used at zero forward speed.

JL2 _ BEElNlIlQh QE INTAKE PERFORMANCE PARAMETERS

UJ _ ifllEBMAL FLOW

There are six properties of the internal flow that are usually measured by the intake aerodynami cist and are of concern to the engine manufacturer whose engine his to operate in the flow delivered by the combination of intake and following duct. These six quantities are efficiency of ram compression, static and dynamic distortion and flow angularity, stability and quant ity.

? — Efficiency of Ram Compression

The most natural definition of efficiency of the ram compression process is:-

Vff “ Work done in compression Kinetic energy available

which for compressible flow is:-

7-1

where station *0' is in the free stream and station ‘2* is at the engine face (Fig. 2.1).

In this condition, it is usual to use a loss coefficient defined as:-

<12

At high free stream speeds and particularly for supersonic flow a more convenient measure of efficiency than is the simple ratio of mean total pressure at the engine face to free stream total pressure Pt2/Pto wh*<* Is widely known

as pressure recovery. This is sometimes designated as »j or ijp but will be presented in this report as the above total pressure ratio or as PR.

It is impractical (and many times impossible) to take measurements at the compressor face when the engine is installed and operating. Consequently, the engine and Intake designers agree upon an Aerodynamic Interface Plane (AIP)(which Is forward of the compressor face but sufficiently close to the compressor face to * similar flow field) for the definition of Pt .

2

The various merits of weighting Individual pitot pressure measurements made at this position to form a moan value have been discussed exhaustively in ref. 2.1. Area mean, mass flow mean, mass derived and constant momentum derived, and entropy flux mean are all options. Although pressure recovery is usually measured by means of a pitot rake, it can also be deduced from measurements of mass fiow and static pressure. Air intakes for missile engines sometimes have to be operated in conditions when large areas cf separation can occur at the engine entry. Fig. 2. 2 shows some curves of pressure recovery

8

for such a condition calculated by conserving: -

1. mass flow - entropy

2. mass flow - momentum

3. static pressure - mo tntura

4. static pressure - mass flow

and confirming the pressure recovery so calculated with a mean obtained from a pitot rake. 1#

Total pressure distortion can also appear in an otherwise distortion free geometry at design conditions due to aircraft attitude which leads to lip separation (Fig. 2. 3d) at incidence or yaw (again related to lip thickness) or to mismatching between engine and intake airflows which results in subcritical (Fig 2.3e) or supercritical (Fig.2.3f) operation.

FLOW SHOCK PATTERN FOR 3 SHOCK INTAKE

FIG 23 (e)-(f ) SOURCES OF FLOW SEPARATION

• CALCULATED AVERAGE f.Ot\ CONSERVING :

- STATIC ARESSURE- MOMENTUM • STATIC PRESSURE - MASSf-LOW

. TOTAL PRESSURE BALANCE

FIG 2.2 PRESSURE RECOVERY DEFINITION

If the flow does not contain large areas of separation, and this Is usually the case for acceptable performance from turbojet and turbofan engines then all the methods of weighting give values which do not greatly differ from each other. Area weighting is the simplest and It is the one mcst commonly used.

Sources of total pressure distortion can be inherent in both the geometric and aerodynamic design - intake shock and boundary layer Interaction on an adjacent aircraft surface or on an intake compression surface can result In separation behind the terminal shock (Fig. 2. 3a); separation In a duct can result because of choice of a too high rate of diffusion and/or the presence of sharp bends (Fig. 2. 3b); the absence of auxiliary intakes combined with thin intake lips can cause separation at take off (static) conditions (Fig. 2. 3c).

(o)

COMPRESSION SURFACE SEPARATION

Flow velocity. Mach number and Reynolds number are required at the compressor entry to determine relative angle of incidence. Mach number and Reynolds number of the flow on to the compressor blades. If both static temperature and pressure can be assumed constant and steady across the compressor face then both velocity and Mach number can be considered as a function only of total pressure and the distribution of this quantity is the only measurement that needs to be made .

In order to obtain this type of flow distortion Information, total pressure measurements are taken at the A.I.P. Total pressure probes arc mounted In a series of equally spaced radial rakes such that they form a series of concentric rings (Fig. 2. 4).

FIG 2.4 PROBE ORIENTATION - VIEW LOOKING FORWARD

Increased accuracy and convenience of data reduction can be obtained If the radius of each ring is set such that all probes are at the centroids of equal compressor face areas. This means that probe radial spacing decreases from the innermost ring to the outermost ring.

Probably the most widespread quantitative distortion parameter available in the literature because of Us use in the earliest measurements on intakes in the 1950s Is the simpler-

LIP SEPARATION

AT ZERO INCIDENCE,

HUH FLOW

FIG 2.3{a)-(d) SOURCES OF FLOW SEPARATION

V Ap, -

and this Is always a useful quantity to measure for comparison purposes and to describe the general 'health' of an Intake flow Irrespective of the type of powerplant (turbojet, turbofen, liquid or solid fuelled ramjet or remrocket) that may be used,

9

✓

More advanced methods introduced in the late 1960a and 1970s take Into account the APt/Pt distortion for each ring of total pressure measurements (with APt now being between ring average and minimum total pressures). These rin^ distortions are weighted by circumferential extent factors 9-, (the sector angle of the ring for which recovery is below face average). Other Improvements Include factors to quantify the relative Influence of the circumferential ana radial distortions.

The effect of a circumferential distortion on compressor sur^e margin is displayed In Fig. 2. 5. Here, the* 180 profile essentially drops the maximum pressure ratio of a constant corrected speed line.

FIG 2.7 EFFECT OF EXTENT ANGLE

CORRECTED

am a

O *7.1 □ 90-1 0 93.2 0 96.2 b. 100 2

SCREEN NJ. I

ciowo *TH*eii

STALL POIHTS

FIG 2.5

Effect of Circumferential Distortion

One of the simplest quoted indices of distortion from Rolls Royce (Fig. 2. 8)

Overall mean total pressure *

Mean total pressure in sector = Pf

FIG 2.8 DEFINITION Of DC*

which rela,''s specifically to engine compressor surge margins I s : -

The effect of a hub radial dlctoition on compessor performance is displayed in Fig. 2.6. In this case the constant corrected speed lines move to the left. That is, a given rotor speed pumps less air flow so that surge is encountered again at « lower pressure ratio.

CORRECTED

FIG 2.6 Effect of Radial Distortion

q2

where Is the mean total pressure In the ’worst’ section of extent 0 and

Ptoand q2 are the mean total and dynamic pressures respectively at the

agreed engine interface plane. Significant 0 values can vary with tne engine design and commonly are 60 , 90* and 120*. For a bypass engine where GC indicates that the index is

taken over the area of the gas generator can be more significant than taking the whole engine face in relating to surge margin.

American engine manufacturers use more complex descriptors of the flow which hove been evolved to take account of both radial and circumferential distortion. One relatively simple radially weighted circumferential index i s : -

Critical circumferential distortion extent angles (9-) are determined in current methods since it has been demonstrated that the blades do not have time to react to narrow dips In the circumferential pressure profile. Above this critical angle, extent factors are calculated to determine the effect of the low pressure region on surge margin loss; however, Fig. 2. 7 illus¬ trates the fact that the bulk of surge margin loss occurs with a 60* to 90* extent angle.

or *D2-

fdPt . 6

r2

—

liptn

max

rn

l r2/r"

n-1

.h.r. APtj- <Pt - PtB|n)n «nd Pt, Pto)n .r.

respectively the mean and lowest pressure in a ring (n is often 5 or 6) and 9 Is an extent factor.

I

10

A multiple per revolution factor may also be used to determine the effect of more than cne low pressure region in a circumferential distribut ion.

Other distortion indices and comparisons between them are given in refs. 2.2 and 2.3.

Using this type of parameter, the loss in surge margin can be calculated directly with the distortion value based on both the intake flow field and the engine design (Fig. 2. 9).

FIG 2.9 CORRELATION OF COMPLEX DISTORTIONS/ GENERAL FORMULATION

Correlations between predicted and calculated loss in surge margin for each type of distortion are displayed in Fig. 2. 10 for a turbo fan compressor and a F-100 fan. These plots show good prediction accuracy of these parameters for a wide variety of flow conditions and engine components. An error band of plus and minus two percent is indicated and the majority of the data falls well inside the two percent band (Ref. 2. 4).

APRS CORRELATION OF TURBOFAN COMPRESSOR

FIG 2.10 PREDICTION /MEASUREMENT CORRELATION- FANS, COMPRESSORS, CORES

2J,-L2,2 — Dynamic Distortion After initial suspicions in the early 1960's that time variant distortion could be responsible for compressor surge it was shown In the late 1960's that although compressor blades need a minimum extent of circumferential distortion to react to, compressor blades do also react If exposed to this critical extent for a sufficient time. Thus surge would follow if the critical steady state distortion index was exceeded fer a time period of the order of that for one engine revolution, typically about 3 milliseconds.

For example, during prototype flight tests of the F-lllA, it became apparent that the desired flight envelope was restricted. Maneuvers of the aircraft at high subsonic and supersonic speeds resulted in engine compressor stalls at steady-state distortion levels which engine tests (using upstreams screens to create these distortions) had shown to be acceptable. This inlet -engine incompatibility gave rise to comprehensive flight test and wind tunnel investigations to identify and correct the causes of the unexplained compressor stalls.

Flight tests of the F-lllA were used to determine how the dynamic nature of intake pressure fluctuations related to engine operational stability. Steady state values of the flow distortion were compared with values taken at 400 sps (samples per second). A typical comparison is shown in Fig. 2. 11 for a Mach 1.6 case. This data clearly shows that while the low response data would not indicate the potential for stall, the high response parameter measurements yield a substantial peak approximately 15 milliseconds prior to surge.

Hitt. IK

FIG 2.11 High and Low Response Comparison for F-IIU Flight Test; Mach 1.6

For a preliminary assessment of dynamic distortion in a given intake, a rough rule of thumb can be applied, based on turbulence of the flow expressed as values of root -mean-square of the fluctuations in static pressure. If (At>)rms/Pt is not greater than one per cent, the problem of dynamic distortion can confidently be excluded: if the same factor is as high as four or five per cent, then detailed distortion measurements are advisable.

The determination of dynamic distortion, requires, much additional instrumentation ond both experiment and analysis take on a different order of complexity.

Unsteady pressures have to be measured at a large number of points at the engine-face position, for representative free-stream conditions In a wind tunnel. The number of points necessary in a development test hay been variously recommended as between 36 and 60. These pressures are recorded on miniaiuie high-response di fferent ia! -pressure transducers .

Details of the means of measuring and collecting the very large amount data required and the aubaequent editing and analysts to produce values of dynamic distortion are given in Chapter 4.

11

AnguUrlty of th« duct flow will elthor lncrewwo or decrows* Incidence on to the compreseor bledee end 1> therefore obviously »n Inportent measurement for compatibility of Intake engine airflows. If an engine Is equipped with Inlet guide vanes that straighten the flow before the first conpressor stage then this Importance may be somewhat diminished.

Flow angularity or swirl develops after a duct bend and Is the result of an Interaction between the centrifugal pressure gradient and a low energy region such as a boundary layer or a reg'on of separation. The centrifugal pressure gradient Is proportional to aJM where V Is

D

mainstream velocity and R Is the bend radius and results In pressure at the outside of the bend being greater than on the inside. If there is a boundary layer on the bottom of the duct with reduced velocity V' the local centrifugal gradient pv' * is insufficient to balance the R

pressure difference between the walls so that the flow in this region is directed towards the inside all. If there is a similar boundary

layer on the top wall, flow is also directed inwards in this region. Both top and bottom inward flows return to the outsido wall in the region of the centre of the duct and the result is two cells of swirling flow (Fig. 2. 12a).

If a large low energy region such as a separation occurs on either the top or bottom wall of the duct upstream of the bend then a single directed or bulk swirl results as shown typically in Fig. 2.32b. The resulting swirl measured at the engine face is a combination of the bulk swirl and twin swirl patterns (Fig. 2. 13).

ClRCUHttRCNTlAl FLOW ANGU -MI R-COWl

FIG 2.13 SUPERIMPOSING OF BULK i TWIN SWIRl.S

Mi,1, .4 _ flow Stability

One form of unstable Flow, compressor surge has already been referred to, but this la a phenomenon resulting from engine malfunction. However when two intakes are closely coupled, the possibility exists that a surge of one engine causes the surge of the second engine. This Is due to disturbance# that the hammershock overpronsure originated by the surge of the first engine *uces on the adjacent intake, and occurs primarily at supersonic speeds, when the adjaoent intake shook system ia effected.

If it is not possible to completely separate the intakes, the problem can be avoided or attenuated by means of a suitable splitter plate dividing the Intakes, and/or cutback endplates to the supersonic compression surfaces which will attenuate the overpressure. There are in addition two forms of flow instability emanating solely from the intake. The first is confined to supersonic speeds and is caused by oscillation of the intake shock system (known colloquially as 'buzz’). The second occurs when two or more intakes supply flow to a single duct and is usually known as twin (or multi) Intake Instabi 1 ity.

BUZZ

External and mixed compression Intakes operating subcritical ly are subject to an unsteady flow oscillation called "buzz". Unless the supersonic freest ream Mach number is very low cli supersonic intakes with external compression surfaces appear to exhibit this instability of the flow at some point in the subcritical flow regime. The Intake flow states of supercritical, critical and subcritical and the probable occurrence and form of shock oscillations in these flow states are described in ref. 2.5 and shown In Fig. 2. 14.

A

Shock acillation f -

liaifi \ -

LARGE AWXITUOE •'SOLUTION

FIG 2.14 SHOCK OSCILLATION & THE INTAKE PRESSURE RECOVERY vi FLOW CHARACTERISTIC

The phenomenon normally occurs at mass flow ratios below design and serves to limit the operating range of the intake-engine combination. Buzz can be responsible for structural damage to the intake, compressor surge or ramjet flame out. Buzz begins when the Intake becomes choked because of massive flow separation. The cause of the separation may be associated with shock wave boundary layer interaction, diffuser flow separation, or shock wave interference ahead of the inlet. Fig. 2. 15 illustrates two suggested mechanisms of buzz triggering resulting from massive separation In the diffuser. In any event* the normal shock (s pushed far out on the compression surfaces to spill the unpassed flow. The flow situation which caused the separation is altered dramatically, reestablishing attached flow with a greatly reduced static pressure created by th* starving engine. The normal shock is consequently sucked Into the subsonic diffuser. As the system stabilizes, the flow structure which ceused the initial separation reappeara auch that the cycle repeats itself.

12

FIG 2.15 Buzz Triggering Mechanisms

Fig. 2. 16 shows the shock cycle and corresponding diffuser pressure. Buzz is characterized by low frequency and high amplitude. Sustained buzz can result in an engine coirpressor stall.

The condition called "hammer-shock" al-eady mentioned may also result from sustained buzz md/or compressor stall. Transient overpressures generated by a sudden engine surge can reach values well above theoretical maximum steady-state Pt . Such a transient fractured the North American F-107 intake ramps during flight test .

Normal shock oscillations can also occur at high mass flew ratios and supercrtical operation (Fig. 2.14). In this condition the normal shock may oscillate in the subsonic diffuser du* to the basic Instability of the strong shock wave boundary layer interaction. These oscillations are of higher frequency and lower amplitude than buzz, but can still generate high turbulence and distortion av the compressor face, possibly surging the engine.

What usually matters to the Intake designer and the engine manufacturer is the definition of the point on the subcritical characteristic of ,'.n intake at which 'buzz* Is Initiated l.e. the stable flow ranje. The match point between engine and intake flows can then always be made to fall within this range Stable flow range or margin is usually characterised either

as

IM

or as

IM .

f A°]

LacJ mi n

stable

ACJ max

J^Ao

lAcJmax

The possible ’triggering' mechanisms thtt could cause the onset of 'buzz* have already been mentioned and are discussed In moie detail In ref. 25. .ogether with measures that can be taken in intake design to maximise the stable subcritical flow region.

1UJJ Multi - Intake Stapllltv

For aircraft this phenomenon Is usually associated with twin Intakes situated on either side of a fuselage feeding air to a single engine. When intake flow is reduced by control of the exit of the common duct a critical point Is reached below which unequal flows develop in the two Intakes. On one side the flow increases again while on the other it falls rapidly to zero and can become negative. The net result on the flow reachihg the engine is that total pressure recovery falls suddenly and total pressure distribution deteriorates as this critical point is passed. The critical point occurs when the slope of the static pressure recovery versus flow curve in the single duct changes from negative to positive so that it Is possible then to have two flows, a high one and a low one that have the same static pressure (Fig. 2. 17).

iwi»:m*i mt: «iK»’ mjic :L*>

id

id *i«n<v immcuitmicj

'■ M

FIG 2.17 TWIN DUCT INSTABILITY

Son times the unequal distribution of flow between t lie two intakes is time dependent. Again as with Intake l jzz the concern is to measure the flow at which the phenomenon occurs so that it can be avoided. If the dividing wall between the two ducts is taken to the compressor entry the pressure equalising process can be transferred to the compressor exit and the instability may be prevented from developing.

The more * nplex multi-intake interactions typical of missile i istallatlons feeding a single ramjet combustion chamber are discussed in Section 7.

Fl9W.Qygn.UAy

All important internal flow phenomena and the external drag are quite ciitlcally dependent on the relative amount of flow through the intake. At subsonic and low supersonic speeds the flow quantity determines the severity cf the pre-entry pressure rise and hence whether or r.ot the boundary layer on the aircraft surface approaching the Intake will be attached or separated (Fig. 2. 18).

FIG 2.18 VARIATION Of STATIC PRESSURE IN ENTRY PLANE FOR INTAKE WITH FORWARD WETTED SURFACE

13

At supersonic speedj for an external compression intake the flow determines the extent to which the designed for shock pattern Is deformed oy the upstream movement of tne final normal shock (Fig. 2.19).

FIG 2.19 DISTORTION OF SHOCK WAVE PATTERN OUE TO SU8CRITICAL OPERATION

For an intake with internal compression it determines whether or not the intake shock system remains in its efficient 'started* configuration (Fig. 2. 20).

PR

_ 'Critical point can be

i approached only from

r/ |t supercritical side

— Subcrif/cal operation

Vac

operc

is of level of ineffic:ent j pitot intake

10

or the non dime '.is tonal mass flow function yyy Tt - Consent x A*

~ap~ r

In intake work it is usual to express measured mass flow as a ratio of flow being ingested to the flow that would be ingested at datum condi t Ions .

Thus the mass flow at the engine face:- n v2 A2 “ Po v0 Ao

and the flow ratio or capture area ratio is:~

Pp Vq A0 - AQ Po V0 A0 Ac

It is evaluated at the engine face or at a venturi section *v' just downstream of the engine face where static pressure only Is measured: -

'STARTED' INTAKE AT CRITICAL FLOW

v. -UNSrARTED' INTAKE AT SU6CPITICAL FLOW -EXPELLED NORMAL SHOCK

FIG 2.20 FLOW STATES FOR INTERNALLY CONTRACTING INTAKF

As has been seen in the foregoing sections either the value of the performance parameter being measured (such as pressure recovery or total pressure distortion) is primarily a function of flow or it is vital to define the flow at which an undesirable flow phenomenon such as ’buzz* occurs so that it can be avoided. In addition the accurate measurement of flow enables other quantities such as supersonic sldespill that are difficult to measure directly to be deduced and it enables one dimensional Mach number to be evaluated at any duct station which is a useful parameter for correlating subsonic diffuser performance.

Flow measurement in a wind tunnel is usuatly done either at the engine face, or In a venturi section Just aft of the engine face or If the control valve that varies flow is choked, Just upstream of this choked exit.

The mass flow or more accurately the mass flow rate (Kg/sec) 1$:-

W - pVA

ts. - vl2 ■ f£l . ^ . [M

Ac Pt(> k J2 Ac La*J,

Each measurement of total pressure at the engine face is associated with an adjacent static pressure (or a mean of a small number of wall statics) so that local Mach number (from P2/Pt2)

is determined ari hence the local value of [_! | Thus for n total pressure points: -

La*J2

2

AA2 fA_ Ac ** o

where dA2 is the engine face area associated with each total pressure tube.

For the venturi sect ion. j can be evaluated

from pv/Pt2& values of Pt2/Pt0 an<* Av/Ac are used in the above equation.

If there is a choked exit:-

or alternatively

For both equations it is necessary to add a calibration factor that is determined In a separate test where flows that are known to a higher accuracy than is required from the intake mass flow evaluation, are measured, by the engine face or choked exit Instrumentation.

Y*1

1+1-1 M*

2(7-1)

Thus A0 -

in p

I **.

jA*l . "2

fA 1

T

1 p<0

U Jj ^

U*J,

From the formula for sonic ares ratio:*-

A - 1

A* M

2 rl + 7-1 M>

y+T T

1*1

2(1-1)

ind Kq - P*2

£ p70 £

i

wher.' 1^2 Is an engine Face callbratlcn factor and

Cd (an exit discharge coefficient)-*** effe<.tjve

*ex geometric

2.2. 1.6 Matching of !ntake_and \<?m

Before considering definitions of external drag it is appropriate to consider the matching between intake and engine airflows.

For a given flight condition and ramjet or gas turbine throttle setting, the intake airflow supply will satisfy the engine airflow demand at one unique point on the intake pressure recovery versus flow character ist ic. The demand

character 1st Ics of both ramjet and gas turbine engines can be simulated by the characteristics of a choked hole of variable area. This area is a function of the fuel burnt in the engine. By writing the equations for mass flow conservation between entry and choked hole exit, the engine demand appears on the intake characteristic as a straight line passing through the origin. The slope of this line will vary with intake enuy area and engine throttle setting. An optimum intake entry area can be chosen to ensure ’matched’ operation at a desired location on the intake characteristic. Tie most desirable location will give a maximum thrust minus drag and will be at or very close to the critical operating point of the intake.

Fig. 2.21 shows the consequences of mismatching engine and intake airflows due to wrong choice of intake size and the effect of engine throttle setting on the match point.

FIG 2.21 EFFECT OF INTAKE SIZE & ENGINE THROTTLE SETTING ON INTAKE /ENGINE MATCH POINT

Fig. 2.12 similarly shows the effect on the match point of altering stagnation temperature (which can occur due to change of day temperature or altitude) or of free stream Mach number) both for a ramjet and (oppositely) for a turbojet engine .

I M OAT IMMUtMlJ I’WOMII

lUaOMO 04- Hil{« K

-}

• 4NHT ItlMONII

FIG 2 22 EFFECT OF CHANGE OF STAGNATION TEMPERATURE ON ENGINE INTAKE MATCH POINT

The position of the match point is further complicated in practice by the change in the shape of intake characteristics and maximum values of ingested flow (particularly for intakes with compression surfaces) as shown typically when free stream Mach number (Fig 2.23) and intake attitude (Fig 2.24) are varied.

FIG 2.23 CHANGE OF TYPICAL CHARACTERISTIC WITH FRCE STREAM MACH NO.

ftECTAHGUUB IHKKC

<9

O INCREASING

awn

%

FIG 2 24 CHANGE OF TYPICAL INTAKE CHARACTERISTICS WITH INTAKE ATTITUDE

All the .xternal forces on an engine nacelle except drag at cruise incidence are usually measured and calculated as part of the complete aircraft and are therefore outside the scope of this review. External drag however is dependent on the flow through the intake, the strength and disposition of compression surface shock waves and the shape of the intake in the immediate vicinity of the cowl lip and is therefore an important aspect of intake design and performance. External drag is usually only a significant part of zero lift or cruise attitude drag of the whole vehicle and therefore measurements on intakes are normally made at zero or small positive incidence.

2-2,2, 1 _ Pre-entry and Cowl Forces for a Pitot

Iniste

A practical thrust definition acceptable to both airframe and engine manufacturer is known as net standard thrust *..J is the difference between pressure and momentum forces at the physical exit to the engine and the incoming strearatube to the intake at free stream conditions (Fig. 2.1) I.e.

** KPex-fo^ 4 Pex ^ex2J \x~Po ^o3

The use of the free stream value of the entering momentum in the streamtube approaching the intake in this equation, implies the existence of a pre-entry thrust force in the above equation, where : -

t«e - l(Pc-Po) + Pc V) VPo V Ao

(2)

15

acts on the boundary of the pre-entry streamtube (Fig. 2. 25). Because the flow Laundary is a streamline the pressure Pjnt acting on the Interna! <,urfr.ce must be balanced by pressures Pext acting on the external surface and these external p-< ssures give rise to an equal and opposite force to Tpgp n the drag direction, Known as the pre-entry *,r in the USA, additive drag (DpRE or DADD) .

FIG2.25PRE- ENTRY & COWL PRESSURE FORCES FOR PITOT INTAKE

When the flow ratio A0 is unity the pre-entry

streamtube Is undeflected upstream of the capture area, Pc - and PCVCJAC- p0V02A0 so that TpRE <-0pR£) <s zero. This can be regarded as a

datum condition from which changes (either A0 > or < Ac) can be studied. From the datum condition, the curved pre-entry flow induces changes in the cowl pressure distribution from the stagnation line position on the cowl lip which divides external and internal flows. For Ao<1.0 these changes In pressure distribution on

the cowl constitute a thrust force which in subsonic potential flow will be equal and opposite to the pre-entry force. In a viscous subsonic flow this cowl thrust Is reduced below the potential flow value by (a) the presence of the boundary layer on the cowl and (b) the appearance of shock waves on the forward facing surfaces of the cowl if the flow over the cowl becomes supercritical.

At the datum flow condition when the external flow is free of shocks the cowl drag Is the summation of friction bnd pressure drag. This is known as profile drag and is often for convenience presented as a form factor \ in which the drag coefficient is normalised by the mean skin friction drag coefficient of a flat plate with the same Reynolds number based on total length of the cowl.

for a pitot intake Tp£E (— DpgE) can be calculated from equation (2) or more conveniently can be expressed In coefficient form in terms of tabulated flow functions as:-

tPRE

.

Pt

‘PRE

- Jr 9o*c Pt

lc lo

PT To

p p

[ ■ J.c. J° ' fol + “pte’ pto' Oo q<J

-2 A0

aT

where at subsonic speeds

and at supersonic speeds

ln the Pr

stagnation pressure change across a normal shock;

or in terms of Mach numbers as:-

r1+r^ wc’n-i l 2 ] (W Mo7)"1

Mo’lv,

i 2 J

2Ao

Ac

In s. jeh calculations the assumption is made that the stagnation line is at the cowl highlight (cap'ure area) position and that the flow is one dimensional. For a sharp lipped cowl errors due to both these assumptions are probably reasonably sm. II until A0 « Ac. For a thick lipped cowl th.- situation should be examined more carefully (especially when correlating values for drag obtained from experimental measurements and those from computational methods) and Is discussed under the heading of spillage drag.

The calculation of C7pRE or CDpR£ from

computational methods using full potential flow at subsonic speeds or the Euler equations at subsonic and supersonic speeds car. bo done directly by integrating surface pressures along the stagnation streamline. However it is often more convenient to use a similar approach to the one dimensional calculation but now using a station downstream of the capture plane for the downstream momentum plane, where the flow can be regarded r.s truly uniform and axial.

At supersonic speeds the pre-entry thrust and drag forces are still equal and opposite but a net drag force is now associated with an increase in wave drag as the attached cowl Up shock changes to a detached rave ahead of the intake capture plane when A0 < Ac. As at subsonic speeds, cowl pressure distribution changes from the datum condition in a favourable sense but the rowl thrust developed Is now much smaller and dots not nearly offset the pre-entry drag (.is It does at subsonic speeds when A0 initially decreases just below Ac and viscous effects Are swaU). Cowl drag in the datum condition is again the summation of skin friction and pressure drag. Pressure drag is now the result of the shock wave emanating from the cowl lip and can be caicula: ’ by linear theory or by the use of the method of characteristics (ref 13 svrnmarises and correlates some results from these methods) or by solving the Euler equations for th combined external and Internal flow.

With the pressure and exit momentum terms now based on station ’2’ (Fig. 2.1) the cowl internal pressure force from the stagnation line* to station '2' has to be evaluated to obtain Oj-pRE.

The internal thrust from upstream infinity 'o’ to stat ion '2' is: -

t2 ' Povoaov2 + (P2‘Po> a2 ' PoV\>

A*T l 2

and Is derived from-

16

A2*

Ao Ac*

- Ac Ao

A

a2

a2

•VPto

cT2

cT|>.t+ cTpre

Ctpre"

CDpRE

CDpRE"

CT2 " CTim

where Cr is the duct internal pressure force 1 itit

coefficient integrated from the stagnation point to station ’2'. The stagnation point can be

found accurately by locating the point at which

the surface velocity changes sign, — 2 “ * ^

Ac

Is a limiting condition at supersonic speeds.

At subsonic speeds it can be regarded as a convenient datum condition but it is not a limiting condition. This is obtained when the Intake flow is choked either at its capture plane or at some throat plane 'th' downstream of the capture plane where < Ac. Thus at low

subsonic speeds if the lip is not thin the flow can increase beyond A0 — 1.0 up to the choking

Ac

Below M0 - My, the datum condition is determined by the position of the compression surface shocks in front of the cowl lip and the maximum flow ratio (A0/Ac^max always less than unity. This maximum flow ratio is associated with a pre-entry drag and thrust corresponding to the fore spillage of 1 - fA0i and this again is a

‘A:-* max

d -.-noted by the suffix 'o'.

At both subsonic and supersonic speeds TpR£ (“DpRE ) can be obtained from application of the momentum equation vo the internal flow in a manner similar to the pitot intak* evaluation of Tpp£. For a single wedge intake oF angle b (Fig. 2. 27) this is

A Po

FI62.27 WEDGE INTAKE AT M0< MS0L..DPre0 V 0

TPRE0" P|vl>Ai cos 4 + <P|-Po> A! cos 4 + (p„-p0) Vp oVao

where Aw is the projected area of the wedge

flow A0 or Ac - Ac and the one dimensional CD

will increase again from zero. At high subsonic and low supersonic speeds Internal contraction to Ath will create a choking condition (M^-1.0) at a flow rat lo AQ < 1.0

Ac

which then precludes the datum condition ever being reached. It then has to be accepted that the intake operates in a spilling condition A0

<1.0 even at full flow and CDpRE will never attain a zero value .

pw at supersonic speeds is the constant wedge surface pressure;

pw at subsonic speeds is cither taken as Po+Pt or is obtained from 2

a correlation of pressures measured on wedges In isolation (ref 2.6).

At supersonic speeds it is often easier to evaluate DpRE directly by summing the pressures on the external limiting streamline. For a single wedge this Is very simply: -

1.2.2 .2 Pre-entry and cowl forces for m intake with a compression surface

At supersonic speeds due to the presence of a compress i'-’i surface the datum condition of A0 - 1.0 is only obtained at Mach numbers above

Ac

which the compression surface shock (or shocks) impinge on or go within the cowl Up 'M0 > Msol,. Fig .2.26).

FIG 2.26 WEDGE INTAKE AT (ct) M0 = Msol

(b) M0> Msol: OPREo=0 FOR (a) & (b)

dPRE0 - (Pw”Po><Ac-Aomax>

The expression for the two wedge compression surface is lengthier and Is given in ref. .2. 5. For the single cone compression surface pw is not constant but can be integrated numerically along the limiting conical flow streamline. The second shock on a double cone is curveu and the flow field can be evaluated by the method of characterise ics .

At subsonic speeds the choice of a datum condition is more arbitrary than for the cases considered hitherto. At low subsonic speeds the value of Aj due to the presence of the

compression surface may not b« sufficient to cause throat choking before the condition A0 - 1

Ac

is reached. In this ca'<e because of tho presence of the compression surface the entering streamline will not be undeflected and parallel to the free stream direction and thus TpRg will not necessarily be zero.

17

At moderate and high subsonic speeds and at supersonic speeds when the wedge shook is detached the throat (or capture plane) choking flow could be considered as an appropriate datum condition and this will now be associated with values of Ac that are less than unity. One other

datum condition is the flow such that the throat velocity Vj cr Vth is equal to the free stream velocity VQ .

Spillage drag cD§pill calculated from:-

Cdspill “ cDpja ■ cDext

O

wh«re CDf;CT - CDpRE(or CDadd) +

and Cdett0 “ CdPRE0 + cDcowi.0

where again the suffix 'o' denotes a datum condlt ion.

Cowl pressures for the maximum flow condition at supersonic speeds should be calculated using the flow direction and Mach number at the cowl lip that result from the compression surface flow at that position. In practice two dimensional calculations of cowl drag show that the values obtained by this procedure differ very little frorr. those obtained Ignoring the compression surface flowfieid and assuming free stream Mach number and direction at the cowl lip.

In the absence of more comprehensive data similarly It is often assumed that the decrease in cowl drag when the intake operates subcrit leal iy is the same as for a pitot intake with the same geometry cowl.

2.2 . 2.3 $Pi 1 loge Drag

As already indicated, a reduction of flow ratio below the datum or maximum flow value results in an increase in pre-entry drag force and a corresponding increase in cowl forebody thrust force (which are equal and opposite in subsonic potential flow) and therefore results in no net axial force. In viscous subsonic flow IF the flow over the cowl remains subcrit leal the change in cowl pressure gradients result in a thickening of the cowl boundary layer and ultimately, as flow ratio decreases further, to separation of the flow From the cowl Up. Under these

circumstances cowl thrust is decreased from the potential flow value and a rapid rise in drag from the datum or maximum flow value is measured that Is called spillage drag.

When the flow o»cr the cowl is supercritical this spillage drag will occur as a result of both boundary layer thickening and shock wave formation. The wave drag component will probably increase initially as flow ratio decreases or free stream Mach number increases and then be succeeded by a flow separation at the foot of a reduced strength lambda shock and finally by a complete collapse of the supercritical flow as separation moves forward to the cowl lip.

At supersonic speeds because of the increased strength of the head shock, directly flow is reduced below the datum level there is an increase in pre-entry drag. This is not balanced by a corresponding increase in cowl thrust so that spillage drag is positive when subsonic forespillage occurs. Cowl drag is reduced as spillage increases and initially cowl flow remains attached. However at greater spillages, flow can detach from the cowl lip and form a small bubble separation followed by a weak reattachment shock (Fig. 2. 28). The size of the bubble and the strength of the reattachment shock grow as flow ratio decreases further.

FIG 2.28 FLOW PATTERNS AROUND PITOT INTAKE AT SUPERSONIC SPEEDS

For spilling flows the stagnation line is on the internal surface of the lip and not at the capture plane position. Thus if the one dimensional value of based on the capture

plane area CppRg is used this should be added to cowl drag CD(X)WL l^at ls l^e suinmat*on of the axial cowl pressure force components downstream from the capture plane. If the true external cowl drag Cdcqwl is used i.e. the

summation of the axial cowl pressure force components from the stagnation line position forward to the capture plane and then back externally to the maximum diameter position then this should be added to a pre-entry drag Cp which can be one dimensionally using the PREt

stagnation line area and not the capture plane area.

I.e. either

0r CdEXT " CDpREt + C|1COWLl Unlike the evaluation of cDpRE the calculation

of CDpRE is not simple in that the position of

the stagnation line has to be found experimentally by pressure plotting the lip in fine detail or from potential flow theory or by solving the Euler equations.

The variation of drag with flow ratio at supersonic speeds for a pitot intake Is shown in Fig. 2. 29a and the more complex situation for a wedge or cone compression surface intake in Fig. 2. 29b.

FIG.2 29 ' LOW STATES % DRAG DEFINITIONS AT SUPERSONIC SPEEDS FOR (a) PITOT t (b) COMPRESSION SURFACE INTAKES

✓

18

Thls latter Figure Illustrates the potential advantages of variable geometry (by translating or varying the compression surface angle) available to the compression surface Intake but not the pitot Intake. Flow can be spilled by supersonic forespill at t much lower rate of drag increase with flow reduction than is obtained by subsonic forespill only or by combined supersonic and subsonic forespill.

REFERENCES

2.1 Livesey, J.L. (1982) 'Flow property averaging methods for compressible internal flows’. A1AA, 82-0135.

2.2 Martin, R.J. and Melick, H.C. 'A Feasibility Study for definition of Inlet Flow Quality and Development criteria'. AIAA 72-1098.

December, 1972.

2.3 Hercock, R.G. and Williams, D.D. (1974) 'Distort ion- induced engine Instability. Aerodynamic response'. AGARD, LS72-Paper No. 3.

2.4 Society of Automatic Engineers 'Aerospace Recommended Practice

ARP 1420 Engine Inlet Flow distortion guidelines. March 1978

2.5 Seddon, J. and Goldsmith, E.L. 'Intake Aerodynamics'. Collins,

London. 1985

2.6 Bryson, A.E. 'An experimental investigation of transonic flow past two dimensional wedge and circular arc sections using a Mach-Zehnder Interferometer NACA TN 2560

19

¥

1A THE INITIAL P6SI« PROCESS CONTENTS

2.3.1 INTRODUCTION

2.3.2 GENERALIZED INLET MODEL & CALCULATION PROCEDURES

2.3.3 SPECIFIC EXPERIENCE WITH THE LEVEL II INLET INSTALLATION PROCRAM (IIP)

2.3.4 CONCLUSIONS & RECOMMENDATIONS

REFERENCES

.2.3.1 INTRODUCTION

Supersonic aircraft design has evolved to the point where even the earliest phases of design can benefit many times from consideration of air intake integration, component design and performance analysis. Figure 3.1a (from Ref 3.1) offers some insight into the rationale behind this statement. In this Figure a trend can be seen toward reduced range factor in fighter aircraft which is associated with an increasing ratio of intake capture area to aircraft wetted area (Ac/A$). Increased relative intake

size in advanced fighters is required to accommodate the higher maximum mass flows associated with increased thrust/weight ratio, but it also generally means a wider range of airflows, leading to greater values of Intake flow spillage at cruise conditions and more sophisticated intake variable geometry. The Figure suggests that the combination of increased propulsion stream size and propulsion installation complexity make integration more difficult. Also, it suggests that the levels of intake and nozzle performance decrements are significant in the determi nat ion of performance. Therefore, in order to discriminate adequately among competing supersonic aircraft concepts, it is useful In preliminary design to show the relative impact of different intake integrations and/or the effect of significant perturbations in an air Intake design. Any such analysis needs to be simple and quick, but give a reasonably accurate intake performance.

ifocwi WTTtO AMU ** j -

FIG 3.1a SUBSONIC CRUISE PERFORMANCE (REF 3.1)

A number of different intako installation prediction programs exist in industry but tend to be guarded rather closely. The Level If Inlet Installation Program (UP) analysis summarized here and described moro at length in Rof.3.2 is similar In concept to other methods but Is not tied to any particular organisation's proprietory data. It was developed for the USAF by Crusuaan Aorospace Corporation primarily to predict the performance of well-defined existing Intakes and

assess the relative performance of Inlets and their subsystems for studies of airframe-propulsion integration. The program was Intended to be accurate and cimprehensive, with Improvement of the method as new data became aval table.

Another object of IIP was rapid calculation and convenient use for inexperienced users. The IIP operates interactively, using a blend of theoretical analysis and empirical correlations to model the flowfields about installed air induction systems. No extensive flowfield solvers or boundary layer codes are employed. A complete on-design analysis starting from an interactive input session can be completed in approximately twenty minutes.

UP is capable of analysing a broad range of inlets and approach flowfield conditions. The program consists of six major software modules capable of analysing both two-dimensional and ax i symmetric supersonic compression types including pitot (normal shock), all external or mixed external-internal. Inlet subsystems including boundary layer bleed, bypass, and auxiliary inlets can also be analysed. The flight conditions may range from static operation up to Mach 3.5 at any altitude. The user can also specify non-atmospheric freest real conditions, such as wind tunnel test conditions. Angle -of- incidence effects can be analysed for horizontal ramp two-dimensional inlets only (no attempt was made at estimating the 3-D flowfields of axisymmetric spikes or vertical 2-D ramps at angle-of- incidence.

Performance output from IIP consists of internal flow thermodynamic losses and external aerodynamic drag. Total pressure recovery losses are followed through the induction process (oblique shock losses, normal shock loss, cowl lip losses, and subsonic diffuser losses). A complete drag component breakdown is given with absolute values of drag components as well as CD changes relative to a specified reference condition. Flow characteristics such as onset oT buzz, oblique shock ingestion, and Inlet 'upstart' are displayed. Also included are massflow ratios for all of the inlet subsystems.

Although IIP is mainly for analysing existing geometries, limited design assistance Is also available. On-design values oi compression surface angles and lengths for maximum recovery and specified mass flow ratio is available. Also, bleed and bypass sizing for maximum recovery can bo determined. Auxiliary inlets can be sized for a specified recovery (trading auxiliary inlet mass flow ratio for reduced cowl lip losses). At off-deslgn conditions, variable geometry ramp schedules or translating spike positions can be determined for a desired mass flow ratio and an assumed throat Mach number of 0.7. This Is done by varying compression ramp anglos for 2-D inlets or translating the spike for axisymmetric Inlets while checking for undesirable shock-shock intersections and/or detached oblique shocks.

IIP is capable of using scheduled oT variable geometry and engine mass flow as functions of Mach number and angle -of- incidence . This option Is used for continuous calculation of inlet performance throughout a '"If flight onvelopo.

20

PROCEDURES

IIP models six basic elements of any inlet: 1) compression system, 2) cowl Up, 3) subsonic duct, 4) boundary layer removal systems, 5) bypass system, and 6) auxiliary Inlet system. Figure 3.1b shows these basic elements in terms of parametric influence.

The procedure for estimating inlet performance involves first calculating all portions of the supersonic flowfield up to the normal shock. The f'owfleld results are then output since the supersonic flow is not affected by engine operating conditions. The procedure then calculates an independent engine operating parameter. ’or pitot and external compression Inlets, the throat Mach number is treated os the independent parameter and governs all subsonic

FIG 3.1b IIP BASIC ELEMENTS (REF 3.2)

portions of the flowfield. Geometry Is defined such that throat Mach number may start at a choked condition (Mach 1) and be incrementally decreased, with Intake performance calculated and printed out at each Increment. As throat Mach number Is reduced below 1.0, the resulting normal shock moves upstream from the cowl lip onto the compression surface(s) until a slipline ingestion (and onset of buzz, see Section 2 of Chapter I) is detected. At this point the calculation terminates. For mixed compression intakes, normal shock position Is the independent parameter sines the throat is still supersonic and does not change with engine operation. All downstream subsonic flowfields are dependent on this position. The normal shock Is initially positioned at one third of the diffuser length or at the downstream end of a diffuser bleed section if there Is one. It Is then moved incrementally toward the throat until the verge of ‘unstart', a» which point the calculation proceeds as for external compression intakes. In this way, total pressure recovery and drag are predicted for a range of ongine operating conditions.

Specific prediction methods are described below. For subsonic operation, 1-D isentroplc flow is assumed up to the cowl Up with a linear variation of static pressuro on the external compression surfaces. For 2-D supersonic flow, oblique shock theory Is used. For oxisymmetr ic supersonic flow, on approximate empirical analysis involving equivalent conical wedges is used. Internal supersonic flow >$ calculated using an equivalent 2-D converging passage. Mass flows for bleed and bypass subays terns are determined by the difference In local surface pressure and ambient back pressure with assumed internal loss factors. Supersonic surface bleed Is unaffected by thu independent parameter, but

subsonic surface bleed or bypass mass flow does change with engine mass flow. Bleed and bypass drag is determined using momentum conservation (taking loss factors Into account) plus a component for pressure drag on exit doors protruding into the freestream. Cowl lip losses are based on empirical correlations that depend on a fairly detailed description of the lip geometry. Subsonic duct friction losses are

based on an average friction factor and the throat dynamic head. Duct divergence and offset losses are determined from empirical correlations. Spillage drag, which Is the sum of pre-entry drag and cowl lip drag. Is based on the previously mentioned reference condition. This reference condition is specified as an engine mass flow ratio. The pre-entry drag is calculated by a sin.ple pressure-area integration of the spilled st reamt ube. Cowl lip drag effects are calculated using the same empirical data for cowl lip losses in conjunction with transonic similarity theory.

Figure 3.2 shows a schematic of these calculation procedures.

FIG 3.2 COMPUTATIONAL FLOW DIAGRAM (REF 3.2)

fho IIP Program Is ueed by a number of UeAF organisations to snalyss potsntlsl or existing Intsks configurations. Thrss examples «re given here to provide in Indication of Its accuracy and show Its use In svaiuatlon.

21

FIG 3 3g SIDE-MOUNTED INLET CONFIGURATION AND INSTRUMENTATION (REF3;4)

The first example, intake data from project Tailor-Mate (Ref i.3) were used to evaluate the method's ability to predict total pressure recovery over a range of Mach numbers and pitch angles. The Tailor-Mate A-l intake (Fig 3.3a Ref 3.4) is a two-dimensional external compression, side-mounted, overhead ramp configuration designed for a Mach 2.5 fighter. The intake has three variable compression ramps, porous ramp bleed, throat slot bleed/bypass, and a subsonic diffuser with both vertical and horizontal offset. The intake was tested from Mach 0.9 to 2.5 and angles-of-Incidence between -5 and 20 degrees. Test data included engine face total pressure recovery and all subsystem mass flow ratios. Unfortunately, no drag data were taken to compare with IIP output. Several problems were encountered in modelling the A-l geometry.

First, sideplate porous bleed was incorporated in the A-l test model but no provisions for sideplate bleed are available in IIP. Second, IIP assumes duct geometry as shown in Figure 3.3b.

MOTt: W*l*001LH«fll»OMI.Lw^«Le„.lw.lWCT»»HlT

NKOIUIIY MOUrniD TO M 100 AC TO THI ACTUAL l (MOTH t AO* THROAT TO INOOM MCI.

FIG 3.3b IIP DUCT GEOMETRY (REF 3.2)

Offset can be input for one direction only for a constant area duct section. All diffusion is assumed to take place in straight sections

upstream and downstream of the offset section. The A-l duct, however, has offset in two directions and has continuously varying area (which is typical of advanced fighter

configurations). The approach taken was to input a similar offset to length ratio Into UP using the total diagonal offset distance of the A-l diffuser. Third, the throat area as calculated by IIP (not a direct input) was always smaller than the A-l data indicated. The predicted throat Mach numbers for a given engine mass flow were generally higher than A-l measured throat Mach numbers. Several sample data comparisons are shown in Figure 3.4. Recovery was predicted reasonably well at moderate angles-of-lncidence despite the geometric discrepancies, but accuracy

fell off at higher angles, were excellent, except at probably underestimated

Transonic comparisons -5 Agrees where IIP the severity of

separation at the intake sideplate leading edges.

MACH - 2.5 LCOEMO

* - leva 1 1

o - TAllOR-MATE

lOO.O r JO 300. o no o

ALPH3 • 10.0

ALPHA - 20.0

100.0 150.0 X0D 350.0

MACH - 0.9

lEGCNO ♦ - LEVEL I! o - tailor-hate

AlPW - 0 0

$6

(T e.'

100.* iM.O 700.0 350.0

ALPHA • 10. 0

lOD 0 lift • 7f0 0 750

♦T»- Cl o'

0

FIG 3.4 SAMPLE IIP COMPARISONS (REF 3.3!

i

22

Next, several advanced short-take-oTf/vert leal- landing (ASTOVL) configurations were analysed in support of an international project (Ref 3.5). The Intakes for all four configurations were designed for Mach 1.8 cruise: two employed a

single fixed compression ramp, and two were simple pitot intakes. All of the configurations involved some degree of shaping for reduced radar signature, such that the intake apertures were three-dimensional. Since IIP cannot model such feature differences, the approach taken was to maintain the correct intake capture area, throat area, and compression ramp angles for an equivalent 2-D intake. Recovery comparisons were generally very good for all the configurations, giving a reasonable degree of confidence in the contractors' quoted performance. Spillage drag comparisons were not as promising, however. Trends with Mach number compared well, but absolute drag levels were different by as much as fifty percent. USAF experience with IIP suggests that its cowl lip drag predictions are particularly sensitive to the input ifp geometry, requiring a high degree of detail. However, engineering sketches rarely give the level of detail required to discern this geometry and it must therefore be approximated. Also, the intake drag is highly dependent on the degree of intake-airframe integration. This effect also cannot be modelled by IIP since it assumes an isolated intake. The combined effect of these limitations is to reduce confidence in predicted drag levels.

Finally, IIP was employed recently to help analyse an advanced supercruise fighter forebody/ Intake test model. The IIP analysis indicated possible problems with oblique shock, detachment and the boundary layer removal system, which after further detailed analysis resulted In test model modifications.

2.3.4 CONCLUSIONS AND RECOMMENDATIONS

The LEVEL II Inlet Installation Program (IIP) has been developed for fast, accurate analysis of intake designs. It is also capable of limited design and optimisation options, particularly for off-design operation. It has been used by a number of USAF agencies and Is similar in concept and methods to US industry programs developed for preliminary intake design. The IIP total pressure recovery prediction accuracy Is considered relatively good. The drag prediction accuracy must be taken with caution, however, because of the sensftivity to input data and the integration aspects. The format of the output information is considered excellent. An entire intake performance envelope can he obtained in a matter of hours. The interactive input sessions

can howevor become tedious due to the number of responses required and IT the information asked Tor Is not available, the process must be prematurely terminated.

Several enhancements would irake a program like IIP more usable as a design tool. The unfamiliar user would be helped by *nenu' driven Input sessions. Graphical aids would be invaluable in defining parameters so that referral to a printer manual Is not required. Also, graphical display of the output would help the user Interpret the results of change of his design parameters.

Improvements of the method would be necessary to increase the application to more advanced designs. The ability to determine the effects of 3-D apertures (to first order at least) and highly offset subsonic diffusers on recovery and flow capture is desired. Sufficient data exists to develop empirical correlations for these concepts. Their inclusion would make IIP applicable to highly survlvable aircraft designs. The current maximum Mach number limit of IIP limits the use of the mixed compression programs, tut extending the empirical correlations up to Mach No. 4 or 6 would allow application to many evolving high speed rircraft.

References :

3.1. Surber, L.E. & Robinson, C.P. Survey of Inlet Development for Supersonic Tactical Aircraft AIAA 83-1164 June 1983

3.2. Tindell, R. , and Tamp I In, C. , "An Inlet System Installed Performance Prediction Program Using Simplified Modelling", AIAA 83-0567, June 1983.

3.3 Numbers, K. , "LEVEL 1 1 Inlet Installation Program Validation Results", AFWAL/FIMM Memo, September !986.

3.4. Surber, Lewis E. , "Effect of Forebody Shape and Shielding Technique on 2-D Supersonic Inlet Performance", AIAA Paper No 75-1183.

3.5. Bowers, D. , Hart, B. , and Numbers, K. , “Propulsion Integration Normalisation Team Report", AFWAL/FIMM Memo, May 1989.

23

2.4 INTAKE DESIGN & PERFORMANCE FOR SUPERSONIC CRUISE/HYPERSONIC OPERATION

CONTENTS

TURBOJET

2.4.1 FINDING A MISSION FOR HICH SPEED

2.4.2 INTAKES FOR MACH N HABER 2 TO 3+ CRUISE

2. 4. 2.1 Characierist Ics of Intake design

2. 4. 2. 2 Intake technology In current Mach 2-3+ aircraft

2.4.3 MACH NUMBER 4-6 INTAKES

2. 4. 3.1 Requisite technologies

2. 4. 3. 2 Specific applications

TURBOJET WITH AFTERBURNER

RAMJET

FIG 41 REPUBLIC XF-103 & POWERPLANT

2.4.4 MACH NUMBER 6+ TO 8 AIR INTAKES FOR

FIRST STACE ACCELERATORS

2.4.5 AIR INTAKES FOR SCRAMJET PROPULSION

MACH NUMBER 8 TO 25+

REFERENCES

2,4,1 _ FINDING A MISSION FOR HIGH SPEED

There was, for many years, a natural trend toward higher speed for almost all types of aircraft, but there seems to have beer, a more or less natural Mach number celling for particular functions/misslons . In fighter aircraft this Mach number cut off point appears to be in the 2.0 to 2.5 lunge. Supersonic strike/bomber concepts have been explored for many years, but subsonic sea level flight remains the preferred mode of penetration. Intercept/reconnaissance missions have consistently excited interest in high speed and in fact have provided the motivation for the Mig-25 (interceptor) and SR-71 (reconnaissance) aircraft, the best examples of high speed (Mach 3+) air intake technology available. Most commercial aircraft operate in the high subsonic regime, but the Anglo-French Concorde and Soviet TUKfc have demonstrated the general technology necessary to cruise at Mach 2.

MOOES OF OPERATION OF MX-1787 COMBINED TURBOJET-RAMJET POWER PLANT

In the US, other forerunners of high Mach number technology can be seen In the B-70 and in two Century series interceptor designs, the F-103 (Fig 4.1) and F-108 (Fig 4.2). All these

concepts employed efficient two-dimensional mixed compression intakes; but whereas the B-70 and F-108 were turbojet powered, the F-103 was designed for a dual mode turbojet /ramjet propulsion system. Engineers and scientists

involved in research and development In the 1960s continued beyond Mach 3 to investigate ramjet propulsion integration Into the high supersonic and hypersonic regimes. The air intake was seen to be a key factor in this development; therefore considerable attention was given to efficient diffusion of propulsion streams over a wide rungo of speeds.

FIG 4.2 NORTH AMERICAN XF 108

Th# earl ie*t work above Mich 3 concentrated on air Intake* for aubeonic conbuet ton ranjets, but when exploration of aver higher speed, reached about Mach 8, th* potential for raaijet effective apeclflc Input** wa* dropping rapidly (Fig 4.3). Thl, observation brought consideration of supersonic conbustlon to th* forefront and originated th* idea In th* US for an aerospace plan*. A number of different concept* for such a

8-25

vehicle were explored In the early 1960’s, all making use of scram jet propulsion above about Mach 6-8. Unfortunately, the magnitude of propulsion technology barriers, resistance from the rocket community and the more Imnediate needs for propulsion integration at lower speeds pulled attention away from hypersontes. There was a strong tendency to concentrate resources on transonic/supersonic Intakes as opposed to very high speed flight. Any high altitude supersonic system that could be conceived was perceived as vulnerable to projected defenses.

FIG 4.3

SPECIFIC IMPULSE OF HYDROGEN-FUELLED ENGINES

When mission analysts began to look beyond Mach 3 significant levels of interest in high speed flight reemerged. In the US interest in high speed flight came about In the resurrection of the single-stage-to-orbit airbreathing aircraft concept. The National Aero-Space Plane (NASP) , makes use of a dual -mode ramjet /scramjet engine to achieve near orbital velocities. A related mission would be the so-called "orient express," a hypersonic passenger aircraft which would provide quick trans-Pacific transport. The work on NASP, then, led naturally to examination of other intermediate speed appi lent ions , both as new missions and as alternatives to NASP. Starting from the lower end of this Mach 4 to 25 regime, mission analysts were able to see utility for systems operating from Mach 4 to 6 for intercept, reconnaissance or interdiction purposes. Beyond that, in the Mach 6-8 range, potential for a ramjet or ramjet -powered first stage of a t wo-st age-to-orbi t vehicle has been examined in a number of places, e.g., the German Sanger vehicle. Also, suborbital Mach 8-12 missions havo been postulated which hold considerable promise in terms of both reaction time and survivability (Ref 4.1 & 2).

Thus, the possibility of high :;peed flight has been revived, and with It, the nood for high speed air intakes. Sovornl distinct areas of Interest can be Identified*

Max Mach number

a. 2-3+

b. 4-6

c. 6-8

High speed civil transport (HSCT) Transport L* Avion ^ Grande Vitesse (AGV), Intercept, Reconnaissance & Interdict ion aircraft First stage of two-atage-to-orbit vehicle (Sanger) or part of a rsnjet /rocket single-stage-to-orbit (British Hotol) concept

d.

o 8-12 Reconna is since, fast -response

global interdiction aircraft

o 12-25 Air breathing slngle-stafe-to-orbit

(SSTO) vehicle

It should be noted that the value of these possible flight vehicles does not hinge on the anticipation of a future major power conflict. Some of them, however, suggest the need for continuing allied power vigilance and the recognition that there will continue to be global "hot spots" which could require high speed

intervention (Ref 4.1). Inherent In the^ie

technologies is the potential for civilian application. Interest has been revived in high speed atmospheric flight, and the challenges for high-speed air intakes are substantial,

increasing in difficulty as Mach number

increases. *hlle there are many overlapping requirements for Intake technology across these regions, several key factors can be considered. The balance of this section will deal with progress in technology and requirements for future solutions.

24.2 INTAKES FOR MACH 2 to 3+ CRUISE

24.2.1 Characteristics of Intake Design

Air intakes designed for supersonic cruise have different priorities to those designed for supersonic dash flight. In regard to dash aircraft, intake designs such as those used on F-15 and Tornado provide relatively high pressure recovery and low flow distortion, even during combat manoeuvres. They place relatively less emphasis on low drag. Supersonic cruise aircraft on the other hand, spend the major portion of their time at or near a supersonic design point; consequently their intakes require reduced cowl lip thickness, more efficient boundary layer bleed and relatively sophisticated bypass systems to provide a precise mass flow match between intake and engine and stable operation. Intake bleed flow in particular must be exhausted efficiently to minimize potential drag contributions. As supersonic cruise takes place at higher and higher Mach numbers, the Importance grows of reducing contributions to bleed, bypass and spillage drag and the need to reduce cowl drag leads to the use of mixed compression. The technology currently available to accomplish these tasks can best bo illustrated by their implementation on current aircraft and in the design of advanced supersonic cruise aircrrft.

2 4.2,2 Intake Technology in Current Mach_2-_3+

Mrp.r.flfi

The Concorde, operating at Mach 2.0, represents a unique intake employed at the lower edge of the supercruise range (Ref 4.3). The Concorde designers were able to employ a low drag external compression design (Fig 4.4). This intako, as can

Wing jechon

25

be seen from the Figure, is basically a three-shock design, but adds some isentropic compression on the second ramp and operates between a complex field of expansions/shocks internally near the ramp bleed gap, and the standard normal shock at intake entry. By allowing a small amount of duct contraction aft of the entry plane the external cowl angle is kept to a minimum consistent with shock attachment at the lip and a self starting intake. The wide bleed slot behind the compression ramp allows a free shear layer to accommodate changes in mass flow. It automatically adjusts the bleed slot streamline and the amount of flow removal required For flow matching. From a purely mechanical point of view, the Concorde intake is a model of simplicity. It uses a variable second compression ramp and an auxiliary intake (for takeoff) combined with a dump door for over-board bypass. Transient controls activate the dump door which is made easier due to the fact that the inlot operates stably Trom supercritical to subcritical conditions. Its advantage is seen in the combination of simplicity and reliability with performance adequate for Mach 2.0 operation. With this type of design (ie without mixed internal/external compression) at higher Mach numbers it would be difficult to maintain high total pressure without unacceptable increases in drag and bleed flow quantity.

Altho-fch in the US a prototype supersonic transport was never built, its air intake was one of the more fully developed designs in the history oT US research and development. Variations on Its axisymmetrlc mixed compression design continued in NASA for several years after abandonment oF the SST itself, exploring mixed compression ranging from 40% to 60% internal contraction, developing highly efficient boundary layer control and refining means of achieving inlet stability. TJonneland outlined some of the chief design characteristics of the Boeing SST intake design in Ref. 4. 4 as it applied to the Mach 2.7 commercial cruise vehicle (Fig. 4. 5a). This reference points out that in the development of the intake design, a number of improvements in intake design were required, especially In the areas of boundary layer prediction and control, intake stability and intake/engine airflow matching. It noted that, even with optimistic pressure recovery and bleed estimates, an SST with a 3500 mile range would experience cruise range loss in cxces.: of 12% due to total pressure losses and bleed drags alone. As a consequence.

It was necessary to set very high performance and stability objectives for the Intake. For example, the 'started' intake at Mach 2.65 operated at a pressure recovery of 91% and was designed to accept a step reduction of 5% in engine corrected airflow without controller action and remain 'started'. Similarly,

operating on-deiign, the inlet was not to 'unstart ' when experiencing a Mach number reduction up to 0.05. The overall system was matched so that no overboard bypass was required for matching during normal climb and cruise operation. The complex Boeing design

Incorporated:

a. A translating spike for matching and bleed opt imitation

b. Four throat doors to obtain flow area variation needed from take-off to transonic condi t ions.

c. Four overboard bypass doors (which In turn incorporatad suck- in doors) to spill excess airflow during descent or other conditions involving an inoperative engine.

d. Distributed centerbody spike and cowl bleed plus a spike shoulder scoop bleed for boundary layer and shock wave boundary layer interaction control (Fig 4.5b). An intricate control system blended these variable geometry features and an automatic vortex valve (Fig 4.5c) to solve problems with:

o Takeoff/landing performance o Takeoff noise abatement

FIG 4.5a INTAKE GEOMETRY

FIG 4.5b BOUNDARY LAYER BLEED SYSTEM

FIG 4.5c VORTEX VALVE SCHEMATIC

FIG 4.5 BOEING SST INTAKE

26

o Stable supersonic "started" operation

o Intake fuzz suppression

o Optimizing supersonic diffuser/bleed system operation

o Minimization of bleed drag o Normal shock system stability

Even recent reports on NASA's Investigation of an advanced SST (Ref 4.5) feature a similar intake design, but operating with advanced higher

temperature variable cycle engines. The intake could employ an expanding centerbody compression surface for noise reduction features such as

intake choking during landing or takeoff to block noise from propagating outside the Intake (Ref

4.6) . Similarly, updated versions of the

Concorde (Ref 4.7) have been suggested (Fig 4.6). The ATSF (Avion Oe Transport Supersonlque du Futur) would operate In the Mach 2.4 to 2.5

regime, possibly with a variable cycle engine which would operate as a turbofan at take off and as a turbojet for supersonic cruise, the "flow multiplier" fan being driven by a secondary turbine for takeoff and subsonic cruise (Fig

4.7) . With such a flight vehicle, crucial needs

include a "smart" control system, a high

efficiency intake and careful integration of all aspects of engine and airframe design from the outset of development.

CU«* KMltultoA

It would have seemed incredible 25 years ago to predict that the YF-12/SR-71 would persist as the state-of-the-art for Mach 3+ manned cruise flight, but even as *t retired, it was the reigning example of high supersonic cruise. A number of reports on this aircraft's intakes (e.g., Ref 4.8-13) offer insight to the flving Mach 3+ cruise state-of-the-art for induction systems. This Mach 3+ axi symmetric mixed compression intake operates essentially as an isolated intake, only a slight downward cant and toe-in being provided to align the intake with the forebody flow field (Fig 4.8). The

FIG 4.8 LOCKHEED SR-71A

compression spike translates for off-design spillage operation and to provide inlet 'restart' capability. Its bleed regions for boundary layer control and shock stability are located on both the spike arid cowl. Spike bleed is by means of a series of surface slots. Cowl bleed uses ( combination of flush slots and a ram scooj: referred to as the shock trap (Fig 4.9). The INLET DETAILS

FIG 4.9 DETAILS OF THE SR-71A FULL SCALE AND 1/3 SCALE INLET SYSTEM

YF-I2/SR-71 Intake system has two bypi.ss systems, using the forward bypuss to position the normal shock and to dump large amounts of flow during a •restart’ cycle. The aft system provides for some engine cooling, but Is also used for engine matching below Mach 3.0 (Fig. 4, 10). A digital Inlet control developed later was able to attenuate disturbance inducod shock excursions for frequencies of about Ihertz and below.

CINTtWOOV BltlO SOCK4NOOOASC«H

ULCIOA 0.05*0

com. auto

CNOMt COOL NOAM

smlMVtNWN

PoamoN * hit

VOJLDUWVAT S**l MOvMIOWMAOTO ftf STA*T TH* HU

•mssoooMaowD

SNSUMDMOOAV- OftNASftCOUMO to WTION SHOCK

OAMCSTAAT HIT

URTIMV DOOMS aOHO CITSMM. MCTQA

r\>**fvuo«N

FIG 4.10 AIR FLOW PATTERNS-STATIC AIRCRAFT (TOP)

-HIGH SPEEOt BOTTOM!

27

Development of the SR-71 Inlet during its lifetime has dealt largely with terminal shock stability problems. Solutions working with existing flight vehicle actuation hardware met with some success, but other systems which would require more drascic inlet modification showed excel ten., promise. In one case (Ref 4.10)) a YF-12 aircraft inlet modified to provide a porous cowl -bleed region just upstream of the intake’s shock trap and bleed flow was controlled by relief-type mechanical valves (Fig 4.11), designed for high Mach number (Mach 3+) flight.

FIG 4.11 YF-12 INLET WITH STABILITY SYSTEM

The valves provided their own reference pressure and, hence, did not respond to the slow response perturbations handled by the inlet's control system. Such a system was shown to be able to absorb diffuser-exlt airflow disturbances that are too fast for the inlet's production control system. If the SR-71 were being developed today, it appears that the level of technology would be adequate to facilitate synthesis of a more stable Intake .

For the most part, the difficulties of designing inta* ; at Mach 3 continue Into the Mach 4 to 6 regime, but with revised priorities. Major changes in emphasis stem from the off-design flow matching problem and from the utilization of a combined cycle turbo-ramjet engine (Fig 4 12).

FIG 412 COMPARISON OF INLET DRAG COMPONENTS

Maximum Mach number propulsion flow requirements produco a large intake capture area relative to other ai rcraFt dimensions (Fig 4.13). This

u~|0 U-*U0

FIG 4.13 SIZE OF CAPTURE AREA & INFLUENCE OF PROPULSION SYSTEM GROWTH WITH SPEED

situation creates an acceleration problem because the inlet sized for maximum Mach number Is unable to pass the capture area stteamtube at transonic and low supersonic flight conditions. Coupling this situation with reduced engine demand at the lower Mach numbers requires spillage bypass of up to 90X of the approaching capture area streamtube (Fig 4.14). Cetting rid oT the excess air, In turn produces high spillage and bypass drags. Consequently a critical need is Identified for Mach 4 to 6 Intakes which exhibit low off-design drag. On the other hand, the requirement to maximize total pressure recovery (as with a Mach 3 turbojet) is relaxed In a high Mach number environment where a ramjet is either included with the turbojet or takes over completely.

Mac' number

FIG 414 COMPARISON OF lN^T SUPPLY & ENGINE DEMAND vs CURRENT PROPULSION SYSTEM

A number of design issues can be identified for air intakes operating in the Mach 4 to 6 range. In the supersonic diffuser the compression ramps are designed not only as to number and length of ramps, but also whether to use hinged ramps exclusively or include flexible ramp(s) to approximate isentropic compression. Variable geometry also helps to provide efficient transonic spillage. An efficient i- -’ndary layer control system for compression ramp(s) and sidewalls is designed not only to maximize pressure recovery, but also to maximize total system performance with absolute reliability, intake throat design for these applications considers the effect of throat Mach number on performance and stability and It provides for normal shock position control, whether by bleed, shock trap mechanism or other device. Design of the flight vehicle itself also considers the forebody boundary layer - whether to divert , bleed or Ingest some or all of it. Because of the high pressures and temperatures in the Intake, mechanical design takes on increased importance for accurately controlled actuation, for sealing of ramps against leakage and for providing adequate thermal protection for subsystems. Even the subsonic diffuser design is critical. Care must be taken to maintain attached flow with the high area ratios associated with maximum Mach number, (a) to provide for smooth transition from turbojet to ramjet operation, (b) to use the ramjet duct appropriately for bypass in low speed (transonic) flight and (c) to assure adequate diffuser wait cool ing.

The foregoing deals with air intake design as a component, but in thla Mach 4 to 6 regime the integration of air intake with the airframe takes on increased importance. As mentioned

previously, the intake size at higher Mach numbers increases with respect to the rest of the aircraft. This forces the vehicle designer to consider utilizing part of the airframe as Inlet comptesslon surface, thereby effectively reducing airframe wave drag and, potentially, wetted area.

l

:» I r frame wave drag and, potentially, wetted area. Air intake design - as well as the entire propulsion system - mist bo accomplished integrally with the flight vehicle from the outset of preliminary design.

?• 4 . 3 . 2

boundary layers are quite different in height and shape and can fill a significant portion of the intake (Fig 4.16). The differences were seen to be determined by parameters such as forebody transverse curvature, transverse static pressure gradient and the degree of Isolation of the forebody compression from crossflow influence.

Intakes may not var;' greatly In general appearance, but still provide dramatically dlffeient engine flow condl t Ions . Reference 4.14 Indicates that the fuselage forobody boundary layer will be turbulent, and could be ingested at least partially by the Intake In ar. integrated configuration. It notes the potential for Improved aircraft performance from forebody precompression and offers experimental data from four different compression surfaces at Mach 6.0 to show potential differences In boundary layer progression (Fig 4.15). The four developed

4 8

ia *ci°

12 16

With this potential influence of forebody boundary layer development in mind, a specific study of Mach 4 to 6 airframe - propulsion system integration is considered (Ref 4.15). This investigation found intakes with low boundary layer bleed requirements and good angle-of- incidence characteristics to be highly desirable. Such attributes were found to permit efficient engine operation including manoeuvre at hypersonic speeds. If, as in many applications, o hypersonic manoeuvring requirement sizes the intake capture area, this study showed it possible to achieve a 20H to 50M reduction in intake capture area by using horizontal ramp intakes or shielded Intakes (Fig 4.17). In this case, airframe integration of the intake has been used effectively to reduce flow spillage requirements. Also, If done properly integration can lead to reduced vehicle weight, fuselage wave and viscous drags (Fig 4.18). On the other hand,

FIG 4.15 SCRAMJET INLET VISCOUS EFFECTS FI6 4.17 MACH 6 RECONNAISSANCE AIRCRAFT

FIG 416

BOUNDARY- LAYER SHAPES FOR TWO ANGLES OF INCIDENCE & FOUR FUSELAGE SHAPES SCRAMJET ENGINE OF HEIGHT h AT THE ENGINE INLET STATION

FIG 4.18

PERFORMANCE BENEFITS DUE TO INTEGRATION

2y

the study warns that non-optimum structural designs can actually Increase weight in integrated concepts and that high drag boundary layer dWerters may be required. In this design task special effort was extended to produce a compact ralxeu compression Intake design which employed Isentroplc compression In both ramp and cowl, and internal shock cancellation in the

FIG U. 19

MACH US INLET DESIGN SHOCK STRUCTURE

throat region (Fig 4.19). Bleed on both r«rou and cowl sides of this shock cancellation corner is used to prevent separation due to shock boundary layer Interaction. This bleed also provides a measure of stability by increasing sharply with forward Movement of the shock. I. is estimated that this type of inlet would require only about 5V» total bleed, Including IS on each of the sidewalls. The idea behind this is that any range decrement due to loss in total pressure recovery would more than be made up by the reduction In bleed drag. Experience with the conventional mixed compression NASA LFRC Mach 5.0 inlet test demonstrates that high Mach number inlets can demand h”-* amounts of Meed flow to produce stable, h.b\ pc; forma nee operation. The focused compression concept of the Ref 4.15 study is inherently high In r'.s. , but offers sizing and bleed drag reductions , rthy of further oxplorat Ion.

Acrorpat ialo, as well as US concerns, have been examining commercial applications of Mach 4 to 6 flight. The French version, L'Avion a Crande Vitesse (AH/;, would cruise at Mach 5^(Fig 4.20).

FIG 4 /0 AEROSPATIALE HYPERSONIC. AIRCRAFT CONCEPT, AGV

Its designers were qtioied (Reference 4.16) as being confident that l hey possess the necessary theoretical and experimental bases required to intiiatr the design of „uch an aircraft, ’’Ticy cite experience with ram-propulsion engines (ramjet and rarnroc'-.et ) in missile work and capability to deal with -supersonic flight on a large aircraft Trom Concorde experience. This would presumably include the ability to deal with air intake problems, but based on recent US experience there is a considerable leap in complexity of Intakes from Mach 2.0 to Mach 5.0,

Judging from the results of investigations ard developments over the past *hirty years since Mach 4 it, 6 propulsion Integration work began In earnest, there appears to he little doubt that the foundation technology for manned Mach 4 to 6 air breathing flight exists. The question, rather, is whether there is sufficient military support or <iv!lian potential to pay for the development and flight test of such a system.

2 4.4 MACH 6+ TO 8 AIR INTAKES FOR FIRST STAGE ACCELERATORS

This application is represented most notably by Germany’s Sanger Vehicle, but numerous versions of this alternative to the NASP-type single-stage-to-orbf t have been generated. It offers lower risk in terms of propulsion system development in general and, consequently, air intake technology. In the case of the Sanger, the large first stage launch platform accelerates to about Mach 6.8 powered by turbojet -ramjet engines (Fig 4.21). At this point it separates

FIG 4.21 SANGER BLENDED BODY/ WING CONFIGURATION

from t ho smaller, top-mounted "Horus" second stage which goes into orbit via a rocket (Ref 4.17). Another proposed system is Hotol , an orbital vehicle conceived ny British Aerospace ard Rolls Royce (Fig 4.22). While Hotol is not

FIG 4.22 HOTOL CONFIGURATION

per se a staged vehicle, the air breathing propulsion system operates only to about Mach 5.0, at which point ascer.t continues on rocket power. Thus many of the air Intake requi reti.onts arc the same as for the S'dnger, the primary difference coming, apparently, from the secondary use of the Hotol inlet during ascent to gather air for production of liquid oxyg'n used In Us rocket phase (Ref 4.18). Air intakes for this type of vehicle would vary from those of the previous section In that , as

accelerator-dedicated components, tnoy would not be so heavily concerned with bleed drag nor design point performance. They might the re fore

30

provide applications Tor the so-called oversoed intake, (Fig 4.23} somewhut undersized for the mr.ximum Mach number condition in order to reduce spillage drag at the transonic condition. Mach 8 Is given as the upper 11. nit Tor this application with the assumption th; ? it would be feasible to

FIG 4.23 OVERSPEO MIXED COMPRESSION UNIT

consider usirg subsonic combustion ramjets to this speed. In any case, avoiding the Jump in technology requirement associated with scramjet operation would have a strong Influence on the maximum Mach number of staging or shifting to rocket propulsion.

Some valuable measurements on research type axlsymmetric Intakes or perturbat ions on ax 1 symmetric Intakes in this Mach number range (5 to 8.5) were done by NASA Langley In the 1970s (fiefs 4.19-21).

2 4..’) A i R INTAKES FOR SCRAMJET PROPULSION. MACH 8 TO 254

A. indicated previously there arc some mission differences associated with various levels of maximum scramjet powered flight Mach number. Suborbital missions have been defined for the Mach 8 to 12 regime, but the greatest attention by far recently has been given to the development of a single-stage-to-orblt flight vehicle using dual mode (ramjet -scramjet ) propulsion. In general the technology challenges are similar across this regime, but obviously escalate with 2 nc reasing Mach number.

Some of the primary challenges of high Mach number vehicles were well defined in 1986 (Kef 4.22). It was pointed out that in ramjet /scramjet -powered vehicles, Intakes and exhaust nozzles would comprise the major part of the entire vehicle so that intake performance would be directly related to forebody precompression. At the same time this inlet precompression creates a significant portion of the vehicle's lift and pitching moment. Also, intake, airframe, combustor and exhaust nozzle must be structurally integrated. While the ah frame and propulsion system are thus inextricably linked, the aerodynamic Interface varies with engine operating mode, complicating the responsibility For the design weight and performance , Ref 4.23 (1987) asserted that there was sufficient knowledge to design and build rocket -boosted scramjet-powered missiles for Mach 3 to 7 operation, but that extension to orbital speeds or integration with a vehicle that operates from take-off would require "additional development." This reference also mentioned the role of CFD (computational fluid dynamics), saying that it was still In n formative stage and lacked fundamental process data to validate a number oT the physical and chemical models being used. Intake development, in particular, would be affected by the need to understand turbulence, especially at supersonic end hypersonic flight speeds, and to be able to predict its effect on wall shear and heat transfer, boundary layer separat ion/reattachment , fuel injection and mixing as well as chemical kinetics. Reference 4.24 (also in 1987) echoed these basic thoughts, pointing out that, in the area of boundary layer

transition and development, factors that had been Insignificant at lower speeds would assume major roles at hypersonic speeds. For instance, rapid growth of high speed boundary layers effectively changes surface chape and "entropy swallowing" pulls energy from the stream that would otherwise energize downstream boundary layers. Wind tunnel verl f k *v. ion testing would be compromised by the fact that wall noise interferes with transition. Inlet development and analysis is further complicated by the significance of real gas effects above about Mach 10.

With this background, it is appropriate to review some recent intake technology developments and notice the advancement of capability in the relatively short period of time from 1987 to 1989. First, Ref 4.25 reports development on Intake boundary layer control at the lower Mach numbers. This work present:; Navicr-Stokes solutions Tor strong shock interactions. Incident oblique shocks, compression corners and shock expansions associated with high speed air intakes (Fig 4,24). Also, it presents results of

sxcweowtMfimrfR

FIG 4.24 SHOCK INTERACTION FLOW FIELDS

tangential air injection to control shockwave-boundary layer interaction in the Mach 3 to 5 region, showing the importance of proper injector location in order to effect flow control. Another source (Ref 4.26) discusses the application of a 3-D Navier Stokes code to the analysis of an intake module where flow Is compressed by swept, wedge-shaped sidewalls. The combination of sidewall sweep and aft cowl placement in this fixed geometry design facilitates efficient spillage and good inlet characteristics over a range of operating Mach numbers . Three wedge-shaped fuel injection

struts double as additional compression surfaces to complete the diffusion process (Fig 4.25).

FIG 4.25 SCRAMJET ENGINE MODULE AND ITS CROSS SECTION

Obviourly, the threat flow Is highly complex, experiencing side wall Interaction with strut shocks and expansion waves, cowl shocks, ond effect expansion waves and separation induced shockwave?. The author points to good agreement oi this complex analysis with experimental data in order to claim good ability to simulate the

31

complex Flow fields- associated with sophisticated hypersonic vehicle air Intakes. The utility of such analysis capability Is particularly evident f*"om the observation In Ref 4.27 of very significant differences «n N4SP Intake eor.'opts. The propuKion modules associated with the two concepts described will be testei. over the Mach 0 to 8 range. Accurate intake <-omputat ional simulation will be crucial not only in the intake designs, but also In assessing the Influence of different combustor approach flow conditions and extrapolation of propulsive performance to cover the full range of scramjet operation (Fig 4.26).

dynamic mtcucnon briwccn flow-field solutions in the s.-nous scram jn enfinr 'low regimes.

FIG 4.26 POSSIBLE CFO METHODS FOR INTAKE ANALYSIS

Even the most optimistic plans for high speed test facility investment (Refs 4.28 and 4.29) will provide only spot checks of Intake flow development and intake f!ow effects on propulsion system performance. Thus for a

singic-stago-to-orbit scramjet vehicle, much of the intake development must be accomplished by computational moans and verified Insofar as possible by experimental test.

There ;.rc some areas of interdisciplinary Intake development which as yet do not lend themselves to any kind of exclusively computational analysis. Actively cooled leading edges

applicable *o intake cowl lips are an example of such developmental work. The designs must be analysed, fabricated and tested In a hot gas facility. Analysis techniques at this point are only Talr, but are still a useful adjunct to continuing experimental development.

While there are many hurdles yet to clear, the advances in knowledge pertaining to hypersonic air-breathing flight have been impressive. Propulsion and propulsion integration have always been regarded as key areas, but progress has been made on a number of fronts, including Intake development. Rapic advances in areas such as turbulence modelling, transition prediction, leading edge cooling and combustion analysis arc helping researchers to define realistic intake and ramjet and Intake - ram/scramjet designs for future hypersonic flight vehicles. As in many cases, increased knowledge brings Increased ui.dcrst and! ng oT the cost Involved. In the re l at we l y near term the cost of development and ownership of hypersonic flight vehicles »vi 1 1 be sufficiently clear to determine whore limited resources should be concentrated.

REFERENCES

4.1. Williams, Robert M. , "National Aero-Space Plane: Technology for America's Future," Aerospace America. Nov 1986, pp 18-22.

4.2. Coleman, Herbert J., "NASA, Defense Dept Award Ccntracts for Aerospace Plane." AW&ST .. 14 Apr 1986, pp 24,25.

4.3. Seddon & Goldsmith, intake Aerodynamics. Ch 12 - "Matching & Control" - The Concorde Intake, pp 329-333 - A1AA Education Series.

4.4 Tjonneland, E. , "The Design, Development

and Testing of a Supersonic Transport Inlet System." ACARD CP 91-71, Sop 1971.

4.5. Driver, Cornelius "How Different a Modern SST Would Bo," Aerosi-ace America. Nov 1986.

4.6. Kandebo, Stanley W. , "HSCT Propulsion Studies Focus on Reducing Emissions, Noise" AW&ST. 10 Jul 89.

4.7. "Birch, Stuart, "Advancing Technologies," Internationa] Viewpoints Section, Aerospace Engl neer Ing , Oct 89 .

4.8. Johnson, Clarence L. , "Development of the Lockheed SR-71 Blackbird," Lockheed Horizons . Winter 1981 <r1982.

,.9. Smeltzer, Donald B. ; Smith, Ronald H. ; and Cubbison, Robert W. , "Wind Tunnel and Flight Performance of the YF-12 inlet System," Journal of Aircraft. Vol 12. No 3, Mar 1975, pp 182-187.

4.*0. Cole, Cary L. ; Dustin, Miles 0.; and

Noiner, Ceorge H. , "Wind Tunnel Performance of a Throat Bypass Stability System for the YF-12 inlet." NASA Conr Pub 2054, YF-12 Experiments Symposium. Vol 1. Sep 1978.

4.11. Reukauf, Paul; Olinger, Frank V.; Ehcrnbcrger, L.J.; and Yanagidate, Craig, "FI ight -Measured Transients Related to Inlet Performance on the YF-12 Airplane," NASA Coni Pub 2054, YF-12 Experiments Symposium. Vol 1, Sep 1978.

4.12. Cole, Cary; Neiner, Coo.; and Dustin,

Miles, "Wind Tunnel Evaluation of YF-12 Inlet Response to Internal Airflow Disturbances With and Without Centro I,"

NASA Conference Pub 2054, YF 12 Exper I, .rents Symposium. Vol, Sep 1978.

4.13. Wilson, J.R. and Wright, V.A. , "Use of Engine speed Trim to Automate the Supersonic Englne/lnlct Match in the AR-71 ," 29 Jun - 2 Jul 1987.

4.14. Lowing, Peirce L. and Johnson, Charles B. , "Inlet B.L. Shapes on Four A/C Forebodies at Mach 6, "Journal of Aircraft Engineering N^tes . Vol 15. No 1 Jan 1978.

4.15. Saiemann, Victor and Andrews, Mark, "Propulsion System Integration for Mach 4 to 6 Vehicles, "AIAA Paper, 1988.

4.16. Birch, Stuart, "Hermes Update. "Aerospace Engineering. Feb 1989.

4.17. Covalt, Craig, "Sanger Aero-Space Plane Cains Increased Support In Europe," AW&ST, 10 Jul 89.

4.18. Burns, BRA, "Horizontal Takeoffs fnr Future Launch Veh i c 1 es , " Aerospace Engl nee ring .

Aug 87.

4.19. M C Torrence, ' Exper imentn) investigation of a Mach 6 fixed geometry fnlot featuring a swept external - internal compression field' NASA TN D-7998 Oct. 1975.

32

4.20 E.H. Andrews, J.W.Russell, E. A. Mackey & A.L.Siwoonds. ' An Inlet analysis for the NASA hypersonic research engine aerothermodynamlc integrat Ion model * NASA TM X-3038 Nov. 1974.

4.21. E.L. Andrews , A.M. Agnone & S . Z. Pinckney , 'Experimental & analytical study of an inlet forebody ior an airframe integrated scramjet concept.’ NASA TM X-3158 Jan.

1975.

4.22 Coons, L.L., ‘Propulsion Challenges of Hypersonic Flight," AIAA-86-2620, Oct 86.

4.23 Waltrup, P.J., "Hypersonic Airbreathing Propulsion: Evolution & Opportunl t ies, " ACARD CP No 428, Aerodynamics of Hypersonic Lifting Vehicles," Paper #12, 1987.

4.24. Johnston, Whitehead and Chapman, "Fitting Aerodynamics and Propulsion Into the Puzzle," Aerospace America. Sep 87 pp 37ff.

4.25 White, M.E.; Thompson, M.W. ; Carpenter A.; Lee, R.E.; Yante, W.I., "Tangential Mas* Addition for the Control of Shock Wave/Boundary Layer Interactions in Sc ramjets Inlets, "IX ISOAGE, Sep 89, Athens, Greece.

4.26. Kumar, Ajav, "Numerical Simulation of Flow Through a Two-Strut Scramjet Inlet," A1AA J.of Propulsion, Vol 5, No 3 May-Jun 1989.

4.27. Kamiebo, S.W. , "NASP Program Office Retains Two Propulsion Contractors," AW&ST . May 8, 1989.

4.28. Anon, "Aerojet Tech Systems Facility to Test Hypersonic Engines Designed for NASP," AW&ST. 12 Jun 89.

4.29. Harvty, David, "NASP Program on Track, " Defense Science. May 1989, ppl6-21.

33

✓

1A _ INTAKE Malta ft PERFORMANCE FOR MULE

EE£

CONTENTS

2.5.1 INTRODUCTION

2.5.2 ISOLATED INTAKES

2.5.2. 1 Internal flow

('•) Flow in the subsonic diffuser (b) Combination of Intake and subsonic diffusers

2. 5.2.2 External flow

2. 5.2.3 Intakes with compression surfaces at subsonic and supersonic speeds

2.5.3 INTAKE- AIRFRAME INTEGRATION

2. 5. 3.1 Fuselage flow fields for side mounted inlets

2. 5. 3. 2 Performance of a rectangular compression surface intake on the side of a fuselage

2. 5. 3. 3 Performance of half cone intakes on the side of a fuselage

2. 5. 3. 4 Performance of a pitot Intake on tne side of a fuselage

2. 5. 3. 5 Fuselage and wing flow fields for shielded intake installations

2.J.3.6 Performance of shielded compression surface intakes

(a) Rectangular intakes

(b) Half axisymraet ric intakes

2. 5. 3. 7 Comparison of performance of shielded and unshielded rectangular and half axisymmetric inles

2. 5. 3. 8 Performance of shielded pitot intakes

2.5.4 TECHNOLOGY IMPLEMENTATION IN CURRENT AIRCRAFT

2.5.5 CONCLUDING REMARKS

2.5.1 INTRODUCTION

Ine design of intakes for high performance over the range of conditions experienced by an agile strike-fighter aircraft is not simple. The difficulty in intake design is trying to achieve good efficiency, low drag, and high margins of stability at all operating conditions together with 1-w w/»i^ht and cost. Compromises in all four characteristics are usually made to maintain acceptable performance levels at various flight Mach numbers, angles of Incidence, sideslip and power jetting. The limits of speed and attitude car, be highly variable depending on requirements which become Increasingly severe with the passage of time, and, in particular, whether or not post stall manoeuvring is regarded as being necessary. Typically, a Mach number range from zero to 2.0 or 3.0, a range of Incidence (without pose stall manoeuvring capability) from -10* to +40* , a

sideslip variation of 110 - 20* and a range of

engine airflows from maximum to flight idling may be required. Under these conditions, separated flow end complex shock and boundary layer

interactions at or downstream of the intake entry plane will be present over an appreciable part of the flight envelope.

The major features of flow patterns around and inside an air intake installed on the side of a body are shown In Figs 5-la-h. At zero and small incidence, internal Up separation can occur at

high engine atrflows and low forward speed (Fig 5-la). At high forward speed with an Ingested streamtube size smaller than the Intake at all

engine flows, separation can occur externally over the cowl and in the pre-entry boundary layer on the approach surface to the intake (Fig 5-lb). Separation may occur Internally also, particular!/ at high intake throat Mach number, if the duct shape bends and/or diffuses too rapidly. At high incidence separation occurs at the lower lip internally. Up?»ash around the body will often lower the threshold of incidence at which this will occur (Fig 5-lc) .

A similar situation occurs at yaw as depicted In the port intake of Fig 5-ld. At supersonic speeds the well known lambda shock formation (Fig 5-le,

starboard intake) occurs as the intake normal shock intersects the body boundary layer and the resulting separated flow may be ingested Into the intake .

Other intake positions may suffer more or less severe regions of flow separation. For any side intake position there Is alway.*. the possibility of ingestion of a vortex shed from the bottom ’corner’ of the body. If the intake is shielded from the direct influence of aircraft incidence by positioning it under a wing or wing root strake (Fig 5—1 f ) then the risk of ingestion of a ’trapped* region of thickened body boundary layer or a shed vortex from the bottom of the body Is increased.

(f)

A dorsal position for the intake behind the cockpit canopy may have the advantage of lower aircraft radar signature (from ground based radar) and reduced hot gas ingestion for a VSTOL aircraft. However, it has the disadvantage of possible ingestion of the canopy wake or a vortex from the wing root or strake at sideslip and at supersonic speeds local Mach number in front of the intake will be in excess of the free stream value (Fig 5-1 g) .

i

34

\> Mo

The underbody position has undoubted advantages at incidence at both subsonic and supersonic speeds (Fig 5-1h). The onset of lip separation will be considerably delayed at subsonic speeds if the intake is well shielded by the body and at supersonic speeds the reduced local Mach number will Increase pressure recovery as incidence Is increased

FIG 5.1 (a)-(h) FLOW PATTERNS FOR BODY MOUNTED INTAKES

Fox & Kline (Ref 5-1) have defined boundaries of flow states in two dimensional ducts. The lowest boundary of 'first stall' defines an area In which the flow is always attached except perhaps for small separated flow areas near to the final area Aj. The next boundary of 'appreciable stall' defines an area in which the flow experiences transitory separation and the flow is no longer steady at the engine face. Tindell in Ref 5-2 uses an average wall slope and diffuser area ratio A2/AJ to relate more general diffuser shapes to Fox & Kline's boundaries (Fig 5-3). Tindell also uses the Himat and Grumman configurations as a basis for the calculation (by a finite difference and by a panel method) of the onset of duct separation as the basic geometry f these ducts is systematically altered to vary area ratio and average slope of the inboard wall.

AT _ _

02l0h r2j*

L

e ayg INB'O WALL

A)

OjtOR R21*

AID

1-35

A 07

19 8*

1 13

HIMAT

0 AS

3 19

11 5"

1 50

GRUMMAN

1 05

t ?0

30 0*

1 89

•use r2 for aaq

To understand and predict the performance of an intake installed on an agile tactical fighter. It Is necessary to build up the performance from simpler component parts. In the next section, first the characteristics of Isolated pitot intakes and then those with a compression surface are examined. For isolated fntakes, the build up Is Illustrated from the duct alone, the duct plus the intake at zero incidence and finally, the combination at incidence. This is then followed In section 5.3 by Illustrations of how the characteristics of an isolated intake and duct are altered by their integration with the flew fields around a body or body and wing combination.

2.5.2 ISOLATED INTAKES

2 5.2.1 Internal flow

(a) Fiow in the subsonic diffusers

Fig 5-2 Illustrates she nasic duct loss due to skin friction on duct walls as measured In a duct with a be 11 mouth intake that is sucked to give ?. range of throat Mach number. The small duct total pressure loss of Fig 5-2 Increases as (a) duct length Is Increased, (b) the duct becomes curved and changes cross sectional shape.

. SYMHETPlCIXL:W.' • OFFSET

FIG 5.3

COMPARISON OF THEORETICALLY DERIVED FAMILIES OF VERGE -OF- SEPARATION CHARACTERISTICS WITH TRANSFORMED CORRELATION OF KLINE

With the advent of more advanced fighter designs, tighter integration of the engine into the airframe sometimes requires highly offset diffusers (Fig 5-4). New methods of avoiding or controlling boundary layer separation are needed to achieve adequate performance. One of the most fundamental approaches is tn provide a favourable area distribution and centreline offset shape. Ref 5-2 .shows a study of these parameters. Three centreline shapes were chosen consisting of: 1) modest turning, 2) rapid turning at the exit, and 3) rapid turning at the entrance (Fig 5-5). These centreline shapes were analysed in conjunction with

35

three area distributions: 1) modest diffusion, 2) high diffusion at the exit, and 3) high diffusion at the entrance. The 3x3 matrix was analysed using flow computation methods; results are shown in Fig 5-6. Obviously, slow turning with high diffusion at the entrance is more favourable in terms of low boundary layer blockage and high total pressure recovery.

FIG 5.4 ADVANCED COMPACT DIFFUSER

oamst* ctNit*uHt shah inC DUIMU'C*.

CtNttM

ASM 0<SU'IUli0*i IA/A.-D

ta.ii

a 0)

it/U IX/SI

FIG 5.5 DIFFUSER CENTRELINE/ AREA DISTRIBUTION

FIG 5.6 EFFECT OF CENTRELINE / AREA

DISTRIBUTION ON DIFFUSER RECOVERY

The S duct flow may include some areas of separated flow followed by reattachment and certainly by the movement of the boundary layer from the outsfde wall of the first bend towards the Inside and the generation of some swirl at the engine face. If the rate of diffusion combined with the sharpness of the bends becomes too high, massive separation will occur downstream of the first bend with consequent unacceptably high total pressure loss and flow distortion at the engine face. The series of ducts shown in Fig 5-7 have constant length, a constant radius final bend, and a common area distribution. As the First bend is usually the primary source of engine face total pressure loss and distortion, the series of ducts successively decreases the extent of the first bend by decreasing its radius and canting the entry plane with the consequent reductions in total pressure loss that are shown.

For highly offset diffusers with separation in the turns, Cerlach area shaping (Fig 5-8, Ref 5-3) has been shown to minimise the extent of the separation. In turning ducts, a pressure gradient normally exists between the Inner and outer walls due to centrifugal forces. High pressures at the outer walls often are the cause of large separated regions. When the duct cross-section is shaped such that the velocity is increased at the outer wall and decreased at the inner wall, the pressure

0 08 tom 007

PRESSURE lOSS 0 06

4F V°05

Pto0 06

0 03 0 02 001 0

FIG 5.7 BASIC DUCT LOSS FOR S BEND DUCT DUE TO CHANGES IN FIRST BEND SHAPE

gradient between the walls can be reduced. Thus, the extent of separation can also be reduced without changing the average flow velocity.

MMMHQ-1 7 AS*C1IUT0*ta mu ihki 1/0-3 5

A2/D-M • -

FIG 5.8 GERLACH AREA SHAPING

Small airfoil shaped vortex generators in the diffuser upstream of a known separation point are still one of the better ways to minimise separation as shown in Fig 5-9. The diffuser separation characteristics must be known a priori, however, because the vortex generators are most effective when piaced just ahead *f the separated regions. Therefore, adequate diffuser analysis methods must be available to determine the location of separated regions before the vortex generators can be insta! led.

FIG 5.9 VORTEX GENERATOR DIFFUSER PERFORMANCE ENHANCEMENT

Active boundary layer control in the form of bleeding and blowing is often more effective than passive control. However, the cost may be higher in Intake bleed drag or increased compressor workload. The amount and distribution of bleed in the diffuser is critical. Too little bleed flow does not control the separation while too much is

i

not economical. The performance enhancement Is seen to be most beneficial in Ref 5-3 when bleed Is located upstream of the separation (Fig 5-10).

5.10 BOUNDARY LAYER BLEED

Computational methods are now coming into routine use which can calculate the flow through S ducts. Fig 5-11 shows a comparison between engine face total pressure contours as calculated by a NASA (Lewis) parabolised Navier Stokes program and measurements at high throat Mach number for an S bend duct with constant circular sections (Ref 5-4). Regions of sepur^'ed flow occur upstream of the engine face. Nevertheless, the general pattern of total pressure loss at the engine face Is reasonably we 1 1 predicted.

Cffsel/lenglh = 0.45, 0.385. Re = 3.90 x IQ6

FIG 5.11 COMPARISON OF CALCULATED (NASA LEWIS) & MEASURED (RAE 2129 S DUCT] ENGINE FACE RECOVERY CONTOURS FOR CIRCULAR SECTION S DUCT

(b) Combination of Intake and subsonic diffusers

The next element to consider is the effect of forward speed and replacement of the bellmouth by a representative intake lip. The magnitude of total pressure loss (other than skin friction) over the full forward speed range from zero to transonic is determined by the occurrence and resulting size of the separation at the lip and the magnitude of throat Mach number. In general at zero and low forward speed (except at very low throat Mach numbers) the Ingested stream tube will be larger than the Intake capture area (A0/Ac > 1.0) and separation will occur on the Inside of the lip If the lip Is sharp (Fig 5-la). The magnitude of the loss due to mixing following separation rises rapidly with Increase in throat Mach number. As forward speed Increases, the streamtube rapidly decreases in size for a given throat Mach number and when it becomes less than the capture area, (A0/Ac < 1.0), lip separation has disappeared and losses rapidly return to the skin friction level of the basic duct loss. In Fig 5-12 total pressure loss Is shown plotted as a function of Mt^ and of inverse capture ratio Ac/A0 rather than capture ratio so that the important static case of A0/Ac - <» (Ac/Ac - 0) can be included.

FIG 5.12 STRAIGHT DUCTS : VARIATION OF LOSS WITH THROAT MACH NO., INVERSE CAPTURE RATIO & INCl6ENCE

37

✓

At incidence, lip separation can occur at all values of Ac/Ac and not just at low values of Ac/Aq as at zero incidence. Fig 5-12 shows a progressive change in loss pattern as Incidence increases from 0* to 20‘ and 30 . At 20* the pattern of losses is similar to that at zero incidence but the thickened boundary layer downstream of the windward lip results in higher losses that are still however invariant with Ac/A0 at Ac/A0 > 0.7. For this particular lip shape at the highest value of Mt^ the effect of a small area of lip separation is Just evident as the losses no longer remain

constant at Ac/A0 > 0.7. At an incidence of 30* lip (Fig 5-12c) separation occurs at all values of Mtj, and this pattern of increasing loss i.s typical at all values of Ac/Ac above about 0.3.

There are two geometric parameters that can markedly affect the magnitude of total pressure loss due to lip separation. The first is lip contraction ratio CR (- Ac/Ath> and Fig 5-13a

illustrates the effect of changing CR from 1.078 to 1,25 at 20* Incidence and low forward speed

conditions. In this case, the differences in

performance between the contraction ratios are principally the result of whether or not separation has occurred. In the second illustration (Fig 5-13b) the streamtube size is smaller than the capture area and separation will have occurred for all contraction ratios; the differences in loss are due to the decreasing size of the separation region as contraction ratio decreases.

EFFECT OF LIP CONTRACTION RATIO ON TOTAL PRESSURE LOSS AT SUBSONIC SPEEDS

The second, which has an even larger effect on loss due to lip separation, Is the effect of intake scarfing or lip stagger. The upper Up turns the flow from the free stream Incidence so that the lower lip is shielded and If the stagger angle Is high enough, lower Up separation Is delayed probably until incidences of 50' - 60' are reached. Fig 5-K shows the variation of loss with Inverse capture ratio and throat Mach number for a 50’ stagger angle at 30‘ incidence. Losses fall continuously with Increase in Ac/Ac In contrast to

the loss variation of the unstaggered Intake (Fig 5-i2c).

VARIATION OF TOTAL PRESSURE LOSS V.’ITH INVERSE CAPTURE RATIO & THROAT MACH NO. FOR INTAKE WITH 50° LIP STAGGER

At low forward speeds with Ac/A0 < 1.0 for a

. aggered intake, more flow is sucked over the .ower lip than the upper one so the unstaggered intake with its symmetrical separation region suffers less loss as show** In the comparison of Fig 5-15. However, as forward speed increases From about M0 0.1 to 0. 2-0.3 (inverse capture ratios of 0.3-0. 6 approximately depending on throat Mach number), a cross-over point occurs and at higher increase capture ratios the favourable effect of lip stagger increases rapidly.

. Jr. 3

--

v*

J „

/

,4

M

h

_ _ l

r

a 0 -E

f 0-.

— o

- 5

- 0

0° STAGGER i N STAGGEREDJ

07 0-4 0 6 0 8 ' 0 07 0:4 0-6A 0-8

7A0

FIG 5.15 EFFECT OF LIP STAGGER AT LOW FORWARD SPEEDS

Thus, in practice, the staggered intake would require slightly larger or more efficient auxiliary Intakes to restore take off and low forward speed performance to the level of the performance of the unstaggered intake.

The influence of other geometry variables that have smaller effects on performance and ,he effect of active geometric variation that can result in large changes In performance aro dealt with in section 5.4.

At supersonic speeds, as incidence Increases from zero, the entry plane of a pitot intake Increasingly Inclines away from the vertical but the intake normal shock remains approximately normal to the free stream direction. As flow ratio is reduced from unity, the shock becomes detached from the leeward lip and moves upstream across the inclined capture plane. At the windward Up the flow remains unaffected until, with increasing spillage, the normal shock finally becomes r/*-; 'etely detached from the entry plane. When the shock is In this position (low values of Mt^) the variation of total pressure loss is then a function

l

38

✓

of capture ratio and throat Mach number as at zero incidence. When however the normal shock is impinging on the internal windward surface, the increase in loss is at a higher rate than at subsonic speeds. In these circumstances, the total pressure loss is now not only a function of capture area ratio and throat Mach number but also free stream Mach number because this determines the strength of the adverse shock and boundary layer Interaction on the inside of the windward lip. Thus, total pressure loss and engine face distortion is correlated over the full Mach number range from 0 to 2.0 by the use of the parameters Mth and M0/Mth as shown In Fig 0-16 rather than the parameters Mth and Ac/A0.

FIG 5.16 VARIATION OF LIP LOSS £ ENGINE FACE DISTORTION FOR AXiSYMMETRIC PITOT INTAKE AT FOR A MACH NUMBER RANGE FROM 0 TO 2 0 (REF 5. 5)

The basic definitions of components of intake drag have been outlined in Section II - Definition of intake performance parameters and description of intake flows - for both pitot and compression surface intakes. Evaluation ox' fore es on pitot intake cowls, from full flow down to the flow at which separation from the cowl lip is initiated, (Fig 5-lb), Is particularly amenable to calculation by full potential flow or by Euler methods. Full potential flow methods are very adequate to calculate drag (or thrust) due to supercritical flow development over the external surface of the cowl up to local Mach numbers of 1.3 - 1.4 (despite being isentropic In concept) at subsonic free stream Mach numbers and Euler (non Isentropic) methods are equally valid at both subsonic and supersonic free stream speeds. Fig 5-17 shows comparisons between force* measured and calculated on axlsymmetrlc cowls by these two methods over the Mach number range 0.4 to 1.8. The departure of measured and calculated curves clearly indicates the onset of cowl leading edge separation which is confirmed by examination of cowl pressure di str ibut ions .

0

‘005

-0-10

-0-15

Co

°C0Wl

-0-20

$

f

0 0

/;

±_

5ai

0_4

i

/I

i A^ 07 0;4 06 0;8 10 ^

°v

0

A

> 0 Oj

/

06

/

/

*0/

/Ac

t.R.

© 1-1771 • V093J

MEASURED

- FULL POTENTIAL 1

- EULER i

CALCULATED

NACA-1-85-100

COWL

oB,*-y °,6 g VY

-0-05

C°cowi -0-15-

-tm

-0-25-

/|

P*

-

A

1%

0

o°-

?

2 0

4 0

6 X

W

8 10

~o —

i

°C0Wl

0

. ?

Mn=V5

V'l

1

i

L

“0 15-

-0 20-

1

1

X

FIG 5.17 COMPARISON OF MEASURED £ CALCULATED AXISYMMETRIC COWL PRESSURE FORCE AT SUB & SUPERSONIC SPEEDS (REF 5-6)

FIG 5.18

COMPARISON OF MEASURED £ CALCULATED PRE- ENTRY DRAG FOR AXISYMMETRIC PITOT INTAKES AT SUBSONIC £ SUPERSONIC SPEEDS

(REF 5.6)

Cowl wave drag at supersonic speeds can calculated using the Euler equations or by method of characteristics. A useful digest axisymmet ric cowl wave drag based characteristics calculations and measurements shown in Fig 5-19.

be

the

of

on

is

Pre-entry drag can also be evaluated by full potential flow or Euler mothods and agroes well with measured values (Fig 5-18). If the intake lip is sharp then the classic one dimensional momentum evaluation of pre-entry drag (also shown on Fig 5-18) is sufficiently accurate for most purposes.

i

FIG 5.19 DRAG OF AXISYMMETRIC COWLS (REF5-7)

2,3 Intakes with Compression, Surfaces at Subsonic & Supersonic Speeds

The design and performance of intakes for long range supersonic cruise aircraft, whose compression surfaces only operate over a small range of incidence and cideslip, is considered in Section IV of this review. This focusses on both external and combined compression intakes with complex variable geometry and multi -bleed systems. However, for tactical fighter aircraft, the emphasis is on simpler designs having only one or two boundary layer bleeds, external compression only and the variation of probably only one hinged or translating surface. Greater emphasis is placed on compromising the supersonic design to obtain good performance over the full Mach number range from zero upwards.

One of the more fundamental problems in supersonic intake design is matching the intake and engine flow rates at off-design conditions. Aside from the engine demand, the intake may have to provide air for an environmental control system (ECS), bypass system, bleed systems, etc. Fig 5-20 shows a typical flow rate breakdown as a function of Mach number. Notice that the most spillage occurs at transonic flight speeds. This transonic flow mismatch is one of the primary factors which complicate supersonic intake design.

The range of manoeuvre, together with the extreme range of hot and cold day operation required by combat aircraft, often leads to the need for matching Inlet and engine airflows other than solely by variation of the geometry of the compression surface. Thus, aft spill from the subsonic diffuser, either direct to the free stream or by bypassing air around the engine to a base area or ejector nozzle, is required and adds further complication of operation and control of moving surfaces (Fig 5-21).

FIG. 5.21 AFT SPILL & ENGINE BYPASS

U_ihlA _ At subsonic speeds

One oi the purposes of variable compression surface geometry is to provide the enlarged thrort area required at subsonic speeds to pass the engine flow without throat choking. Ideally, this entails the complete collapse of the wedge or axi symmet r ic centrebody. In practice, this is most ccnveniem ly approximated when employing a rectangular intake by reducing the second wedge surface to the free stream direction or slightly below this if the actuating mechanism beneath this surface will allow it (Fig 5-22). Although the data shown in 5.2.1 is strictly for pitot Intakes, the flow states discussed there will In general apply to intakes with 'collapsed' compression surfaces. The inward camber and sharper lips of a compression surface Intake will however increase losses and engine face flow distortion at zero and low forward speeds and there will often be a need for large auxiliary intakes to obtain the required engine flow at high intake efficiency for take off and low speed operation (Fig 5-23).

- SUBSONIC

- - SUPERSONIC

FIG. 5.22 TORNADO INTAKE RAMPS IN SUBSONIC & SUPERSONIC SPEED POSITIONS

FG 5.23 AUXILIARY INLET OPERATION

FIG. 5.20 INLET/ENGINE AIRFLOW MATCHING

Numbers of auxiliary door arrangements for the BAe Tornado Intake are shown In Fig 5-24 without Intrusive doors in Configurations A & D and with internally intrusive doors In the other configurations. Their performance under static conditions is summarised in the bar charts and engine face distributions of Fig 5-24.

ifWf.i 0

FIG 5.24 EFFECT OF VARIOUS AUXILIARY INTAKE GEOMETRIES ON PERFORMANCE OF A COMPRESSION SURFACE INTAKE AT STATIC CONDITIONS

The form of the rectangular intake, with a collapsed second ramp and endwalls to the compression surface that are swept from the front ramp tip to the cowl lip, is then very akin to the staggered pitot intakes discussed in Section 5.2.1. Thus, if the intake lip is not sharp, performance of the wedge compression surface supersonic intake, (when suitably orientated with the first wedge leading edge horizontal) at high incidence at subsonic speeds, should be adequate.

No such happy coincidence occurs when using axisymmetric centrebody intakes. These pose larger problems for obtaining adequate throat area by centrebody translation. Indeed, on the Fill aircraft (Fig 5-25), the second cone angle consists

Fit. 5.251/4 AXISYMMETRIC (Fill) INTAKE WITH TRANSLATING & COLLAPSING CENTREBODY

FIG 5.261/2 AXISYMMETRIC (MIRAGE) INTAKE WITH TRANSLATING CENTREBODY ON CURVED TRACK

of interleaving elements that allow the centrebody to collapse (it also translates) to obtain adequate throat area. This would however be very difficult if not impossible if the intake was an isolated full axisymmetric nacelle. Another solution to this problem, only possible with the Installed half axisymmetric centrebody intake, is to translate the half conical centrebody on a circular track to obtain a larger throat area, as exempli fie-1 by the well known 'mouse' in the intake of the Mirage series of aircraft (Fig 5-26).

25,2.3,2 At supersonic speeds

As flight Mach number increases, multiple compression surfaces and variable geometry may be employed to improve performance, but the increase in complexity also Increases weight and cost. l’he aircraft designer must look at the effect of intake design on total aircraft performance before selecting an intake type. In some cases, the designer may make a sacrifice fn recovery performance in order to get the lower weight and cost of a simpler design.

A family of curves depicting the maximum pressure recovery attainable through a series of oblique or conical shocks and a terminal normal shock is shown in Fig 5-27 and reflects the ideal performance potential of axisymmetric supersonic intakes.

creasing the number of shock waves yields higher pressure recovery values. Thus, an isentropic compression surface generating an infinite number of Mach wavas yields the maximum recovery.

FIG 5.27 CONICAL SHOCK RECOVERY

Fig 5-27 also shows the performance of the Lockheed F-104 intake which has a single fixed geometry cone designed for Mach 1.8. The recovery drops off rapidly at off-design flight conditions.

FIGHTER AIRCRAFT INLET TOTAL PRESSURE RECOVERY

VARIABLE GEOMETRY INLET DEVELOPMENT (REF 5.8)

41

An interesting trend In US variable geometry development Is Illustrated in Fig 5-28. Attempts at Improving performance over the fixed geometry F-104 were made in the F-105 and F-4. Acceptable performance was obtained over a wider range of Mach number, but the full benefit of variable ramps was not realised until the advent of the F— 1 11, F-14 and F-15 aircraft .

FIG 5.29 COMPARISON OF ADVANCED 2-0 INLET PERFORMANCE WITH CURRENT .NLETS, MATCHED AIRFLOW a=0? 0 = 0°

Some of the recent progression in intake design efficiencies are illustrated in Fig 5-29. Although these are for installed intakes, at cr - 0 - 0, the results should be very close to that of isolated intakes .

FIG 5.31 OBLIQUE SHOCK- BOUNDARY LAYER INTERACTION

Cowl drag associated w'th a given amount of external compression can be determined at an early stage in the design process from the requirement of oblique shock attachment on the inside surface of the cowl lip. The consequences of selecting an internal cowl angle (and hence external cowl angle and thus external drag) to Just give an attached undersurface shock on maximum shock pressure recovery that can possibly be attained (using an idealised isentropic compression surface) is shown in Fig 5-32. Thus, at Mach number 3, a cylindrical undersurface that would be associated with a very low cowl drag is limited to a shock recovery of 0.72, whereas a moderate cowl drag with an undersurface angle of 12* would have an upper limit to shock recovery of 0.83.

A successful bleed design is one that gives large increases in pressure recovery for relatively small bleed flows. When this is achieved, pressure recovery values that differ from the theoretical shock recovery by only the duct skir. friction loss are obtained (Fig 5-30) for bleed flows in the region 4-6K.

FIG 5.30 EFFECT OF VARIATION OF BLEED FLOW ON PERFORMANCE OF A DOUBLE RAMP SUPERSONIC INTAKE

The drawbacks inherent in using large amounts of external compression at supersonic speeds are well known. It results in high external drag due to Urge flow turning from the free stream direction and the development of long boundary layers on the compression surface that are subject to high adverse pressure gradients which cause them to thicken and/or separate, particularly at mul t 1 -compression surface intersection points (Fig 5-31).

FIG 5.32 MAXIMUM SHOCK PRESSURE RECOVERY FOR EXTERNAL COMPRESSION WITH A GIVEN COWL EXTERNAL ANGLE (REF5.7)

The Increase in surface wetted area in front of the capture plane can be linked to the Increasing number of supersonic ramps or cones and hence increasing shock pressure recovery (Fig 5-33).

o Kecrangutar

i Reef angular romp no swept sidewalls > Axisymmetric conical centrebody • Halt axisymmctric conical cenfrebody with swept ewjvalls

Shock-on-lip Mach No. Msoi 2.5

^3

Double ram?

0.5

FIG.

5.33

0.6

J Axisymmctric

0-2 0,8 0.9 H»lf Mlsyamih-ic

Shock pressure recovery with swept endwall

VARIATION OF WETTED SURFACE AREA FOR EXTERNAL COMPRESSION INTAKES (REF.5.7)

v

42

This diagram also emphasises the large differences In approach surface area between square and axi symmet ric Intakes and the Increase due to enclosing the shock system of the square intake between swept endwalls. It is no surprise that rectangular Intakes invariably feature boundary layer control on the compression surfaces whereas single or even double cone centrebodies often do not .

The realisation that cowl internal angles do no* need to be at the same Inclination as the fina. ramp of the compression surface, together with close integration of the design uf the bleed on the compression surface just downstream of the entry plane, with the shocks from the low angle cowl und^ ^urfac? , has led to development of optimum low drag exit-.nal compression intakes. The resulting wide slot bleed unchoked at its entry also has the advantage that matching initially occurs by spillage through the bleed (Fig 5-34) and does not Involve conventional subsonic forespillage over the cow* (Fig 5-35) with its attendant high drag and possible shock instability.

FIG 534 WIDE THROAT BLEED OPERATION

r-Compression Surface Spillage - supersorre forespill \ -Subsonic A

FIO 5.35 SUPERSONIC &. SUBSONIC FORESPILL

At incidence, if the ramp compression surface is horizontal, then the variation of shock loss can be calculated until the ramp shock(s) become detached. Fig 5-36 shows a comparison between calculated and measured pressure recovery for a two ramp intake with a wide slot throat bleed.

EFFECT OF INCIDENCE ON PRESSURE RECOVERY S SHOCK WAVE PATTERNS FOR A HORIZONTAL DOUBLE- RAMP INTAKE (REF 5.7)

Other purely measured values of pressure recovery referenced to pressure recovery at zero Incidence for a wider range of conditions are shown in Fig 5-37.

-riis

n

*

n

B

□

Q

A—

”

/

~~

£

M0 ^501 □ 2 0 2-45 RAE

*

OL

Kx-

i 2 0 2-7 ONLRA v 2-4 2 45 RAE

M

o 2-7 2 7 ONERA

8 -

-l

° ‘

j, 10 1

FIG 5.37 VARIATION OF PRESSURE RECOVERY WITH INCIDENCE. FOP TWO-RAMP INTAKE

If the intake Is yawed, the variation of shock recovery Is no longer readily predictable and again

rig

rocov<

PR,, O’-

pfy

5-38 shows measured variation of pressure ery .

L 1 * 1 1 f. ! ! 1 1

PRPtP

q

• -4 M !

P

p.

i ! Mo Msol

W i ' V. I „ 7ft OAF

07-

J

p'i. ‘l\l '* A 2-0 2-7 ONERA ' '''[ p - ° 2-/i 2 45 RAE

Li |i

t

1 I v ] 0 2 7 2 7 ONERA

0 5- C

r it i

_ j

rfiTT

• 2 A t

- e

' to 1? U 16 15

FIG 5.38 VARIATION OF PRESSURE RECOVERY WITH SIDESLIP ANGLE FOR TWO- RAMP INTAKE

In this case, performance is sensitive to changes in endwall shape and quite small changes in configuration can result in relatively large changes In pressure recovery (Fig 5-39).

FIG 5 39 EFFECT OF SMALL CHANGES IN ENDWALL SHAPE ON PERFORMANCE OF YAWED RECTANGULAR INTAKE

With axi symmet r i c external compression Intakes, the flow in detail at Incidence is complex but the haste pattern of shock angle variation is akin to that of a pitot Intake. The conical rArobody car. be regarded as pivoting about Its apex with shock angles changing much less than cone surface angles relative to free stream direction. As incidence increases, the cone shock moves into the windward lip and the leeward lip moves away from it. When this happens an increasing proportion of the flow Is compressed through a normal shock only and tho result is a sharp fall off *n pressure recovery. Thus, tho change in pressure recovery with incidence is largely Influenced by whether or not tl.c cone shock at zero incidence is on or well outside the cowl lip, as illustrated In Fig 5-40.

I

CRITICAL

\ :NE SHOCK WOUTSIOt COWL \ \llP M 0C=0*

\

COWL SHOCK oil1'

OS 1MSIOE cowi up *1 a.o*

* i :NAl »« _£ OF 1SCNTBOPIC SUBFdC!

F IQ 5 40 ' 7 •— ’ ‘ 6 *

EFFECT OF INCIDENCE ON PRESSURE RECOVERY OF AXISYMMETRIC FOREBOOY INTAKES(RF.F5.7)

A second influence on pressure recovery enmes from the tendency for the boundary layer on the centrebody to be swept up to the leeward side. Thus, a long centrebody with a high shock on lip Much number MS0L will have a worse performance than one with a iow (Fig 5-41). This trend in

principle is of courst in the opposite sense to that of Fig 5-40.

INFLUENCE GF FCRE80DY LENGTH ON PRESSURE RECOVERY AT INCIDENCE OF DOUBLE- CONE INTAKES (REF 5.7)

There is no simple way of predicting maximum flow ratio for axisymmetrlc intakes at incidence. Fig 5-42 shows some measured values referred to the zero incidence value for slngie cone, double cone and isentropic centrebody intakes at M0 1 9.

FIC 5 42

EFFECT OF INCIDENCE ON MAXIMUM FLOW OF AXISYMMETRIC FOREBODY INTAKES (REF5.7)

LAAlmaKerfllrfuiiM Integrailpn

It was discovered early in jet aircraft design that allowing the forebody boundary layer tc be Ingested into the Intake imprs^d a serious loss of recovery at the compressor. The nub of the problem, as has been seen (Fig 2-7 Section 2) is the Interaction of the boundary layer with the pre-entry pressure rise which is incurred in the process of producing a relative retardation cf airipeed from the flight velocity towards that required at inlet to the engine. Broadly speaking, the severity of the problem is greater the higher the flight speed and the presence of shock waves at supersonic speeds adds a special dimension to it. If cite boundary

layer separates or comes close to separation, the effects are particularly adverse and generally not res; Meted U' * lowering of 'tan pressure recovery: other adverse le&tures include total -pressure distortion and flew instability (Section 2).

Rcrroving the boundary layer at some stage from the intake provides an escape from, or easement of, the difficulties This is done by means of bleeds or diverters. The term ’bleed* denotes a separate duct which leads away the boundary layer. The term 'diverter’ implits that the intake stands off from a particular surface, a P owing the boundary layer on that surface to escape down the int viced i at * channel. in either case the boundary layer removed from the Intake usually becomes a pari of the aircraft system, that is to say it represents ar additional Item in the aircraft drag but the effect of the increased pressure recovery invariably outweighs the drag penalty.

Common forms of bleed and diverter are illustrated in Fig 5-43. biverters may be of the step type (a) or channel type (b), (c). A step diverter is a useful form in the .ng root of a subsonic aircraft, because the forward extension allows a goou wing-root profile to be preserved. Step diverters are not generally recommended for supersonic airc lft, however, because fresh boundary layer initiated on the surface of the diverter may itself produce most of the interaction l^ss of the original longer boundary layer. Channel diverters p re suitable for both subsonic and supersonic application, provided thai a reasonably aerodynamic ’prow' shape car* be obtained between the Intake and the boundary- layer surface. The recommended width for step diverters is abcut one and a half times the thickness of the boundary layer, when this is undistujbed by the presence of the intake, and for channel diverters about one such boundary layer thickness.

(h) Channel d-verter - subsonic intake

(c? Channel diverter - supersonic intake

FIG 5.43 EXAMPLES OF STEP & CHANNEL DIVERTERS

Boundary layer bieed systems can become more complex at supersonic speeds as on the F-lll aircraft (Fig 5-44) and need more careful design to ensure their successful operation over all engine flow rates and aircraft attitudes that will be encountered

Tactical- aircraft are required to he manoeuvrable at *iubsonlc, transonic and supersonic conditions without giving up good subsonic cruise efficiency. Consequent ly, proper integration of the engine intake with the airframe Is of paramount interest. The range of incidence and sideslip, although reduced at supersonic speeds from the values outlined in tne Introduction, still remains wide compared with those required by the high altitude supersonic cruise vehicle. Within these reduced ranges, at supersonic speeds for the installed Intake, not only Mach number and flow direction will vary in the approach flow field to the intake but also total pressure due to the presence of an upstream curved and In some cases detached shoe*: from the body nose.

location of boundary layer and cone probes used to define flow fields approaching the intakes.

. want mm u uw-mom :«*m Mini

FIG 5.46 FOREBODY FLOW FIELD MODEL INSTRUMENTATION

Design for optimum al rframo-inlet integration has

the following specific goals:

1 Minimise approach flow angularity with repect to inlet leading edges

2 Deliver uniform, high pressure recovery flow to the inlet entrance at velocities equal to or below free stream conditions

3 Prevent boundary layer ingestion by the inlet

4 Reduce the probability of foreign object or Lot gas ingestion to acceptable levels

5 Minimise the potential for flow field Interference from weapon carriage/firing, landing gear deployment, fuel tanks, pods, pressure probes etc.

Flow fields are characterised by vector plots or by lines of constant local angle-of- incidence, a^, or local angle-of-:. idesl ip, 0l- The vector plots give an overall impression of flow angularity whereas the anc* 0L contours provide more precise definitions. Flow field variations for relatively high manoeuvre conditions at Mach 2.2 are examined to aid understand* ng of the limits of performance.

Fig 5-47 shows a vector plot depicting combined c*l and vectors around the BASELINE body shape at MQ - 2.2, cv0 - 15 . Flow angularity is highest near the body and there is a significant sideslip flow condition in v.he region of the inlet inboard surface .

Similarly, the airframe-intake design should minimise any deleterious effect on the design/operation of lifting surfaces, landing gear and avionics and how they integrate with aircraft structures .

Because of the many factors which influence airFrame flow fields, it is expedient to discuss a sot of coordinated data so that the effect of isolated parameters may be illustrated. To this end. the series cf fighter conf igurai ions shown in Fig 5-45 (Till lor-M&te Program, Ref 5-10 to 5-14) will he used to isolate some of the primary effects of fuselage shaping inlet location and how the inlet conf igurat io.t interacts with location in sections 5.3. 1-5 .j. 5.

:au»

«SC l*W !*

tmiKOMwa

viNG.smriNe

kmc-skuk;

FIG 5.47 BASELINE FOREBODY VECTOR FLOW FIELD; M0=2-2, 0Co=15°, I30 = 0°

A mor; careful examination (FI 5-48 and 5-49) shows lines oT constant uL and 0L for the BASELINE shape and the SQUARED shape. There Is little difference in oL distributions, but the flow in the region of the lower Inboard sideplate at sideslip conditions varies measurably between the shapes. The maximum value In the region of the intake Inboard sjdeplate Is least Tor the BASELINE shape U?L ~ 8.6’) and greatest for the SQUARED shape (0. -9.2’).

FIG 5.43 BASELINE FOREBODY FLOW FIELD ANGULARITY MAPS, M0= 2 2, 0Co= 15°, [30= 0°

FIG 5 45 TAILOR-MATE AiR FRAME INLET CONFIGURATIONS (REF 5.1C -5.14)

L3.-JL1 — r_ OS cl age flow fields lor sid? mounted

.Qi.tflR<L£

Wind tunnel tests wider the Tailor-Mate program were conducted in the Arnold Engineering Development Center (AEDC) Propulsion Wind Tunnel 16 foot test sect Ions.

Fig 5-46 depict^ the forebody cross-sect ion shapes Investigated for side-mounted intakes and shows the

FIG 5.49 SQUARED FOREBQDY FLOW FIELD

ANGULARITY MAPS; 22, O^15fl3o=0°

45

Figs 5-50 and 5-51 show 0^ contours associated with a combination of aircraft pitch and yaw for baseline and square bottom fuselages at M0 - 2.2, oD - 15*, 0 - 4*. The leeward (right) side 0 ^ plots for both fuselage shapes at 0O - 4* are similar in appearance to the 0O - 0 plots but with approximately 2* in 4" 0Ladded to each measurement point. The flowfield differences may appear to be minor but inlet data will show a significant effect on performance.

FIG 5.50 BASELINE FOREBODY FLOW FIELD IN SIDESLIP. M0=2-2, Oo=15°,

SIDESLIP; M0= 2-2. a0=15° fi0 = 4°

systems and bleed on the perforated side plates was est inuted from plenum and exit static pressure measurement s .

Intake instrumentation was provided at the left-hand compressor face to measure totrl pressure recovery and both steady state and dynamic Ir.let flow distortion by use of 40 combination steady-state and high response tcta? pressure probes. Additional instrumentation on the opposite (right-hand) side of the double Intake system, consisted of total pressure rakes and sta: 1c pressure measurements through the duct as noted in the figure, in the following discussion, "positive sideslip" is defined as that condition where the inlet under consideration is on the leeward side of the Fuselage.

Figure 5-53a,b takes two intake data points for the BASELINE forebody at M0 - 2.2 (oQ *» 0*. 20 ) and depicts the progression of flow through the Intake In order to shed some light on the variation in levels of performance experienced with increasing oQ. At oD - 0 total pressures are all high at the inlet cowl lip and intake throat or aft ronp station. Pressures farther down the duct indicate that some low energy flow exists in the upper part of the duct, but this situation shifts rapidly, with the low energy flow showing up only in the lower portion of the compressor face. Since no lew energy flow Is in evidence in the lower part of the duct at the first three data stations, it must be supposed that the flow defect originated In the region of the eft ramp leading edge (behind ti.e throat slot) and that static pressure gradients in the duct were sufficient tc move this region to the lower portion of the compressor face.

25.3,2 Performance of a rectangular compression

class. J niaLs_ag_Hi£_aldfc .0 La. .f.use U&e

The variable geometry ••ectangular intake shown In Fig 5-52 was tested with both the baseline ana squared forebody shapes. It featured a variable first ramp ar well as variable second and third ramps for efficient compression at supersonic condi t ions .

felMCOt BUflJ

to pg te

UUllW MAM* DWO

FIG 5.D2 SIDE-MOUNTED INLET CONFIGURATION AND INSTRUMENTATION

It had a long subsonic duct (L/D - 5.23) with a low diffusion rate. Boundary layer control on the ramps was provided by performed bleed on the compression ramps anJ on the side plates. The variable gap offset between the th‘rd ramp and aft rnnp allowed a combination of flow bypass and boundary lay»r control at the throat region. Throat bleed flow was measured with flow metering

At or0 20 , flow at the inlet Up is relatively uniform, but with a lower total pressure than In the previous case. In the upper throat region, low pressures are measured, suggesting a greater I low interference of the flow with the aft ramp leading edge. More low energy flow and flow separation in the upper part of the duct are in evidence midway in the duct. In this case, rotation of the low energy flow through the duct to the inboard and lower part of the compressor face appears to be quite evident .

FIG 5.53b EASELINE INLET TOTAL PRESSURE MAPS; M0 = 2-2, 00=20? Bo = 0c

46

Tfc* n*xt F'gure (Fig 5*541 shows an example of the effect of sideslip at a0 ~ 15' cn progression of ftow through s right -hand intake integrated with the baseline forebody. At 0O - 0 , there is evidence from the |»p rakes of inboard s.'deplate Ivedlng edge flow interference. This defect is still in evidence at the inlet throat, and further on oovn the Suet the low energy flow region extends ove- an extensive por* ion of the inboard s’de of tho duct. Once again, at the - onpressor face, the low energy air appears to have emigrated downwards, but flow conditions are relatively uni Toro. With the intake exper fencing 4* leeward sideslip, there is now extensive flow separation on the inboard sfdeplate at the cow! tip. The separation and its effects spread sign! f ivant ly at the throat, and continue o spread in the duct. As before, the region of lowest energy air rotates downward so that, at the compressor face, the very lowest pressures have moved to the bottom and the only remnant of the high pr«$su.*e flvw from the lower throat rcg’on now exists in the upper region of the compressor face.

(b)

%.4'

FIG 5.54 INLET TOTAL PRESSURE MAPS WITH BASEL INE FUSELAGE M0 = 2-2, 0Co=15° (a) T30= 0°, { b) B0= 4°

figure 5-55 gives an extmple of throat flow variations associated with the different forebody influences. Influence of the forebody is seen in all three cf the integrated configurations, but Is by far t he most prominent with the SQUARED forebody, where a substantia) region of flow

un.:« Hstott ikii iwutit i»n

WWII «irtli»W

'xon fVJtiMJ j«.ct

FIG 5.55 SIDE-MOUNTED INLET THROAT PRESSURE CONTOURS. M0 = 2 2, 0Co= 15°, (V 0°

separation dominate* the lo**r inboard region of the throat. Thus, It Is seen that the flow field difference# between the BASELIKE and SQUAXED fuselage shapes do indeed make a difference In intake performance.

?5.3.3 Performance of rxi symmetric . halfagnt

The previous discussion and examples centred or. the horizontal *a«p intake. The choice of rectangular versus axisyroctrlc or vertical versus horizontal ramps depends upon the Moch number rnnge and manoeuvrability required. At low angles of Incidence, axlsyrrftetr le intakes can provide better pressure recovery tivan rectangular intakes at supersonic Mach numbers because of the sculler

fundament# 1 lotsec associated with coin cal shocks. Also, axisymaet r ic intakes tend to be much lighter because of the high load carrying capability of "hoop* tension. However, ax i symmetric Intel s and vertical raop intakes exhibit nonsyemetry of the flow at high angles of incidence which result in lower pressure recovery. The manoeuvring

horizontal ramp Intake is able to deflect the

one oir. I ng flow and retain Its quasi two-dimensioiial character. In addition, variable geometry to

control shock position, spillage, and recovery is easier to accomplish. Variable geometry in axisymmet r ic intakes is usually confined to spike translation in single or twin cone configurations.

The Ref 5-14 prograwne employed ha 1 f-axi synvwt r Jc intakes as well as rectangular intakes, offering a comparison of their relative performance in the s ide-mou tied (and wing-shielded) configurations. Figure 5-56 shows the side -mounted axi synrnet r ic intake with its instrumentation. These Intakes were tested with the forebody having the ROUNDED lower shoulder shape (see Fig 5-46) so the flowfield approaching mis intake was at least as good as that ahead of the BASELINE rectangular Intake .

INLET CONFIGURATION & INSTRUMENTATION

Flow progression through the side-mounted he 1 f-cone intake is shown in Fig 5-57 et aQ - 10 . In this case, a massive flow separation originates in the upper part of the intake throat due to the oversped

FIG 5.57 SIDE-MOUNTED HALF AXI-SYMMETRIC INLET TOTAL PRESSURE CONTOURS; M0=2 2, 0o=0°

47

condition »nd hone* strong normsl shock in that region. Further -!nwn the duct the flow in this upper region Is still sepsrmted. but has been joined by * separated flow region In the lowr par; of the .hroat. adjacent to the spike. At the cospressor face evidence of separated flows is obvious in the low pressure regions a.id large distort ion.

Based on this test data. the side-'aounted half-axlsyussetrlc Intake would appeal to be a poor candidate for highly manoeuvrable fighter aircraft. Yet several current aircraft systems employ this intake effectively, and it can be attractive Trom structure, system drag and reduced observables points of view. Design modifications to provide optimum compression surface angles, more efficient boundary layer removal and/or better variable geometry control can improve axisymmetric intake performance, possibly without giving up some of the Important advantages noted. Care would also have to be taken with this type of an Intake to design th. divert er/cidep!at< carefully enough to avoid sideplate leading edge separation.

?5 3.4 Performance of a pl'-al- Lmm Pn thejUd.t— iC a fuselage (Ref 5-15)

Although detailed body flow field data Is not available In the following examples, the main influence of the Increase in local angle of Incidence due primarily to upwash around the body, particularly In the region of the upper shoulder, shows up In the performance of pitot intakes. The local variation of sidewash Is probably not important for a pitot Intake with blunt Up. Comparison of isolated pitot intake recovery with recovery measured on an installed pitot Intake