Evaluation of laminar flow control systems concepts for subsonic commercial transport aircraft

CONTENTS OF POOR QUALITY Section

Page

l.O SUMMARY 2.0 INTRODUCTION 3.0 SYMBOLS & ABBREVIATIONS 5 4.0 MISSION DEFINITION 11 5.0 CONFIGURATION DEVELOPMENT 5.1 CONFIGURATION STUDY GROUND RULES 15 5.2 CONFIGURATION INTEGRATION STUDIES 18 5.3 STUDY I INITIAL BASELINE LAMINAR FLOW & TURBULENT FLOW AIRCRAFT . 20 5.3.1 General 20 • ° 5.3.2 Aero dynami cs 5.3.3 Structural Design 5.3.4 Suction/Manifolding System 37 5.3.5 Propulsion System 5.3.6 Surface Contamination Prevention 5.3.7 Performance - Initial Baseline Aircraft 45 5.3.8 Economics - Initial Baseline Aircraft 5.3.9 Study Conclusions - Initial Baseline Aircraft 5.4 STUDY II - FOUR VS THREE PROPULSION ENGINES 61 5.5 STUDY Ill - FOUR VS TWO WING MOUNTED SUCTION ENGINES 65 5.6 STUDY IV - FINAL AIRCRAFT CONFIGURATIONS 66 5.6 l Laminar Flow Aircraft 70 6.1.1 Aerodynamic Considerations 6.1.2 Wing Structure 6.1.3 Wing Leading Edge Protection 6.1.4 Suction/Ducting System 5.6 2 Turbulent Aircraft Comparison 5.7 OPERATIONAL CONSIDERATIONS 100 5.7.1 I00 FAA Regulations 5.7.2 Airline Comments I01 APPENDIX 5-A - DRAG ESTIMATION 6.0 ECONOMICS ANALYSIS 6.1 AIRPLANE COST ESTIMATING 109 6.2 COST SUMMARY 6.3 SENSITIVITY OF DOC COMPARISONS TO FUEL PRICE k v

CONTENTS (Continued) Section AERODYNAMIC DESIGN & ANALYSIS 7.0 119 7.1 119 AIRFOIL DESIGN & SUCTION REQUIREMENTS 7.l.l 119 Design Guidelines for LFC Airfoils 7.1.2 Airfoil Analysis 7.1.3 Suction Requirements 7.1.4 Variation of Suction with Airfoil & Flight Parameters 7.1.5 Effect of Loss of LFC 7.1.6 133 Off Design Conditions 7.1.7 133 LFC Airfoil Design Study Summary 7.2 NUMBER & LOCATION OF SUCTION ENGINES 135 7.2.1 135 Effect of Engine Noise on Allowable Engine Location 7.2.2 Number of Propulsion Engines 7.3 141 ASPECT RATIO, TMRUST & WING LOADING ANALYSIS 7.3.1 141 Wing Configuration 7.3.2 Aircraft Sizing - Base Case 7.3.3 Aircraft Sizing - Upper Surface Suction Only 149 8.0 AERODYNAMIC TESTING 155 8.1 AERODYNAMIC SMOOTMNESS TESTING 155 8.1 .l 155 Model Description & Installation 8.1.2 157 Test Specimens 8.1.3 158 Testing Procedure 8.1.4 Test Results 161 8.2 SWEPT WING WIND TUNNEL TESTS 8.2.1 166 Model Design 8.2.2 Model Description & Installation 8.2.3 Model Instrumentation 8.2.4 Checkout & Calibration with Nonporous Panels 8.2.5 Calibration & Testing with Porous Surfaces 8.2.6 198 Boundary Layer Stability Analysis 8.2.7 202 Conclusions - Wind Tunnel Testing 8.3 REFERENCES - AERODYNAMIC TESTING 204 vi

CONTENTS (Continued) Section 9.0 STRUCTURES 9.1 WING BOX DESIGN 9 l.l General Requirements 205 91.2 Initial Concepts Considered 206 91.3 Preliminary Design for Strength 91.4 Flutter Penalty 211 91.5 Evaluation of Initial Concepts 91.6 Aeroelastic Penalty 221 91.7 Concept Selection .

91.8 Aspect Ratio Study 226 91.9 Revised Wing Design Details 9 l.lO Wing Box Structure Conclusions 9.2 POROUS SURFACE SELECTION & LFC PANEL DESIGN 9 2.1 Initial Survey of Possible Materials & Construction 235 92.2 Aerodynamic Smoothness Test Panels 242 92.3 Pressure Drop Through Combined Porous Layers 9 2.4 Porous Surface Selection.

92.5 Thermal Aanalysis of Dissimilar Materiais 25O 92.6 LFC Panels Arrangement 254 92.7 Superplastic Formed/Diffusion Bonded LFC Panel 254 9.3 STRUCTURAL TESTING 9 3.1 Porous Materials - Strength Properties 93.2 Porous Materials - Fatigue 93.3 Impact Testing 9 3.4 Rain Erosion Testing 266 9 3.5 Compression Crippling Tests 9 3.6 Panel Joint Strength 93.7 28O Panel Joint Fatigue 9 3.8 Panel Development Testing 282 9 3.9 Structural Demonstration Component 9.4 WING LEADING EDGE STRUCTURE INTEGRATION 290 9.4.1 Leading Edge Design - Phase I 9.4.2 Leading Edge Design - Phase II 9.4.3 Leading Edge Fabrication .

9.4.4 Improved Leading Edge Glove 9.5 SWEPT WING WIND TUNNEL MODEL - POROUS PANELS 9.6 CONCLUSIONS & RECOMMENDATIONS - STRUCTURES .

9.7 REFERENCES - STRUCTURES 3O5 vii

CONTENTS (Continued) Section

Page

I0.0 SUCTION SYSTEM lO.l INTRODUCTION I0.2 REQUIREMENTS 307 I0.3 POROUS SURFACE CHARACTERISTICS I0.4 DUCTING I0.5 SUCTION PUMP SYSTEM lO.5.1 Compressor Design 319 lO.5.i.l Pressure Ratio I0.5.1.2 Compressor Sizing 321 I0.5.1.3 Compressor Type .

I0,5.1.4 Compressor Efficiency I0.5.2 Power Unit for Compressor 328 I0.6 SUCTION ENGINE ARRANGEMENT lO.6.1 Number of Suction Engines 337 I0.6.2 Location of Suction Engines 338 I0.6.3 Number and Location - Conclusions I0.7 POSITIVE-PRESSURE PURGING SYSTEM I0.8 SUCTION AND PURGING CONTROL I0.9 SUMMARY - SUCTION SYSTEM II .0 ENVIRONMENTAL II.I INTRODUCTION ll.2 CONTAMINATION AVOIDANCE & ICE PRO+EC+ION SYSTEM REQUIREMENTS 352 II.3 POSSIBLE SYSTEMS CONSIDERED II.3.1 Mechanical Systems ll.3.1.l Retractable Shield .

II.3.I.2 Retractable Cover II.3.1.3 Mechanical Scraper . 359 11.3.1.4 Recommended Approach - Mechanical System 361 11.3.2 Liquid Systems . 361 11.3.2.1 Liquid Film 362 ll.3.2.2 Washer System 365 II.3.2.3 Erosion by Water Spray 365 II.3.2.4 Recommended Approach - Liquid System 366 II.3.3 Electrically Powered Systems 367 II.3.3.1 Electro Impulse . 367 II.3.3.2 Ultrasonic Vibration II.3.3.3 Thermal Decomposition 368 II.3.3,4 Magnatohydrodynamics 369 II,3.3.5 Recommended Approach - Electrical viii

CONTENTS (Continued) Section

Page

11.3.4 Covers Disposable 369 113 .4.1 Temporary Coating . 369 113 .4.2 Chemically Removed Coating 113 .4.3 Electrically Removed Coating 113 .4.4 Mechanically Removed Cover 372 II 3 .4.5 Thermally Removed Coating 373 II 3 .4.6 Cover Recommended Approach - Disposable II .3.5 for Contamination Speci al Surface Coatings Avoi dance II .3 .5.1 Anti-Sti ck Coating" II .3 .5.2 Elastic Coating 11.3 ,5.3 Hydrophylic Coati ng 11.3.6 Contamination Protection- Initial Conclusions 375 II .4 SHIELD PRELIMINARY DESIGN & PERFORMANCE EVALUATION II .4.1 Preliminary Design Studies II .4.2 Contamination Avoidance Effectiveness Anaiyses 11.4.

2.1 Computer Programs .

11.4.

2.2 Shielding Analysis 381 II .4.

2.3 Trajectory Analysis . . .

II .4. 2.4 Effectiveness Analysis - Conclusions 11.4.3 Wind Tunnel Test of Shield Effectiveness 39O II.4.4 Shield Operational Use 390 11.5 SHIELD ICE PROTECTION SYSTEM 11.6 LIQUID-FILM PRELIMINARY DESIGN & PERFORMANCE EVALUATION 11.6.1 Preliminary Design Studies 394 Liquid-Fi Im Concept 11.6.2 Experimental Evaluation of 397 11.6.2,1 Liquid Distribution 397 11.6.2.2 Removal of Debris 11.6.2,3 Residual Film 11.6.2.4 Clogging and Clearing" 403 11.6.3 Liquid Formulation 11.6.4 Liquid Film Testing & Analysis - Summary 421 11.6.5 Liquid Film System Design .

11.7 SPRAY SYSTEM PRELIMINARY DESIGN & PERFORMANCE EVALUATION 11.7.1 Shield Spray System Evaiuation II .7.1 .I Liquid Droplet Trajectory 428 II .7.1,2 Nozzle Selection and Performance II .7.1.3 Drop Shattering 11.7.1.4 Preliminary Shield Spray Design 438 II .7.1.5 Shield Spray System Evaluation 438 II.7.2 Wing Spray System Evaluation .

11.7.2.1 Liquid Droplet Trajectory 439 II .7.2.2 Droplet Shattering . .

II .7.2.3 Nozzle Selection and Performance .

11.7.2.4 Preliminary Wing Spray System Design 11.7.2.5 Wing Spray System Evaluation 450 ix

CONTENTS

(Continued)

Section ll.8 EFFECT OF ATMOSPHERIC ICE CRYSTALS ll.9 EFFECT OF ENVIRONMENTAL CONDITIONS ON FLOW THROUGH POROUS SURFACE .

ll.9.1 Environmental Contamination II.9.2 Effect of Ambient Pressure & Temperature . 464 ll.lO CONCLUSIONS - ENVIRONMENTAL ll.ll RECOMMENDATIONS - ENVIRONMENTAL ll.12 REFERENCES - ENVIRONMENTAL 12.0 CONCLUSIONS & RECOMMENDATIONS 12.1 LFC SURFACE 12.2 LFC AIRCRAFT CONFIGURATION 12.3 COMPARISON WITH ADVANCED TURBULENT AIRCRAFT 12.4 RECOMMENDATIONS 12.4.1 Full Chord LFC Glove 12.4.2 Supporting LFC Programs 487 12:4.3 LFC Demonstration Aircraft / x LIST OF ILLUSTRATIONS Figure

Page

4-I Seat-Mile Demand by Aircraft Types 4-2 Aircraft Development Trend Lines 5-I Interior Arrangement 5-2 a Initial LFC Base Case Airplane 5-2b Initial Turbulent Base Case Airplane 5-3 Base Case Airfoils Assumed 5-4 Buffet Boundary for Turbulent Base'Aircraft 5-5 Cruise CL vs Wing Thickness/Chord Ratio 5-6 Preliminary Wing Thickness Definition 5-7 Preliminary Wing Twist Distribution 5-8 Lift Curves for LFC Aircraft 5-9 Preliminary Variation of Flap Characteristics for LFC Aircraft Estimated Taii-Off Lift Curves for Turbulent 5-I0 Base Aircraft 5-11 Effect of Composite Usage in'Fuselage on LFC Aircraft Takeoff Gross Weiflht 5-12 Pseudo-lsotropic Composite Corrugation 5-13 Suction Pump/Dry Bay Area Installation 5-14 Predicted Suction Requirements for LFC Airfoil 5-15 Preliminary Estimate of Fuel Flow for Suction System Pumping Aircraft Altitude Profiles 5-16 5-17 Climb Parameters 5-18 Mission Sizing - LFC Baseiine -'Takeoff Gross Weight 5-19 Mission Sizing- LFC Baseline - Fuel Burned . .

5- 20 Mission Sizing - Turbulent Baseline - Takeoff Gross Weight .......

5-21 58 Mission Sizing - Turbuient Baseiine- Fuel Burned 5-22 Effect of Aspect Ratio on Aircraft Sizing 5-23 a Four-Engine Configuration - Study II 5-23 b 62 Three Engine Configuration - Study II 5-24 a LFC Aircraft - Upper & Lower Wing Surface Laminarized to 70 Percent Chord 5-24 b LFC Aircraft - Wing Upper'Surface Laminarized to 85 Percent Chord 5-24 c Advanced Turbulent Aircraft 5-25 69 Effect of LFC Extent on Profile'Drag 5-26 Variation of Design Lift Coefficient with Airfoil Thickness 5-27 Comparison of Wing Sections ....

5-28 Low-speed Lift Curves - Upper Surface Suction Only 5-29 Low-speed L/D Characteristics - Upper Surface Suction Only 5-30 Low-speed Lift Curves - Base Case - Suction on Both Surfaces . . .

5-31 Low-speed L/D Characteristics - Suction on Both Surfaces xi

LIST OF ILLUSTRATIONS

(Continued)

Figure Page

5-32 Horizontal Tail Sizing - Upper Surface Suction Only 78 5-33 Horizontal Tail Sizing - Upper & Lower Surface Suction 78 5- 34 LFC Aircraft - Balance Diagram - Upper Surface Suction Only. 80 5-35 LFC Aircraft -Baiance Diagram - Upper & Lower Surface Suction 80 5- 36 Spoiler Lift Effectiveness at Low Speeds 84 5-37 Structural Arrangement - Suction Through Both Surfaces 84 5-38 Typical Wing Rib - Suction Through Upper Surface Only 85 5- 39 Suction System Manifolding Integration 85 5-40 Structural Arrangement - Upper Wing Box Panel at CL of Suction Pump System. 86 5-41 Structural Arrangement - Lower Surface & Fuel Bulkhead at CL of Suction Pump System 86 5-42 LFC Glove Panel Layer Identification 88 5-43 Leading Edge Design - Suction on Both Surfaces 88 5-44 Leading Edge Design - Upper Surface Suction Only 90 5-45 Spray Concept 90 5-46 Outboard Wing Shield 91 5-47 Wing Structure/Suction Ducting Integration - Upper Surface Only Laminarized . 91 5-48 Manifold Ducting to Suction Pump - Both Surfaces Laminarized to 70 Percent Chord 92 5-49 Integration of Suction Ducting in Dry Bay Area - Laminarization of Upper Surface Only to 85 Percent Chord 92 5-50 Estimated Center-of-Gravity Limits - Final Turbulent Configuration 98 5-51 Turbulent Aircraft Sizing . 98 5-Al Parasite Drag - Laminar & Turbulent I06 6-I Effect of Fuel Price on Fuel Cost Savings ll8 6-2 Reduction in DOC as a Function of Fuel Cost ll8 7-I Typical LFC Wing Chordwise Pressure Distributions and Airfoil Shapes - 250 Sweep . . . 121 7-2 Typical LFC Wing Chordwise Pressure Distributions and Airfoil Shapes - 30 o Sweep 121 7-3 Variation of Design Lift Coefficient With Airfoil Thickness 125 7-4 Typical Suction Distributions for Upper Wing Su;face - 30 o Sweep 125 7-5 Typical Suction Distributions for Lower Wing Surface - 30o Sweep 127 7-6 Variation of LFC Suction Coefficient With Design CL 127 7-7 Variation of Suction Required for LFC With Thickness/Chord Ratio- 30 o Sweep 128 7-8 Variation o# Suction Required for LFC With Thickness/Chord Ratio - 25 v Sweep . 128 xii LIST OF ILLUSTRATIONS (Continued)

Page

7-9 Variation of Total Suction Coefficient With Reynolds Number 7-10 Variation of Profile Drag With Reynolds Number 7-11 Effect of Loss of LFC on Airfoil Pressure Distributions 7-12 Effect of Trailing Edge Flap on Airfoil Pressure Distributions 7-13 Change of Airfoil Pressure Distribution With Lift Coefficient 1 34 7-14 Effect of Charging Lift Coefficient on LFC Suction Velocity Distribution 1 34 7-15 Near Field Acoustic Environment Due to One Engine 137 7-16 Allowable Acoustic Environment 7-17 Effect of Engine-lnduced Acoustic Environment on Extent of LFC - Wing Mounted Engines . . 139 7-18 Effect of Engine-lnduced Acoustic Environment on Extent of LFC - Fuselage Mounted Engines 7-19 AspectRatio Sizing - Wing Planfo_ms" 7-20 Effect of Aspect Ratio on Angle of Zero Lift 7-21

High-Lift System for LFC Base Case - Suction on S rf ce

to 70 Percent Chord 144 7-22 High-Lift System - Upper Surface Suction Only 7-23 Elastic Wing Bending Stiffness Criteria 7-24 Elastic Wing Bending Stiffness & Weight Factors 7-25 Mission Sizing Matrix for A.R. = lO - Suction on Both Surfaces 7-26 a Aspect Ratio Siling Summary'l - Suction'on'Bo;h Surfac;s 7-26 b Aspect Ratio Sizing Summary II Suction on Both Surfaces.

7-27 Mission Sizing Matrix for A.R. = I0 - Upper Surface Suction 7-28 a Aspect Ratio Sizing Summary Comparison I 7-28 b Aspect Ratio Sizing Summary Comparison II 8-I Wind Tunnel Model Test for Laminar Flow With Porous" Surface Specimens 8-2 Wind Tunnel Model Instaliation 8-3 Wind Tunnel Model - Solid Surface'Pressure Distribution 8-4 Transition Location on Wind Tunnel Model - Solid Plate Surface 160 8-5 Boundary Layer Signatures - Probe Signal Displayed on Oscilloscope 8-6 Typical Wind Tunnel Results - Transition Location 162 8-7 Comparative Effectiveness of Laminar Flow Control Surfaces 168 8-8 Swept Wing Model Test - Representative Chordwise Pressure 8-9 Swept Wing Modei- Airfoil Profile in Wind Tunnel 8-I0 Swept Wing Model - Chordwise Pressure Distribution of Airfoii Installed in Test Section 169 xiii LIST OF ILLUSTRATIONS (Continued) Figure

Page

8-11 Swept Wing Model - Upper Surface Pressure Distribution & Suction Velocity 8-12 Swept Wing Model Test - Reference Steamline on Upper Surface - Infinite Yawed Wing 171 8-13 Swept Wing Model Test - Effect of Tunnel Sidewalls on Pressure Distribution 8-14 Swept Wing Model Test- Sidewall Fairing Development 8-15 Swept Wing Model - Components 17_ 8-16 17_ Swept Wing Model - Basic Structure 8-17 Wind Tunnel Model Porous Panel Configurations 176 8-18 Swept Wing Wind Tunnel Model With Dynapore Porous Surface 8-19 Swept Wing Model With Upper Porous Panel Installed . 177 8-20 Perforated Titanium Surface Near Front Spar Joint Showing Non-Porous Region and Welded Joint 8-21 Perforated Titanium Leading Edge Showing Suction & Static Lines• 1 79 8-22 Swept Wing Model Test - Model Instaliation 1 79 8-23 180 Swept Wing Model Test - Upper Surface at Leading Edge 8-24 1 80 Swept Wing Model Test - Suction System Schematic 8-25 181 Wind Tunnel Model With Suction System 8-26 Arrangement of LFC Panel Insert in Wing Leading Edge 8-27 183 Wind Tunnel Model - L.E. Insert Removed.

8-28 Leading Edge Insert Substructure Configuration and Suction Outlet Tubing . . , 8-29

LFC Swept-Wing WithPerfOrated Titanium Insert in D_ugias

Long Beach Wind Tunnel .

8-_ 1 85 Schematic Layout of Data Acquisition 8-31 LFC Theoretical Streamline Above Model Surface on Tunnel CL .....

8-32 Tunnel Velocity Calibration - Model Installed in Tunnel 8-33 Streamwise Pressure Coefficient Versus Surface Station. 188 8-34 Swept Wing Model Test - Spanwise Pressure Distribution Compari son 8-35

ReEvaluation -'Su_faLe _or_al'Veiocity Versus Surface

Distance 189 8-3

Non Porous Swept-" Wing" - Baseline FlowVisullizationWith

1 89 Ti 02 Oil Mix ....

8-37 Swept Wing Modei Test'- Naturai Transition Location 191 8-38 Effect of Crossflow on Streamlines With Non-Porous Surface on Swept Wing Model 8-39 Swept Wing Model Test - Comparison of Reference Streamline and Surface Flow Visualization 194 8-40 1914 Porous Surface Velocity Distribution I xiv LIST OF ILLUSTRATIONS (Continued) Fi9ure

Page

8-41 Leading Edge Porosity Distribution 8-42 Leading Edge Insert Porosity Distribution 8-43 Swept Wing Model Test - Effect of Suction on Transition - Dynapore Surface 8-44 Swept Wing Wind Tunnel Model - Suction Velocity Distribution 8-45 Swept Wing Model Test - Effect of Suction on Transition - Dynapore L.E. Insert 199 8-46 Swept Wing Model Test - Effect of Suction on Transition - Perforated Titanium Leading Edge Insert. 199 8-47 Swept Wing Model - Leading Edge Insert Suction Veiocity Distribution 9-I LFC Structural Design Concepts 9-2 Initial Baseline Structural Concept 209 9-3a Baseline Wing Structure 210 9-3b Corrugated Structure (Compositei 9-4 Wing Moments & Shears 212 9-5 Wing Panel Bending Load Intensity - Spanwise Variation 9-6 Preliminary Wing Thickness Definition 213 9-7 Flutter Speed versus Percent Fuel 9-8 Wing Torsional Stiffness Increase Required for'Flutter 9-9 Archweb - Stiffened Concept 9-10 Isogrid-Stringer Concept 218 9-11 Peel-Ply Concept 218 9-12 Honeycomb Concept 224 9-13 External Blade Stiffener Concept 9-14 Control System Attachment - External Blade 9-15 Internal Blade Stiffener Concept - Chordwise Air Collection 9-16 Control System Attachment - internal Blade Stiffener Concept 229 9-17 Increased Stiffness'Required versus Aspect Ratio 9-18 Spanwise Variation of Increased Stiffness 231 9-19 Effect of Thickness/Chord Ratio on Increased Stiffness Required 9-20 Structural Arrangement - Suction Through Both Surfaces 9-21 Structural'Arrangement - Upper Wing Box Panel at CL _f Suction Pump System 232 9-22 Structural Arrangement - Lower Surface at CL Suction 9-23 Typical Wing Rib - Suction Through'Upper Surface Only . . 234 9-24 Lock-Core Basic Panel Def nition 9-25 Lock-Core Panel Xv LIST OF ILLUSTRATIONS (Continued) Figure 9-26 238 Early Modified Lock-Core Panel Concept 9-27 238 Porous Isogrid Panel.

9-28 239 Kevlar/Epoxy Isogrid Core 9-29 239 Isogrid Core Splice 9- 30 245 Isogrid & Honeycomb Coves Bonded to Dynapore Facings 9-31 Effect of Calendering on Porosity of Dynapore Materials 9-32 Pressure Drop Panel Design - Porous Composite Layers 248 9-33 80 x 700 Dynapore Surface Plus Diffusion Bonded 80 x 80 Sublayer 9- 34 Electron Beam Perforated Titanium 9-35 Electron Beam Perforated Hole 249 9- 36 Strength & Strain Requirements for LFC Glove Panels 252 9-37 Thermal Analysis Model & Materials 252 9-38 Glove Panel Installation 9- 39 Typical Wing Box Glove Panel Arrangement 9-40 LFC Glove Panel With Perforated Fiberglass Sublayer 9-41 Glove Panel Splices 9-42 Typical Field Fasteners 9-43 257 LFC Gloved Wing Structural Concept .

9-44 Super-Plastic Formed/Diffusion Bonded Titanium Sandwich Panel Before Machining . . .

9-45 Super-P_astic Formed/Diffusion Bonded +itanium Sane_vich Panel 259 • • • ° • 9-46 Super-Plastic Formed/Diffusion Bonded Titanium Sandwich Panel - Suction Holes 260 9-47 Test Beam 260 9-48 262 Stress/Strain Curve - Dynapore on Fiberglass 9-49 262 Stress/Strain Curve - Nitronic 50 Dynapore 9-50 Fatigue Test Beam 9-51 264 Electron Beam Perforated Titanium Sheet Fatigue Specimens 9-52 267 Porous Surface Fatigue 9-53 267 Impact Test Specimen 9-54 Impact Resistance 9-55 Impact Resistance Comparisons 9- 56 Rain Erosion Test Specimen 9-57 Tested Dynapore Rain Erosion Specimens 9-58 Tested Dynapore Rain Erosion Specimens 9-59 Compression Test Specimen . 275 9-60 Failed Compression Specimen 277 9-61 • 277 Joint Test Specimen 9-62 Joint Compression Test Set-up 9-63 Failed Joint Compression Test Specimen 9-64 Joint Fatigue Specimen 9-65 Compression Test Panel 9-66 285 Compression Test Panel Set-up 9-67 Large Compression Panel Test Arrangement xvi LIST OF ILLUSTRATIONS (Continued) Figure 9-68 Large Compression Test Panel 287 9-69 LFC Demonstration Panel Failure 9- 70 LFC Panel Failure 9-71 Graphite/Epoxy Wing Box Panel Failure 288 9-72 Demonstration Panel - Average Strain History 9-73 LFC Leading Edge - Phase I 9- 74 LFC Leading Edge - Phase II 9-75 Leading Edge Design - Separate Collector Ducts 9-76 Leading Edge Design - Integral Collector Ducts 9-77 Leading Edge Box Details at Front Spar 9-78 Leading Edge Initial Co-Cure Assemblies 9- 79 297 Leading Edge Structure - Titanium Surface 9-80 299 Swept Wing Wind Tunnel Model LFC Surface 9-81 Wind Tunnel Model Substructure in Molding Form Tool 9- 82 301 Wind Tunnel Model Leading Edge After Curing 9-83 301 Leading Edge STM - Before Bonding to Dynapore Surface 9-84 303 Dynapore Lap Shear Joint Welding Technique .

9-85 Leading Edge Panel With Porous Surface 30 3 lO-I Wing Surface Pressure & Suction Requirements I0-2 Effect of Surface Porosity on Suction Velocity I0-3 Typical Flow Characteristics - Porous Surface I0-4 Effect of Suction System Pressure Drop on Compressor Requirements• I0-5 314 Chordwise Air Collection System I0-6 Spanwise Air Collection System 315 I0-7 Typical Upper Surface Construction and Suction System I0-8 Typical Wing-Skin Cutout for Suction Duct I0-9 Ducting Aft of Rear Spar . .

lO-lO Ducting Aft of Rear Spar - Suction on Upper Surface Only 320 lO-ll Typical Suction System Pressures lO-12 Effect of Main Compressor Ratio lO-13 Typical Compressor Map lO-14 Schematic Showing Use of "Boosted" U.S. Air to Drive Ejector . .

lO-15 Effect of Boost Compressor Pressure Ratio lO-16 Typical "Bleed & Burn" System Results lO-17 Suction Airflow in Wing-Box Ducts lO-18 341 Typical Area for Wing-Box-Duct Cutouts lO-19 341 Suction System Manifolding Integration I0-20 Schematic of Suction System ....

lO-21 Effect of Suction Design Macn Number on Purging Mach Number with Positive Pressure 346 ll-l Insect Frequence vs Weight - Near the'Ground ll-2 354 Insect Ballistic Coefficient (k) vs Weight ...,' ll-3 355 Insect Population Density vs Altitude ll-4 Effect of Angle of Attack on Limit of Insect Impingement .

xvii

LIST OF ILLUSTRATIONS

(Continued)

11-5

Preliminary Concept of Internally Stowed Shield

II-6

Preliminary Concept of Retractable Shield.

11-7

Freezing Point Depressant L.E. Distribution Strips Alternate Air & FPO Passages .

ll-8

Preliminary Concept for Freezing Point Depressant Distribution System .....

ll-9

Preliminary Concept of Freezing Point Depressant Spray on Fixed L.E.

Insect/Ice

ll-lO

Retractable Shield - Deployed for Protection Fixed Camber Shield - Extended 378

ll-ll

ll-12 Variable Camber Shield - Extended

ll-13

LFC Wing Characteristics

ll-14

Insect Trajectories - Takeoff, _=0, K=O.2, 0.9 b/2

ll-15

Insect Trajectorles - Takeoff, m=15o, K=O.2, 0.9 b/2

If-16

Insect Trajectories - Takeoff, _=0, K=O.l, 0.8 b/2

ll-17

Insect Trajectories - Takeoff, m=-4 °, K=O.l, 0.8 b/2

ll-18

Insect Trajectories - Takeoff, e=15o, K=O.l, 0.8 b/2

ll-19

Insect Trajectories - Takeoff, m=15o, K=0.25, 0.8 b/2

ll-20

Insect Trajectories - Takeoff, m=15 °, K=0.35, 0.8 b/2

ll-21

Insect Trajectorles - Takeoff, _=0, K=O.2, 0.13 b/2 .

II-22

Insect Trajectories Takeoff, _=15o, K-O.2, 0.13 b/2

II-23

Insect Impingement Tests Schematic Diagram of LFC LIE. Ice Protection'System

II-24

II-25

First Approximation of Active Porous Panel Width Required on Shield for Ice Protection

II-26

FPD Liquid Flow Tests in Low Speed Wind Tunnel

II-27

Clogging & Clearing of Porous Material

II-28

Viscosity Vs Temphrature for Candidate Liquids

II-29 a

Effect of Porosity on Liquid Clearance at 12.4 kPa.

II-29 b Effect of Porosity on Liquid Clearance at 15.5 kPa.

I 1-30 Effect of Surface Porosity on Liquid Clearance ll-31 Effect of Overall Porosity on Liquid Clearance I 1-32 Time Taken to Clear Diluted Ethylene Glycol II-33 Time Taken to Clear Diluted Ethyl Alcohol II-34 Time to Clear Diluted PGME II-35 Liquid Clearance Time vs Pressure -'Various Liquid) Surface Combinations II-36 Liquid Clearance Time vs Pressure -'Various Liquid> Surface Combinations 11-37 Shield Spray Concept 11-38 Retractable Nozzle Concept 11-39 Schematic Diagram of LFC L,E. Contamination Avoidance System - Shield Spray Concept I 1-40 Minimum Droplet Diameter to Impinge on Wing vs Initial Drop Velocity 11-41 Visual Aid for Comparing Relative Particle Sizes of Nozzles with Equal Capacities and Spray Angles xviii

LIST OF ILLUSTRATIONS

(Continued)

Figure Page

11-42 434 Spray Particle Size Vs Pressure 11-43 436 Shield Spray Performance 11-44 436 Drop Size Distribution 11-45 437 Cumulative Drop Size Distribution.

11-46 Droplet to Air Relative Velocity - Shield Spray System 11-47 441 Maximum Droplet D ameter, Without Shattering - Sea Level 11-48 441 Dro)let Trajector les - a=O, 279 Micron, 45.7 m/s 11-49 442 Droplet Trajector les - m=O, 260 Micron, 45.7 m/s 11-50 442 Dro)let Trajector les - m=O, 245 Micron, 45.7 m/s 11-51 443 Droplet Trajector les - _=0, ll60 Micron, 45.7 m/s 11-52 443 Dro)let Trajector les - re=O, I025 Micron, 18.3 m/s 11-53 Dro)let Trajectorles - m=lO, 187 Micron, 45.6 m/s 11-54 444 Dro)let Trajectories - m=l ° 170 Micron, 45.7 m/s II-55 Dro)let Trajectorles m=-l _, 410 Micron, 45.6 m/s 445 II-56 445 Dro)let Trajectorles - m=-l °, 348 Micron, 45.7 m/s II-57 446 Range of Drop Size for Impingement, m=l ° II-58 446 Range of Drop Size for Impingement, m=O II-59 447 Range of Drop Size for Impingement, m=-l ° I 1-60 447 Droplet-to-Air Relative Velocity - Wing Spray System ll-61 Maximum Droplet Diameter, Without Shattering - 35,000 Ft. Altitude 11-62 449 Liquid Impingement Efficiency vs Angie of Attack 11-63 451 Wing Spray Droplet Size Distribution 11-64 Liquid Impingement Efficiency vs Angle of Attack - Optimum Nozzle Pressure . .

11-65 452 Wing Spray Droplet Size Distribution 11-66 Estimated LFC Performance in Clouds and Light Haze at 7,620 m (25,000 ft) Altitude 11-67 Estimated LFC Performance in Clouds and Light Haze at 12,190 m (40,000 ft) Altitude 11-68 a 462 Effect of Environmental Contamination - MPP 11-68 b 462 Effect of Environmental Contamination - Dynapore 11-69 Effect of Environmental Contamination - E.B.

Perforated Titanium 11-70 Effect of Ambient Pressure on Pressure Drop Characteristics With Internal Laminar Flow Characteristics - Dynapore.

II-71 Effect of Ambient Pressure on Pressure Drop Characteristics With Internal Turbulent Flow Characteristics - Doweave 11-72 Porous Surface Pressure Drop at Various Fluid Temperatures - Higher Porosity .......

11-73 Porous Surface Pressure Drop at Various Fluid Temperatures - Lower Porosity 12-I LFC Aircraft - Wing Upper Surface Laminated to 85 percent Chord ....

12-2 LFC Aircraft - Upper and'Lower'Wing Surface Laminarized to 70 percent Chord .

Advanced Turbulent Aircraft .

12-3 486 12-4 486 Proposed LFC Glove on DC-9 Wing Box 12-5 488 LFC Glove Wing on DC-9 Aircraft 12-6 488 DC-9 With LFC Wing xL, c C, LIST OF TABLES Table

Page

4-I Aircraft Sizing & Mission Ground Rules 14 5-I Basic Study Ground Rules 17 5-2 Initial Baseline Study Ground Rules 5-3 Tail Sizing Criteria 5- 4 a/b Structure & Material Assumptions Compatible with 1990-95 Aircraft .34/35 5-4 c Composite Factors 36 5-5 Composite Weight Reduction 5-6 Suction System Ducting Areas 5-7 Suction System Sizing - Base Case 5-8 Aircraft Characteristics 5-9 Weight Estimation Methods & Assumptions . 47 5-10 a Weight Summary 49 5-10 b Wing Weight Comparison 5O 5-11 Design Data & Geometry Summary 5-12 Performance Characteristics - Study'l .

5-13 Preliminary Costing- Initial Baseline Aircraft 5-14 a Four vs. Three Propulsion Engines - Drag & Weight Differences 5-14 b Characteristics of Sized Aircraft - Study'll 63 5-14 c Direct Operating Cost Summary - Study II 5-15 Comparison of Turboshaft Engine Arrangements for Suction System . . 65 5-16 Final Configuration- Ground Rules 5-17 Lateral Control Contributions Comparison 5-18 Weight Breakdown of Operational Items 93 5-19 Laminar Flow Control Suction/Ducting System Weights 5-20 Weight Summary - LFC Aircraft Comparison .

5-21 LFC Aircraft Characteristics Comparison 5-22 Aircraft Characteristics Comparison - Turbulent vs Laminar .....

5-AI LFC Initial Base Case Aircraft Parasite Drag 107 6-I General Costing Ground Rules 110 6-2 DOC Factors & Coefficients 6-3 Characteristics of Configurations Analyzed 6-4 Cost Summary . . . 113 6-5 Acquisition Cost Comparison . .

6-6 DOC Comparison 7-I Configuration & Performance- Sized'Aircraft" 9-I Initial Evaluation Matrix . . .

9-2 Glove vs Integral LFC. Surface 208 9-3 Initial Baseline Wing Stiffness Distribution 215 9-4 Preliminary Structural Design Assessment 9-5 Wing Structure Weight - Effect of Aeroelastic Requirements 9-6 Wing Weight Comparison - Corrugated'vs'Blade'Stiffening- " - - " " " PRZCED;N3 F'_,L;£ _' ;".,r'"<" NOT FiL/v;_.D >:xi LIST OF TABLES

©

(Continued) Tab I e

Page

9-7 Wing Weight Comparison - Blade Stiffened Structures 9-8 Aerodynamic Smoothness Test Panels 9-9 Effect of Thermal Strains with Combined Materials 253 9-10 263 Fatigue Load Requirements 9-11 Fatigue Specimen Identification 9-12 268 Impact Test Panels .

9-13 Erosion Specimen Materiai Configurations . . . 270 9-14 273 Erosion Specimens with Smaller Sublayer Perforations 9-15 276 Joint Specimens - Test Schedule 9-16 289 Demonstration Panel - Strain History I0-I 324 Off-Design Requirements 10-2 325 Compressor Comparison 10-3 Typical Suction Power Required" 329 10-4 Effect of Bleed Temperature for Bleed-only System 10-5 332 Cycle Comparison - Assumptions 10-6 Turboshaft & Bleed-and-Burn Cycles 10-7 Comparison of Fuel Required - Turboshaft and Bleed-and- Burn Systems 10-8 Qualitative Comparison - Suction Power Systems .

10-9 Typical Parameters - Suction & Purging II-I 370 Available Coatings Effect of Suction Plenum'to'Local Ambient Differential 11-2 Pressure 11-3 Physical Characteristics of Test Liquids 11-4 Construction & Overall Porosity - Clogging & Clearing 4O8 Test Specimens . .

11-5 Properties of Test Liquids 11-6 Candidate Glycol Ethers 11-7 421 Candidate Alcohols, Fluorocarbons & Heat Transfer Fluids 11-8 424 Liquid System Weights 11-9 Probability of Encountering Cirrus Clouds . .

II-I0 Environmental Contamination - Test Sample Descriptions. 460 II-II Effect of Contamination/Clearing on Pressure Drop of Moderate Porosity Samples x.xLi

©

1.0 SUMMARY This "Evaluation of Laminar Flow Control (LFC) Systems Concepts for Subsonic Commercial Transport Aircraft", considered all aspects of the application of LFC to commercial transport aircraft in operation. The problem areas were identified and tackled systematically until resolved. Program activities included configuration design and analysis, performance and economic analysis, fabrication development, environmental studies, contamination avoidance systems design and testing, structural design/analysis and testing, and wind tunnel testing. The results of LFC program activities up to December 1980 are covered in this report. For summary reports see References 1.0-I and 1.0-2.

Laminar Flow Control was achieved by using controlled suction through the external surface to stabilize the laminar boundary layer and prevent transition to turbulent flow, thus achieving significant drag reduction.

An objective of the program was to take advantage of any new and advanced technology consistent with a mid 1990's aircraft time frame. With this in mind, it was decided to examine the possibilities of using porous materials at the surface to control suction airflow rather than use a series of very fine slots, as used previously on the Northrop X21 aircraft program. Due to the very limited data base available on the use of porous materials for achieving LFC, an extensive survey of possible porous materials and their application was undertaken. This involved design studies, fabrication development and structuraland aerodynamic testing. The field was eventually narrowed down to two promising materials, a smooth finely woven stainless steel mesh manufactured under the trade name Dynapore and electron beam (EB) perforated titanium sheet material, perforated with Steigerwald equipment. The EB perforated titanium surface was finally selected, after exhibiting excellent LFC characteristics in the wind tunnel, and because of its better structural and damage resistance properties.

A number of configuration trade studies were undertaken, the most significant being a comparison of LFC on both upper and lower wing surfaces compared with LFC on the upper surface only. Using suction to 70 percent chord for LFC on both surfaces and to 85 percent on the upper surface only, it was found that the reduction in drag coefficient and the total suction airflow required were ,i of the same order. The advantages of having suction on the upper surface only with respect to simplicity, reduced damage vulnerability and the availability of access through the lower surface for maintenance are obvious. Not so obvious is perhaps the main advantage, the possibility of using a shield at the leading edge to avoid surface contamination. This shield can also function as a high lift device and be retracted into the lower surface after use. The trade study showed the superior performance of the upper-surface-suction-only configuration. This configuration was therefore selected for the baseline LFC aircraft to be used in subsequent studies.

Another significant trade study was a comparison of the LFC aircraft with an advanced turbulent aircraft configuration. This clearly showed the advantages of LFC with respect to reduced fuel consumption and reduced operating cost, particularly with rising fuel prices.

In examining all aspects of the practical application of LFC to commercial transport aircraft, no problem was found for which a practical solution could not be identified, as shown by analysis, design studies, and development testing undertaken in this program. The overall results indicate that the LFC aircraft configuration, suggested by Douglas in this study, could be developed into a practical design that would bring signifiant fuel saving and operating cost benefits.

SUMMARY REFERENCES 1.0-I Pearce, W. E., Evaluation of Laminar Flow Control Systems Concepts for Subsonic Commercial Transport Aircraft - Executive Summary, NASA Contractor Report 159252.

1.0-2 Pearce, W. E., Progress at Douglas on Laminar Flow Control Applied to Commercial Transport Aircraft, Paper ICAS-82-I.5.3, Proceedings of the 13th Congress of the International Council of the Aeronautical Sciences, August 1982.

2.0 INTRODUCTION \ This investigation into the possibilities of using laminar flow control (LFC) on commercial transport aircraft was initiated by NASA in response to the growing need for energy conservation.

Fuel saving results directly from the drag reduction that can be achieved by using LFC. The successful application of LFC to commercial airplane operation would result in a major reduction of fuel consumed by airline fleets throughout the world. With rising fuel costs, increasing economic benefits are also obtainable.

The airflow over the surface of an airplane is initially laminar within the boundary layer but this low drag condition is unstable and transition to turbulent flow occurs normally over a very short distance. On a swept wing this instability is aggravated by cross flow conditions in regions of steep pressure gradients. Transition can also occur due to the spanwise flow along the attachment line at the leading edge. In all of these cases, transition to turbulent flow can be avoided by the use of suction through the surface to stabilize the laminar boundary layer.

Ideally the suction airflow would be distributed over the whole area using a porous surface but when this study was undertaken a practical solution to achieving this did not exist. Very fine suction slots had been used previously to create intermittant suction at frequent intervals in order to sustain laminar flow. Although slotted systems have been tested successfully, full scale flight testing of a slotted system on the Northrop X21 airplane wing in the early 1960's demonstrated many of the difficulties of making such a system reliable and it was not considered to be commercially practical at that time.

The approach adopted by Douglas was directed towards taking full advantage of recent advances in technology to achieve a practical, reliable and economic LFC system for commercial transport aircraft, by using suction distributed through porous surfaces.

Preliminary design studies resulted in an initial LFCairplane configuration

that was used as a baseline for LFC system and structural design and

configuration trade studies. The baseline was updated at intervals and was

comparedwith a turbulent design to determine relative economic and

performance advantages.

Particular emphasis was placed on the design and development of suitable

porous surfaces, their supporting substructure, and the use of integral

ducting for the suction airflow. A variety of designs was considered. The

most promising were tested to determine structural strength, airflow

characteristics, and LFC performance in the Douglas wind tunnel at Long

Beach. The 2.14 m (7 ft) chord swept wing LFC wind tunnel model tested was

funded by Douglas in support of the LFCprogram.

This report is divided into sections covering the work accomplished under the

principal disciplines involved. Section 5.0 describes the aircraft

configuration studies and provides an overview of many of the activities

covered in greater detail in the other sections. Each of these sections is

virtually self contained but references are given to related activities. The

conclusions and recommendations are presented in Section 12.0.

3.0 SYMBOLS AND ABBREVIATIONS APU = Auxiliary power unit AR = Aspect ratio b = Wing span B = Boron = Chord c, C = Drag coefficient C d = Flap chord CF c.g., C.G. = Center of gravity = 2-Dimensional lift coefficient = Wing lift coefficient CL = Center line

t

C : Moment coefficient m = Side slip force coefficient Cn 8 = Pressure coefficient Cp(N) CPR = Compressor pressure ratio = Suction airflow coefficient

CQ

= Nozzle velocity coefficient CV d = Diameter DOC = Direct operating cost e = Induced drag efficiency factor E : Young's modulus of elasticity E 3 = Energy efficient engine - NASA program EB = Electron beam LET = Energy Efficient Transport - NASA program E-I = Electro impulse E = Epoxy P f = Equivalent flat plate drag area or friction loss f/a = Fuel/air ratio = Engine thrust FN = Takeoff thrust FNTO FOD = Foreign object damage FPD = Freezing point depressant FS = Front spar = Acceleration due to gravity g = Shear modulus G = Graphite (carbon) Gr h = Height = Horizontal stabilizer angle of incidence i H = Wing angle of incidence i W = Zero wing lift angle of incidence iZLL I = Moment of inertia J = Polar moment of inertia -- Ballistic coefficient K Kev = Kevl ar L = Length L/D = Lift/drag ratio L.E.

= Leading edge LFC = Laminar Flow Control : Lower surface LS

M = Mach number

MAC

= Aerodynamic mean chord

MEW

= Manufacturers empty weight

MLW

= Maximum landing weight

MPP

= Micro perforated plate

MTOW

= Maximum takeoff weight

MZFW

= Maximum zero fuel weight

N = Load factor

NWLO = Nose wheel lift-off

OASPL

= Overall sound pressure level

OEW

= Operator's empty weight = Air pressure

p, P

PGME

= Propylene glycol methyl ether

PLM

= Plastic laminating mold = Pressure ratio

PR

P

= Ultimate design load

ult

= Airflow dynamic pressure

q

Q = Porosity

= Quiet engine - NASA program

QCSEE

= Nozzle flow rate QT R = Fatigue stress ratio = Reynolds number R c , Re S = Distance measured along surface SCFM = Standard cubic feet per minute SFC = Specific fuel consumption SLST = Sea level static thrust SPF/DB = Super plastic formed/diffusion bonded

SPL = Sound pressure level

SRLT

= Silicone rubber laminating tool -- Wing area

S w

t

= lemperature or thickness = Thickness/chord ratio

t/c, T/C

= Ti tani um

TI, ti

TKS

= TKS Company, Cumberland, England

TOGW

= Takeoff gross weight

T/W

= Thrust/weight ratio

U+L

= Suction on upper and lower surfaces

U.S.(O) :: Upper surface (only)

V

= Aircraft speed -- Aircraft cruise speed Vc VCK = Variable can_ber Krueger flap = Initial droplet velocity Vd = Aircraft design diving speed VD = Horizontal tail colume coefficient VH = Local velocity through porous surface V_ = Minimum control speed Vmcg = Average velocity through porous surface VN, VW = Vertical tail volume coefficient Vv = Free stream velocity V = Suction air mass flow Wa W_ = Fuel flow or consumption i = Chordwise distance from L.E.

X = Spanwise distance from aircraft CL along wing XW = Distance from L.E. to boundary layer transition XTR YEHUDI = Wing root trailing edge extension = Angle of attack = Flap angle 6 F = Relative pressure _(t) e = Meniscus angle of liquid against surface = Relative temperature

°(t)

= Height of insect deposit n = Efficiency or fraction of semi-span from aircraft CL X, A = Sweepback angle = Viscosity = Kinematic viscosity , P = Density c = Relative density of air ¢ = Angle of nozzle to wing reference plane f--,

©

4.0 MISSION DEFINITION In accordance with the NASA Contract NAS1-14632, the selection of the aircraft mission and passenger sizing to be considered throughout this LFC study is the contractor's choice. The aircraft is assumed to be operational in a 1990 - 1995 time period.

The mission selection was guided by Douglas Marketing analysis based on the passenger traffic of 140 airlines and cargo traffic of 62 airlines. The airlines represent over 95 percent of the !CAO world passenger and cargo traffic, excluding USSR and China. The market analysis and prediction of the 1990-95 passenger aircraft demands included the following assumptions: Typical airline route structure would not change substantially over the forecast period; Retirement of the current fleet is considered to be after 15 years of service: Types of aircraft selected to fill additional needs are normally the same as those currently in service with each airline; Production of new aircraft types is based on - • Speed and utilization of aircraft replaced, • Assumed seating arrangements of the new aircraft, • Assumed increase in load factors and yields support the presumption that the airlines can raise capital to purchase equipment to meet passenger demand. Over the long term, the commercial aircraft industry is assumed to maintain its ability to meet the world demand for aircraft• ;_ --_r=_._ -, r-_ : r_L:_L,_,_u FAG£ BL.AS:K NOT FILMED The anticipated seat/distance demandby aircraft types, as a function of time,

(Figure 4-I), shows the major demandin the 1990 time period to be for

long-range jet and the short/medium range jet types. Required seating

capacities are estimated at 250-425 and 180-380, respectively. For purposes of

this study, a 300-passenger/9260 km (5000 n. mile) range aircraft was selected.

The rationale for this selection as a design point for the LFC base case

aircraft is exemplified in the aircraft development trend lines summarized in

Figure 4-2. In the time period of 1990-95, it is shown that an LFC aircraft

of the 300-350 passenger size class could replace a large group of current

aircraft such as the B747SP,DC-IO, DC-IO Stetch, L-I011, DC-8 Stretch, DC-8,

B-707, B-727-200, and the A-300. The estimated time periods required

for development of the aircraft are indicated by heavy horizontal lines which

span a 7-9 year period.

The aircraft sizing and mission ground rules for this study are given in Table

4-I below.

©

OF POOR QUAL|TY 2OOO 2OOO SUPERSONIC--

STO'_FEEDER JETS-, _

1200 1200 AVAILABLE SEAT MILES (BILLIONS) SEATING CAPACITY -- 250,425, I_ 80O 8OO __ RT-/MEDIUM-RANGE JETS (2')_ , ,_ SEATING CAPACITY -- 180-380 _\_ 4OO LONG-RANGE JETS_) ""' ' ........ ' _'_ _"_\\_\\\\ _\\\\\\\\\\\ ;,i ,, 400 ........ AND (_) tJ P,OP : 1975 _ '1977' '1979' '1981' 1983 1985 1987 1989 1991 * EXCLUDES U.S.S.R AND CHINA YEAR ENDING "_ STANDARD-BODY TYPES WIDE-BODY TYPES FIGURE4-1. SEAT-MILE DEMAND BY AIRCRAFT TYPES 4OO 747 • o.oeoo.o.oel '*-I°e •"'_-'CF OI ADVANCED TECHNOLOGY ENERGY- PASSENGER __EFFICI DERIVATIVE DC-10 .... " ......

=.,,,.

CAPACITY 300 L-1Ol 1 0... A_Ol%747SP ..- ENERGY-EFFICIENT =e._o. oe*°° SHORT HAUL _..e°°!

oe.q ,-°''- 200 DC-a STRETCH_.,,.-'" .I!1,,Z...I.7t_ ._,..,. ......

OC-8 .....-" '°:....':_ ............... ........... --_-;;;; 100 _ ..............

1955 60 65 70 75 80 85 90 95 1OO YEAR 7-GEN-22946"I

C _

FIGURE4-2. AIRCRAFT DEVELOPMENT TREND LINES Table 4-I Aircraft Sizing and Mission Ground Rules Takeoff distance 3,050 m (10,000 Ft) Max.

67 m/s (130 Knot) Max.

Approach speed M =0.8 Cruise Speed crui se Initial cruise altitude (LFC off) 10,670 m (35,000 FT) Step climb cruise altitude increments (Standard practice) 1,220 m (4,000 Ft) International Standard Reserves (LFC off) The takeoff distance restriction of 3,050 m (10,000 Ft) feet is consistent with available airport runways from which long-range aircraft are expected to operate.

The 67 m/s (130 Knot) approach speed ensures growth capability with acceptable landing characteristics.

The cruise Mach number of 0.8 and the step climb cruise procedure are consistent with current commercial aircraft operating practice.

The selection of 10,670 m (35,000 ft) as the initial cruise altitude is to reduce the probability of encountering ice crystals in the atmosphere that can cause loss of laminar flow. Ice crystals are encountered more frequently at lower altitudes. The initial cruise altitude and the international reserves are calculated assuming LFC off. This is definitely conservative but would minimize the impact of losing LFC on aircraft operation.

5.0 CONFIGURATION DEVELOPMENT

5.1 CONFIGURATION STUDY GROUND RULES

In order to make a logical and meaningful evaluation of the laminar flow

control aircraft, a side-by-side comparison with a comparable turbulent

aircraft is necessary. Therefore, throughout the study, weight, performance,

and economics data are generated for both laminar flow control and turbulent

aircraft. The laminar flow and turbulent aircraft are configured

to the sameground rules, see Table 5-1, and to the same level of technologies

except for laminar flow control. The turbulent aircraft provides a basis for

evaluation of the benefits and performance gains associated with the LFC.

The interior arrangement of the basic 300-passenger fuselage, shown in

Figure 5-1, is the same for both the turbulent and the laminar flow cases.

DC-IO loftlines are used with a fuselage extension of 2.03 m (80 in). The

cabin is actually identical to a Japan Air Lines International mixed class

interior, with the exception of the replacement of one lavatory by two seats, and the addition of two rows of seats for the lengthened fuselage.

As on the DC-IO commercial transport, cargo is carried under the floor with

cargo containers aft of the wing and pallets forward. In the design weight of

the aircraft these cargo-carrying provisions are considered to be onboard but

empty, except for passenger baggage. The aircraft structural weight is

compatible with carrying a full cargo load. Fuselage under-the-floor volume

is sufficient to carry four extra cargo containers if desired.

---. 24 FIRST CLASS -- 965- and 990-mm (38 AND 39 IN.) SEAT PITCH F 275 ECONOMY -- 864-mm (34 IN.) SEAT PITCH 45.94 m (1848 IN.)

l] -0t_ _1-2_t116 FIGURE 5-1. INTERIOR ARRANGEMENT - MIXED CLASS ORIGINAL PAGE IS OF POOR QUALIT_, f. :o Table 5-I BASIC STUDY GROUND RULES PERFORMANCE CHARACTERISTICS 31,300 Kg (69,000 LB) - Payload (300 Passengers + 10% Cargo) - Range 9,620 km (5,000 N MI) - Cruise Mach No. 0.8 - Field Length 3050 m (I0,000 Ft) MAX - Approach Speed 66.9 m/s (130 KN) EAS MAX 0 LFC - Porous Suction Surface ADVANCED TECHNOLOGY - Consistent with 1995 operation - Advanced Airfoils - Advanced Composites - Advanced Engines - Active Controls 0 COMPARE LAMINAR & TURBULENT ADVANCED AIRCRAFT For the initial base case LFC aircraft, the first 70 percent of both surfaces of the wing is assumed to be laminarized. This is compatible with the use of a conventional trailing edge high lift system. The initial LFC surface is a porous glove concept composed of a sandwich of 50 x 250 Dynapore stainless steel outer surface supported by honeycomb and Kevlar epoxy stiffener materials. No laminarization is assumed for the fuselage and tail surfaces.

5.2 CONFIGURATION INTEGRATION STUDIES Four major trade studies were made in the development of the final LFC aircraft configuration, as summarized below: Study I Initial preliminary design of 1990-95 laminar flow control (LFC) and turbulent aircraft configurations, feasibility and performance comparisons - Model No. 3128 (Laminar Flow) Model No. 3127 (Turbulent Flow) 4-engine arrangement Suction over 70 percent chord on both wing surfaces for LFC.

See Section 5.3.

Study II Trade study of four-engine vs three-engine configurations resulting in selection of the three-engine arrangement. See Sections 5.4 and 7.2.2.

Study III o Trade study of four vs two wing mounted suction engines resulting in the selection of two suction engines. See Sections 5.5, 7.2.3 and 10.6.

Study IV Comparison of effects of extent of laminarization - upper and lower airfoil surface suction to 70 percent chord compared with upper airfoil surface only to 85 percent chord comparable turbulent aircraft all 3-engine arrangements Three-views of the above-mentioned configurations, the associated performance characteristics, and the weight comparisons are presented in Section 5.6.

Discussion of these studies and pertinent subsystem descriptions for the LFC and turbulent aircraft follows. The detail substantiation of subsystem design may be found in the other sections of this report dealing with particular disciplines. However, discussions concerned with the integration of the subsystem into the aircraft configuration are included in this section. The economic evaluation of the laminar flow control versus turbulent aircraft should be based on the final configurations only, as presented in Section 6.0.

Douglas consulted with United Airlines (passenger) and with The Flying Tigers (freight/cargo) airlines during the course of these studies. Their comments are included under operational aspects, Section 5.7.

C !

5.3 STUDY I - INITIAL BASELINE PRELIMINARY DESIGN OF LAMINAR FLOW CONTROL

(LFC) and TURBULENT AIRCRAFT CONFIGURATIONS

5.3.1 General

The initial baseline laminar flow aircraft serves as a configuration into which "first-go-round" feasible LFC subsystems may be integrated. Trade studies are then performed to evaluate the merits of the LFC subsystem designs.

Updating the base case LFC configuration, as the study progresses, assures that the recommended subsystems such as the airfoil section, LFC structure, LFC suction and manifolding systems, suction compressor drives, sizing, and locations are all compatible and capable of integration into an efficient practical aircraft. For the initial baseline laminar flow aircraft, only the first 70 percent of the wing chord on both surfaces is assumed to be laminarized; there is then sufficient room for a conventional trailing edge high lift system to be installed. A porous suction surface concept is used for laminarization. All fuel tankage is in the wings. Three-views of the aircraft and the initial study ground rules are given in Figures 5-2 a & b and F" Table 5-2 respectively.

\< Discussions of configuration characteristics, rationale for selection, and the effects of technology assumptions on aircraft sizing and performance characteristics are included in subsequent paragraphs.

5.3.2 Aerodynamics a) Wing Planform - The wing planform characteristics selected for the base case aircraft are as follows: Laminar Tu rbu Ient Sweep angle @ I/4 chord 30 o 300 Taper ratio .3 .3 Aspect ratio 12 12 2O WING HORIZ VERT /_ '- i Y f AREA m2(FT 2 ) 381 (4100) 64 (693) _) //_

SWEEP _ _ 40" / r_

..,c_...._o o.,,._,,,,o o.,., o.,., / ]_ _ _

DIHEDRAL 4" -_ -- _ - - _17._ m

1202 IT) -I 7-GEN-22 726 -3 FIGURE 5-2a. INITIAL LFC BASE CASE AIRPLANE WING HORIZ VERT AREA m 2(FT2) 267 (2870) 89 (960) 46 (495) ASPECT RATIO 12 4 1.6 TAPER RATIO 03 0.35 0.35 SWEEP 300 300 350 THICKNESS RATIO 0.132 AVG 0.10 0.10 DIHEDRAL 40 70 -- TAIL VOLUME -- 1.52 0.078 56.56 m (185.58 FT) nn m == 56.82 m (186.42 FT) 7-GEN-22727 "1 FIGURE 5-2b. INITIAL TURBULENT BASE CASE AIRPLANE Table 5-2 INITIAL BASELINE STUDY GROUND RULES GENERAL - 2.5 maneuver load factor - limit Design N - I0 FPS at design landing weight Design Sink Speed - Gust and maneuver load alleviation Active Controls - Relaxed static stability WING --'--Ai rfoi I Sect i on - Turbulent-Supercritical; Laminar-Shock-Free L.E. Devices - Turbulent-VCK; Laminar-None T. E. Devices - Flaps, Ailerons, Spoilers FUSELAGE - Circular Cross Section Pressurization - 60kPa (8.7 PSI) LANDING GEAR - Hard Surface Field - Carbon Brakes Advanced Materials POWER PLANT - Advanced Technology Turbofan (4) Engines -7.0 Engine T/W - Long Fan Duct Length Thrust Reversers - Fan, primary -6 By Pass Ratio FUEL SYSTEM Fuel Tanks - 4 Main tanks plus center wing all integral - 778.9 kg/m 3 (6.5 Ib/gal) Fuel Density FLIGHT CONTROLS & HYDRAULICS Actuators - Integral servo pumps - Titanium Plumbing Controls - Fly-by-wire P res su re - 27.6 MPa (4,000 psi) AUXILIARY POWER PLANT Unit - Advanced technology APU Location - Aft fuselage INSTRUMENTS - Fiber optics Advanced Technology ELECTRICAL SYSTEM - VSCF Generators Primary Power - VSCF Generators on APU Supplemental Power - E-Mux Circuitry AV IONI CS - Mini-comp, A-Mux Circuitry - DC-IO type Cool i ng ICE PROTECTION Protecti on - Engine inlet, wing L.E., windshield The wing sweepback of 30 o is consistent with current Douglas turbulent advanced transport studies and is selected on the basis of compatibility of i the wing thickness permissible for an operational CL of .5 to .6 considering the internal volume available for landing gear stowage and the suction manifolding installation. Preliminary checks with a sweep angle of 25 o and the associated thinner airfoil sections indicated that volume limitations may exist.

As shown later under item 5.3.7, preliminary aspect ratio trade study results show that, for both the turbulent and laminar flow cases, aspect ratios of 12 and 14 gave essentially the same minimum takeoff gross weight. Consequently, the lower aspect ratio of 12 which gave more space for suction ducting, was selected.

The planform trailing edge fairing or "yehudi", required to submerge the landing gear strut within the wing depth, extends from the fuselage intersection to 42.8 percent semispan. Definition of the airfoil thickness at the fuselage wing intersection is based on the extended chord including the "yehudi" rather than on the trapezoidal wing chord.

b) Airfoil Section - On the basis of the supercritical and advanced airfoil work which has been accomplished and is underway at Douglas, the following airfoil sections are selected for the base case configurations: (See Section 7.1) Laminar Turbulent Airfoil Section Shock-free Supercritical Root t/c DSMA 679 Type/13.3% DSMA 679 Type/15% Aero Break t/c DSMA 691Type/11.1% DSMA 684 Type/12.5% Tip t/c DSMA 691Type/10.5% DSMA 686 Type/t1.8% .50 .64 CL Design A comparison of the base case airfoil contours, at three spanwise locations is presented in Figure 5-3. The airfoil thickness is compatible with buffet-free operation at the design CL. Figure 5-4, which shows the _CLbuffet variation with M/Mdivergence for the turbulent aircraft, Inalcazes that LAMINAR FLOW CONTROL _DSMA 679 TYPE -- T/C = 15 PERCENT ........ TURBULENT FLOW r/ FDSMA 679 TYPE -- T/C = 133 PERCENT FUSELAGE _ ................ r .........

INTERSECTION -(,. ..... - - 0.093 u/2 ............

I I I I I L J I ] ] J DSMA 684 TYPE -- T/C = 12.5 PERCENT AERO BREAK ___ DSMA 691 TYPE -- T/C = 11.1 PERCENT 0.4283 b/2 L L I I i I I _ I I I DSMA 686 TYPE -- T/C = llJ_ PERCENT TIP I_A 691 TYPE--T/C = 105 PERCENT 1.00 b/2 I i I I I 1 I [ I I I 0 20 40 60 80 100 PERCENT CHORD ,-_E_-22_,, FIGURE 5-3. BASE CASE AIRFOILS ASSUMED ORIGINAL PAGE i9 OF POOR QUALITY 1.2 1.0 0.8 C L BUFFET 0.6 0.4 • I I I 1 I I 0.75 0.80 0.85 0.90 0.95 1.00 1.05 M/M DIVERGENCE AT C L = 0.5 83_N-23278

©

FIGURE 5-4. BUFFET BOUNDARY FOR TURBULENT BASE AIRCRAFT CLbuffet is .9 or more in the operating regions of interest. In the case of the laminar flow aircraftp the C L of .5 is a more limiting des ign condition than buffet; the corresponding C L variation for the buffet laminar flow case is therefore not defined. The airfoils selected for the laminar aircraft are designed to be shockfree at the design CL, with no significant amount of separation even if turbulent flow should occur on the laminarized portion of the airfoil. More than sufficient buffet margin exists with these shock-free airfoils.

The turbulent airfoils are designed to provide satisfactory compressibility drag characteristics at the cruise Mach number of 0.8 as well as an adequate margin before buffet onset. Figure 5-5 summarizes the effects of average winq thickness and airfoil design on the design cruise CL with a wing sweep of 30 ° . For the LFC wing, an operational CL of .5 corresponds to an average wing thickness of 11.5 percent. The use of advanced supercritical airfoils for the turbulent design allows a thicker wing to be used for a given CL (for example, CL = .6 allows an average wing thickness of operating 14 percent). Figure 5-5 which is derived from previous Douglas aerodynamic studies, clearly illustrate the rationale for selection of a thinner wing for laminar flow aircraft than for the turbulent. The aircraft sizing work of this study shows that the design CL of .5 is essentially optimum for cruise the base case laminar aircraft.

The maximum thickness of the wing varies spanwise, as shown in Figure 5-6.

The thicknesses from the fuselage intersection to the aero break are based on the total wing chord. Relative to the trapezoidal planform chord, the root aircraft thickness would be 22 percent.

Wing Twist - The wing twist distribution, based on Douglas studies of similar wings, is as shown in Figure 5-7 for both the laminar and turbulent aircraft.

For these preliminary layouts, no distinction is made between the laminar and turbulent twist distribution.

ORIGINALPAGE 14

C'}FPOORQUALITY

30-DEG SWEEP M CRUISE = 0.8 0.8 0.7 0.6 CRUISE CL 0.5 0.4 0.3 8 9 10 11 12 13 14 15 AVERAGE WING t/c 7-GEN-22651 NOTES: (1) AIRFOILSDESIGNEDTO HAVE LAMINARFLOWTO 70% CHORD (2) AIRFOILSHAVEATFACHEDTURBULENT FLOWWITH LFC OFF FIGURE 5-5. CRUISE C L VS WING THICKNESS/CHORD RATIO /_ c/4 ----- 30 DEGREES MICRUlSE = 0.8 k- Z gJ p,- I,LI % O. 14 I LAMINARFLOW (MODEL 3128) / C L DESIIGN = 0.50 p- TURBULENT (MODEL 3127) C3 C L DESlIGN "_ 0.64 O _'_I ........

= U'J w Z

\

-r 14 -J v I p-

.J I

t/c BASED ON TOTAL CHORD INCLUDING "YEHUDr' < O

I

.J 0 10 20 30 40 50 60 70 80 9O 100 7"GEN-22647-| PERCENT SEMISPAN FIGURE 5-6. PRELIMINARY WING THICKNESS DEFINITION Empennage Sizing - The size of the horizontal tails for the base case aircraft is influenced by the use of active controls compatible with a 1990-95 time period. The horizontal tails are not sized as is usual by stability considerations at the aft c.g. limit, but by the criteria given in Table 5-3 below.

Table 5-3 TAIL SIZING CRITERIA LFC AIRCRAFT Horizontal Control for Nose-Wheel-Liftoff at 1.15 Vstal 1 (Elevator deflected) Vertical Directoral Stability TURBULENT AIRCRAFT Horizontal Trim at 1.4 Vstal I in landing approach with ice on the tail Vertical Control, with full rudder, at 56.6 m/s (II0 KN) minimum ground control speed The stringent requirement imposed on the turbulent aircraft horizontal tail is due to its powerful high lift system and associated high pitching moments; the nose wheel lift-off criterion is less critical. For both aircraft, the amount of negative stability permissible is limited by the physical arrangement of the aircraft such that the tip-over requirements are satisfied. Tail sizing criteria and aircraft balance for each configuration are discussed separately in the following paragraphs.

el Laminar Flow Control Aircraft - Considering the aircraft balance associated with the four rear engines, the T-tail installation and f_ the landing gear stowage requirement in the wing, the main landing gear is located at the 60 percent wing MAC. With this arrangement, the aft c.g. is limited to 39 percent wing MAC in order to prevent the c.g. from moving aft of the main gear in the tip-over attitude (tai l-on-ground) With aft-mounted engines, the aircraft c.g. and the wing are further aft relative to the c.g. of the payload carried in the fuselage. The required c.g. travel is therefore greater than with wing mounted engines and it is estimated that a c.g. range of 28.7 percent MAC is appropriate for the LFC aircraft. This percentage MAC corresponds to a 69.6 inch c.g. travel or 3.3 percent of the fuselage length. With this c.g. range, the forward limit is at 10.3 percent wing MAC. The horizontal tail is then sized to provide sufficient longitudinal control (with elevator deflected) for nose-wheel-lift-off (NWLO) at 1.15 Vs . With this horizontal tail, neutral stability occurs at 21 percent MAC, and at the aft c.g. limit of 39 percent MAC, the aircraft has 18 percent negative stability.

The vertical tail is sized to provide adequate directional stability since the minimum ground control speed is less than 56.6 m/s (110 knots). The relatively simple high lift system assumed for the laminar flow configuration does not impose sizing requirements on the tail.

be Turbulent Aircraft - Considering the balance of the aircraft with four wing-mounted engines, the landing gear is placed at 71 percent wing MAC. (As indicated earlier by Figure 5-3, the thicker supercritical airfoil provides adequate stowage for the landing gear strut at a more rearward % MAC than for the laminar flow aircraft.)

The aft c.g. limit for this turbulent configuration is also established by the tipover limit and is at 47.5 percent wing MAC. A c.g.

travel of 20 percent MACis judged appropriate for this aircraft, equal to 41.6 inches. This is consistent with DC-IO aircraft c.g.

travel in terms of percentage of fuselage length.

Critical factors in sizing the horizontal tail are:

Static longitudinal stability is provided at the tail-on

(dCm/dC L = O) neutral point located at 42.5 percent MAC, at a

speed midway between Vc and VD (assumedM = .77 @4,570 m

(15,000 Ft) altitude).

At the forward c.g. limit, the horizontal tail size must be adequate to provide longitudinal trim at 1.4 Vstal I in the landing approach with ice on the tail. This stringent requirement is due to the powerful high lift system with associated high pitching moments. The nose wheel liftoff criterion is less critical.

With the forward c.g. limit at 22.5% MAC, the tail sized for neutral stability at the 42.5% MAC, and the aft c.g. limit allowed to go to the tip-over limit of 47.5%, the aircraft has 5 percent negative stability. It should be noted that this modest amount of negative stability is the maximum physically possible with this configuration.

Active Controls - For these base case aircraft, except for the yaw damper, active controls are not considered in the vertical tail sizing. Rationale for this decision is based on the safety consideration that with active controls on both the longitudinal and directional axes, a failure on both axes could result in the pilot not being able to control the aircraft.

A parametric tail sizing study was performed in the early stages of configuration definition. Variations of tail volume, c.g. location, and c.g.

",L range as a function of wing area and aspect ratio as applied to the LFC and the turbulent aircraft were taken into consideration. Maneuver/gust load alleviation was not included. Douglas experience indicates that other criteria become critical before any significant advantages of maneuver or gust load alleviation can be realized. For strength-designed aircraft, gains of I to 2 percent are realizable in some cases. The decision not to consider maneuver/gust load alleviation on these base case aircraft was validated since the wing stiffness required to maintain aileron effectiveness was shown to be more critical than strength for the wing structure. See Section 9.1.6.

Low Speed High Lift System - With LFC limited to 70 percent wing chord, a conventional trailing edge high lift system can be used. Either a single slotted extensible flap or a double slotted flap are capable of providing the high lift characteristics that are given in Figures 5-8 and 5-9. Leading edge devices for high lift are not considered feasible for this LFC configuration.

The landing flap angle is limited to that for which 1.3 Vmi n occurs at zero angle of attack to avoid landing on the nosewheel first. For the base case laminar aircraft, this is a flap angle of 22.5 degrees, Figure 5-8.

4 ¸ The low speed high lift system assumed for the base case turbulent aircraft is an advanced system consisting of both a variable camber Krueger (VCK) leading edge device plus trailing edge flaps. Characteristics of this high lift system are shown in Figure 5-10.

5.3.3 Structural Design With the exception of the LFC features, the structural design of the two base case aircraft is conventional but includes advanced materials and concepts appropriate to the 1990-95 operational time period. The design life, relatively crack-free, is 60,000 hours with a scatter factor of 2, or 120,000 hours for fatigue. The design load factor is 2.5 g. Section 9.1 gives details of structural concept studies.

3O ( r ¸, = 30 DEG Ac/4 = 0.8 MCRUISE TWIST (DEGREES) -2 -4 I ] I I I I I I I 9O 100 0 10 • 20 30 40 50 60 70 80 SEMISPAN (PERCENT) 83-GEN.23272 uc594 ¢1_oi-2 FIGURE 5-7. PRELIMINARY WING TWIST DISTRIBUTION _! =50 DEG

/// oo o

///,,_/// _V°,n

AR 14\ _ ---_r/ ,_" /// -20 -16 -12 -8 -4 0 4 8 12 16 ANGLE OF ATTACK (DEG) ;P-GEN-22859 FIGURE 5-8, LIFT CURVES FOR LFC AIRCRAFT OF POOR _dl-+.LirY' FLAP DRAG CHARACTERISTICS FLAP LIFT CHARACTERISTICS AR i4 LANDING I0 CDtotal = |basic +ZICD S.

|4 AR w 4.0 1.5 3.0 CL untrimmed CL2 2.0 1.0 12 i "1 ARw _ 1.O 10 15 25 ° FLAP DEFLECTION 20 I , I I 0 I I I I 0.04 0.08 0.12 0 10 20 30 40 50 0 FLAP DEFLECTION,(DEGREES) FIGURE 5-9. PRELIMINARY FLAP CHARACTERISTICS FOR LAMINAR FLOW CONTROL AIRCRAFT Ac/4 = 30 DEGREES,AR = 12, VARIABLECAMBERKRUEGER EXTENDED

,oI

I 3o.o

3.0 5/11/0 C L TAILOFF 2.0 .

1.O __'_J 1.3 v s 0 I I 1 i I 0 4 8 12 16 20 24 ANGLE OF ATTACK, o_ T (DEGREES) B3-GEI¢23275 _594 mlr 01*4 G_ FIGURE 5-10. ESTIMATED TAILOFF LIFT CURVES FOR TURBULENT BASE AIRCRAFT Composite Material Utilization - Advanced composite materials are used extensively as described in Tables 5-4 a and b, consistent with the current composites development programs. In the case of the wing and tail components, composite materials are used whenever practical but all attachments and fasteners are metal. The effect of this use of composite materials on component weights is shown in Table 5-4c.

In the early design study phase, composite utilization in the fuselage was subject to a trade study. Initially the optimistic assumption of 80 percent composite material was considered because NASA planning in the ACEE program originally included the development of a composite fuselage. However, it is now doubtful that the development of a full-scale fuselage will take place in time for utilization on a 1990-95 design and a much lower percentage of composite materials was therefore assumed. The effect on aircraft sizing of the more conservative assumption (15 percent composite shell panels and 0 percent composite frames) was checked and the effect on the turbulent aircraft weights_for example, is as follows: A Weight empty +2244 kg (+ 4,948 Ib) A Takeoff gross weight +5806 kg (+12,800 Ib) +14.4% Z_ Fuselage structural weight The effect of these assumptions of composite usage in the fuselage on the LFC aircraft takeoff gross weight and sizing is shown in Figure 5-11.

..... ;_;W MATERIAL UTILIZATIOtl CO),;PO'S ITE t.IETAL COMPONEiIT BASIC COt¢STRUCTI Orl T'YPE'" _TYPE WING Skin Panels Skin/Stringer Gr/Ep lO0 .Al 20 Stiffened Web Spars Gr/Ep 80 Al lO Ribs Stiffened Web Gr/Ep 90 Hul ti-Rib Al lO Leading Edge Gr/Ep 90 Al lO Multi-Rib Trailing Edge Gr/Ep 90 Ti 30 Slats Hul ti-Rib Gr/Ep 70 Al 20 Multi-Rib Flaps Gr/Ep 40 Ti 30 Kev/Ep lO Al 20 AiIerons Multi-Rib Gr/Ep 65 Kev/Ep 15 Multi-Rib Al 20 Spoilers Gr/Ep BO Multi-Rib Al 20 Wing Tips Kev/Ep 80 Hul ti-Rib 0 Fairings Kev/Ep lO0 (!

Multi-Frame Ti lO Gr/Ep 80 Pylons A1 lO AI lO Nacelles Multi-Fran_ Gr/Ep 90 AI HLG Doors Hul ti-Beam 2O Gr/Ep 65 Kev/Ep 15 VERTICAL TAIL Skin Panels 0 Skin/Stringer GrlEp lO0 Stiffened Web Spars Gr/Ep 80 Al Stiffened Web lO Ribs GrlEp go Al lO Multi-Rib Leading Edge Gr/Ep. 90 Al Multi-Rib I0 Gr/Ep go AI Trailing Edge Rudders Multi-Rib 20 Gr/Ep 65 A1 Kev/Ep 15 ,Mul ti-Rib 2O Kev/Ep 80 Al Tip |, I i i i, TABLE 5-4a. Structure and Material Assumptions 1990-95 Aircraft COmpatible with (Continued).

MATERIAL UTILIZATION COMPOSITE BASIC CONSTRUCTION I NETAL COI,IPONENT TYPE % TYPE HORIZONTAL TAIL Skin Panels Gr/Ep Skin/Stringer 80 A1 Stiffened Web. Gr/Ep Spars I0 90 A1 Ribs Stiffened Web Gr/Ep I0 9O AI Multi-Rib Gr/Ep Leading Edge A1 I0 Multi-Rib Gr/Ep.

Trail ing Edge 65 A1 Elevators Multi-Rib Gr/Ep Kev/Ep 80 A1

Mul if-Rib zO

Kev/Ep Tip FUSELAGE f 15 A1 Shell Panels B/Ep Skin/Stringer I00 0 A1 Stiffened Web Frames 80 A1 Floor Panels Stiffened Panel Gr/Ep I0 90 A1 FI oor Beams Stiffened Web Gr/Ep I0 90 A1 Gr/Ep Keel Stiffened Web 65 A1 Multi-Beam Gr/Ep Doors & Stiffened Panel Hatches Kev/Ep I00 Multi-Frame Radomes Kev/Ep LANDING GEAR: 25 Ti 75 Beam Fig GrlEp Torque Links Ti 60 Gr/Ep Monocoque Tube Drag Links Ti 60 Gr/Ep Side Brace Honocoque Tube ,P TABLE 5-4b. Structure and Material Assumptions Compatible with 1990-95 Aircraft Table 5-4C COMPOSITE FACTORS COMPOSITE WEIGHT FACTOR* Wing Box .73 Leading Edge •73 Trailing Edge •73 Wing/Fuselage Fairing .70 Tips •76 Ai I erons .76 .85 Flaps Slats .79 Spoilers •76 Horizontal Tail Bending Material .70 Spar Webs .76 Ribs .73 Leading Edge .73 Trailing Edge .73 Elevator .76 Tips •76 Fairing .85 Vertical Tail Bending Material .70 Spar Webs .76 Ribs .73 Leading Edge • 73 Trailing Edge .73 Rudder • 76 Tip • 76 Fuselage Shel 1 •96 Ra dome •70 Wing & Landing Gear .85 Support Tail Support .85 Cockpit Floor .74 Main Cabin Floor .74 Entrance Doors .76 Cargo Doors .76 Enclosure .95 Landing Gear Doors •76 Miscellaneous •88 Propulsion Engine Cowl .75 Factor is a ratio of (composite and conventional material ) to (all conventional material)•

©

A comparison of the effect of the finally selected use of advanced composite

in both the laminar flow and turbulent aircraft is given in Table 5-5. The

result is 15 to 18 percent reduction in structural weight over that of the

all-metal aircraft. In terms of manufacturer's weight empty, the weight

saving due to composite usage is 9 to I0 percent.

Structural Design/Integration - LFC Win 9 - The wing outer surface is a porous sandwich glove concept. It is attached to the corrugated main wing hox by self-sealing fasteners in press-fit bushings. The porous cover is thus removable without affecting the primary structural integrity. This simplifies maintenance activities such as cleaning, inspection and repair of the internal st ructu re.

The design selected initially for the LFC outer glove panel is shown in Figure 5-12. The porous surface is supported by the corrugated main box panel creating an integral spanwise duct for suction air collection. The porous surface is of woven stainless steel Dynapore and the corrugations are of graphite/epoxy composite material. The diffuser is a vacuum formed thermo-plastic sheet that provides plenum compartmentation and airflow metering.

5.3.4 Suction/Manifolding System General Arrangment - As shown in the three-view, Figure 5-2a, four compressor drive units are used on the initial baseline LFC aircraft. Final selection of the number of these compressor drive units was the subject of a separate trade study, Sections 5.6, 7.2.3, and 10.6, resulting in the selection of only two units. At this stage, however, the use of four units was based on the followinq design rationale: The design of the manifolding must maintain high efficiency throughout; therefore, the ducting must have gentle bends and the losses of momentum or velocity kept to a minimum.

.c t-, t.- _ r._ c_', oC, I-....

ii t-, ,_ t- el.f) co oo cx_ la._ co oo t't3 l-- z I l.l.J ,_J I-- ¢,,.)

.--J oo I •-J I-- o 0_ ._ _- o_ oo _ _ I.Ll i II "I- l,/3 l.,,3 Ch I-- I.I.I e,,J Q 00 -..J I e,,."

Z _0 !

.-t I--- ¢_,, _.0 Ud,_ ,._I ..J Zl,1 C_I--- .-J _ ..-J ;---_0 I-- .._J i,i (_) (-JO z _ c.o N _ .--I Z U.JC_ Z _" _" i_O (..5 OF POCI: _._.,._ i'Y ..... 66.9 m/s (130 KN) APPROACH SPEED

_90 -(,2o)

i .............. 3D50 m (10;000 F'r) TAKEOFF DISTANCE COMPOSITE FUSELAGE (4101 Sw = 390.2m 2 ASSUMPTIONS 18f - (4,200 FT"2 )_......_._.__.__ CONSERVATIVE

f----'_:-=_.__ ___

TAKEOFF GROSS WEIGHT OPTIMISTIC (3901 " -- .... 1" ....

__wS =297.3m212 _ 14 J" 12 14 1000 Ilg AR (3701 = (1000 LB) BI=GEN-223Q_ FIGURE 5-11. EFFECT OF COMPOSITE USAGE IN FUSELAGE ON LFCAIRCRAFT TAKEOFF GROSS WEIGHT (NONSTRUCTURAL) 126 2 _\DIFFUSER 1"---'-- 14.97) _ Q _\ . POROUS SANDWICH i,__,\ /,NONSTRUCTURA', OUTBOARD SUCTION ENGINE _----_1_ _'L _ _, I_ _,__--..--., LOCATION L304 '(i.-_; POROUS CSANDWICH _ (7.52I STRUCTURAL BREAK LOCATION _ i (0.'375) \DIFFUSER I I\\ //1 -.s k\ // 7.2,

L=o_ L-,:.',;,-J L,,%'_

{16.50) '1 FUSELAGE INTERSECTION LOCATION POROUS SANOWICHI DIFFUSER 7 $2 12 7 , .4 I / ,,o._,,, _,o._oo, d142-2 7*GEN.22&S3 I \ FIGURE 5-12. PSEUDO-ISOTROPIC COMPOSITE CORRUGATION To maintain high efficiency in the suction system, it is desirable that the air from the surface collection points be individually ducted as far as possible before being ducted to the suction pump.

The individual ducting arrangement avoids large losses in momentum, or velocity, which are associated with a plenum chamber configuration.

The internal ducting area available with corrugated stiffening is compatible with a four unit suction pump system; one suction pump per side would require larger individual ducts incompatible with the outboard wing structure.

Four suction pumps versus two pumps reduces the effect of single pump failure. In the case of four suction pumps, the malfunction of one pump should result in loss of LFC over only one fourth of the wing surface rather than one half of the surface.

The four suction pumps are located spanwise under the wing such that the areas to be sucked are divided equally. The suction air is collected spanwise through the structural corrugations using two collector stations per side.

,f Dry bay areas are used to provide space for the plumbing required for suction air to be ducted from the multiple spanwise corrugations (or flow channels) into mixing/collection chambers ahead of the pump inlets below the wing.

Cross ducting of the suction flow between left and right wings for reliability purposes is not necessary and would incur ducting problems and weight penalties.

Integral ducting is formed by the corrugated wing structural panel and the outer surface. Sixteen spanwise corrugations run, top and bottom of the airfoil contour, tapering from the fuselage intersection to the tip of the wing. The structural/ducting integration is a critical factor in the design of the laminarized wing. In this particular case, the outboard suction engine is the critical location for ducting requirements; here the width and depth of i ' i 4O the corrugations as sized by structural requirements are consistent with the duct sizes required for suction manifolding. Dimensions of the C structurally-designed corrugations at all other stations are in excess of those required.

Ducting requirements are based on a maximum flow velocity of M = .2, assuming 20 percent excess airflow over theoretical requirements during cruise conditions. The resulting duct areas are presented in Table 5-6 below• Table 5-6 SUCTION SYSTEM DUCTING AREAS Location Flow Area Upper Wing Surface cm (in 2) (X/C = .15 to •70) t . Pump Location 322.58 (50.00) . Minimum Flow Points 15.55 (2.41) Lower Wing Surface (X/C : .15 to .70) . Pump Location

364.32 (56.47)

. Minimum Flow Points

15.35 (2.38)

Wing Leading Edge (x/c = o to .15 . Pump Location 276.45 (42.85) • Minimum Flow Points 13.81 (2.14)

Figure 5-13 presents a preliminary layout of the LFC suction air collector

system for the initial base case configuration. This layout is of the dry bay area at the outboard suction pump station. For clarity, the upper and lower wing surface ducting are shown in separate drawings. The compressor drive/suction pump is mounted below the wing such that ducting may be taken through an opening in the lower wing skin.

The air collection ducting is divided into three separate manifolding systems: namely - the upper wing surface or low pressure air, the leading edge ahead of the front spar, and the lower wing surface; the latter are both higher pressure air. The lower pressure air from the upper wing surface is ducted into the low pressure ratio axial compressor where the pressure is increased to the level of that from the higher pressure areas of the wing.

The two flows are then mixed and ducted into a higher pressure ratio compressor that exhausts the air at a velocity equal to the free-stream velocity. Previous studies indicate that efforts for any thrust recovery from such a system are not profitable.

Further details of the manifolding system requirements are described in .

Section I0.0 of this report.

Suction Pump and Drive Unit Sizin 9 - Estimated suction requirements for laminarization, based on aerodynamic calculations, are as shown in Figure 5-14. The amount of suction air and the corresponding sizes for the suction compressors and drive units are thus determined. An axial flow compressor and a turboshaft drive were selected for the base case aircraft. The sizing ground rules are summarized in Table 5-7. See also Sections 7.1.3, 10.4 and 10.5.

Estimated fuel flow requirements are summarized in Figure 5-15. The brake specific fuel consumption is estimated to be achievable with a new technology turboshaft engine having a pressure ratio of 6 to 8.

OR!GINAL PAGE i_ CORRUGATED STRUCTURE OF pOO!;_ QUALITY 4r--_ SPANWISE DUCTING COLLECTOR STATION AT OUTBOARD PUMP \ T-- 7-GEN-22922 FIGURE 5-13. SUCTION PUMP/DRY BAY AREA INSTALLATION M = 0.80, R c = 30.5x 106 , x=30DEGREES,h = 10,970m(36,000FT) 1.4 CQ UPPER SURFACE LOWER SURFACE X/C 1.2 / 0.907797 x 10 -4 0.162702 x 10 -3 0.7 / 0.156214 x 10 -3 0.226288 x 10 -3 0.8 / 1.0 0.29647 x 10 -3 0.330295 x 10 -3 0.9 0.349259 x 10 -3 0.367206 x 10 -3 0.95 m 0.8 CQ (xl,000) 0.6 0.4 _LOWER SURFACE / i 0.2 R SURFACE / I _ "" _ _ "=,, _ _ ,.,g _ _ _ ,_bam.am _ ..d_" " I I I I I I I I I 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 X/C 83_EN-23277 FIGURE 5-14. PREDICTED SUCTION REQUIREMENTS FOR LFC AIRFOIL Table 5-7 SUCTION SYSTEM SIZING -- BASE CASE ° SIZING CONDITION: Altitude = 12,190 m (40,000 Ft) Wing Area = 381 m2 (4,100 Ft 2) Aspect Ratio = 12 Wing Sweep = 30 o M = 0.8 SUCTION AIR QUANTITIES PER AIRPLANE K_/s (LB/SEC) WaUPPER = 3.04 (6.7) WaLOWER = 5.44 (12.0) corrected to pressure at compressor face Wa _ = 26.22 (57.8) t Upper f J" • Wa _ : 34.61 (76.3) _t Lower COMPRESSOR PRESSURE RATIO: P.R. = 1.36 CompressOruppe r P.R. : 1.87

(for F --o)

CompressorMain n TURBOSHAFT POWER REQUIREMENTS PER AIRPLANE At 12,190M (40,000 Ft) Altitude, 555.5 kW (745 HP) Sea Level Equivalent = 1,566 kW (2,i00 HP) 5.3.5 Propulsion System The main propulsion units are advanced turbofan engines. Consistent with engine manufacturers' advanced technology levels, the engine is assumed to have a thrust/weight ratio of 7 and an 11 percent improvement in SFC over present day engine performance. Both of the initial laminar and turbulent flow aircraft are four engined configurations. The turbulent flow aircraft is a conventional wing installation but, the laminar flow aircraft engine installation is an aft-fuselage mounted arrangement. The aft engine mounting is necessary to avoid noise levels at the wing surface inconsistent with laminar flow. Study results of the near field noise environment due to the main propulsion engines (mixed flow CF6-50 type - E3 concept, and a variable pitch fan - QCSEE concept) show that wing-mounted engines are not compatible with achieving LFC over the wing surface, see Section 7.2.

5.3.6 Surface Contamination Prevention/Removal Several alternative systems were being considered at this time and reference should be made to Section 11.0.

\ i 5.3.7 Performance - Initial Baseline Aircraft With the mission and study ground rules listed previously in Tables 5-1 and 5-2, the characteristics of the competitive LFC and turbulent aircraft that were illustrated previously in Figures 5-2a and b, are as presented in Table 5-8.

Estimated weights, based on the methods and assumptions given in Table 5-9, are as listed in Tables 5-10 a and b. Drag estimation is presented in Appendix 5A of this Section.

Geometric data for the sized aircraft are given in Table 5-11 and the resulting performance characteristics are compared in Table 5-12.

Table 5-8 AIRCRAFT CHARACTERISTICS (STUDY I) LFC TURBULENT ENGINE Advanced Turbofan Advanced Turbofan Thrust per Engine (4) kN (LB) 108.40 (24,370) 114.32 (25,700) Installation Aft Fuselage Mount Pylon Mount on Wing WING Area m2 (Sq. Ft.)

381 (4,100) 267 (2,870) 30 30 Sweep Degree 0.3 0.3 Taper Ratio Aspect Ratio Ai rfoi 1 Shockless Supercritical Root T/C 13.3% 15% 11.1% Break T/C 12.5% 10.5% 11.8% Tip T/C 0.5 0.64 CL Design High-Lift System No Leading Edge Device 22% LE VCK !_ ' 25% Double Slotted 35% Two Segment Flap (or single Tracked Flap slotted extensible 80% Span flap) 80% Span Takeoff Distance Sizing Criteria Approach Speed 66.9 m/s (130 kN) 3,050 m (I0,000 Ft) Mi ni mum TOGW Mini mum TOGW Configuration Selection Basis 4B

Table 5-9

WEIGHT ESTIMATION METHODS & ASSUMPTIONS

COMPOSITE MATERIALS USE As described in Tables 5-4a, b and c.

WING Multi-station analysis based on critical gust and maneuver load conditions, Weissinger load distributions, aeroelastic effects, material allowables and weight/geometry relationships.

Gust and Maneuver Load Alleviation: Seven percent wing weight reduction based on past studies.

Flutter: Turbulent Flow Wing: Weight Penalty added to higher AR wings to provide stiffness equal to that of the AR = I0 wing.

Laminar Flow Wing: No penalty with fuselage mounted engines.

<

TAIL Multi-component analysis based on critical gust and maneuver load conditions and weight/geometry relationships.

FUSELAGE Shell Structure: Multi-station analysis based on pressure, critical gust and maneuver load conditions. Pressure critical shell structure is determined by a IP hoop tension stress allowable to prevent skin crack propagation.

Remaining Structure: Wei ght/Geometry relationships.

Gust and maneuver load alleviation: Two percent fuselage weight reduction based on past studies.

LANDING GEAR Empirical weight relationship to takeoff gross weight based on correlation L of several aircraft. Weight reduction due to carbon brakes based on Douglas, Goodyear, Goodrich, and Bendix studies.

Table 5-9 (Cont.)

POWER PLANT Nacelle: Empirical multi-component weight/geometry relationships.

Pylon, Thrust Reverser, and Exhaust: Each component is based on empirical weight/geometry relationships.

In addition, pylon weight considers engine position and supported wei ght.

Engine and Engine Systems: Advanced technology engine weight based on T/W:7. Engine systems weight based on average of several turbofan engines.

Fuel System: Empirical weight relationship to wing span and number of fuel tanks.

Gust and Maneuver Load Alleviation: Pylon weight reduced i0 percent based on past studies.

FLIGHT CONTROLS AND HYDRAULICS System Weight: Empirical weight relationship to total control surface area based on correlation of several transport aircraft with multiple hydraulic systems.

Advanced Technology: Total weight reduced 7 percent based on studies considering fly-by-wire, 4,000 PSI system, titanium tubing, and integral servo pump actuators.

INSTRUMENTS Instrument weights based on DC-IO modified for four engines and advanced technology.

Fiber Optics: Instrument wire weight reduced by 30 percent.

AVIONICS Weight based on DC-IO.

Advanced Technology: Circuitry weight reduced by 50 percent for mini-comp wiring and 10 percent for A-Mux.

Table 5-10a F WEIGHT SUMMARY (STUDY I) LAMINAR FLOW TURBULENT FLOW K 9 (LB) Kq (LB) Wing 24,812 (54,700) 17,620 (38,846) Horizontal Tail 1,329 (2,929) 1,715 (3,781) Vertical Tail 1,763 (3,886) 1,353 (2,983) Fuselage 17,305 (38,152) 17,155 (37,820) Landing Gear 6,972 (15,371) 7,173 (15,814) Nacelle 2,432 (5,361) 2,826 (6,231) Propulsion 8,889 (19,597) 9,375 (20,668) Fuel System 1,082 (2,386) (I ,857) Flight Controls 2,530 (5,577) 2,151 (4,741) Auxiliary Power Unit 470 (1,037) (1,037) Inst rument s 794 (1,751) (1,815) Hydraulics 1,047 (2,309) (1,963) Pneumatics 845 (1,862) (1,864) Electrical 1,574 (3,470) 1,574 (3,470) Avionics 1,427 (3,146) 1,427 (3,146) Furnishings 16,521 (36,422) 16,521 (36,422) Air-Conditioning 1,217 (2,684) 1,225 (2,700) Ice Protection 223 156 (492) (344) 28 28 Handling Gear (62) (62) 84.170 Manufacturers Empty Weight 91,260 (201,194) (185.564) Operator's Items 5_596 (12,336 (1_,336 Operator's Empty Weight 96,866 (213,530) 89,766 197,900 Payload 31,298 (69,000) 31,298 69,000 Fuel 54,190 66,179 (119_470

(145,9oo)

Takeoff Gross Weight 182,344

(4o2,ooo) 187,243

(412_800)

Table 5-10b

WING WEIGHT COMPARISON

(STUDY I)

LAMINAR FLOW TURBULENT FLOW Kg (LB) Kg (LB) Wing Box 13,820 (30,467) 9,980 (22,002) 2,210 (4,873) Wing Box Flutter Penalty Leading Edge 690 (1,521) 564 (1,243) Trailing Edge 755 (1,665) 615 (1,356) Wing to Fuselage Fairing 572 (1,261) 457 (1,007)

Wing Tips 14 (30) 11 (25)

Ailerons

313 (689) 264 (582)

Flaps 3,137 (6,917) 2,462 (5,427) Slats 717 (1,580)

Spoi I ers 341 (751

-f'/--,-,-,-,-_ (38,8461

( ,553)

Basic Wing Weight Laminar Flow System: Leading Edge Penalty 491 (1,083) Laminar Flow Panels 2,633 (5,805) Engine/Pump Assembly 746 (1,645) I , 851 (1,876) Ducting - Wing to Pump

Fairings 335 (738)

Laminar Flow Penalty -"_056 (11_147) Total Wing Weight 24,812 (54,700)

17,621 (35,846)

5O Table 5-11 DESIGN DATA AND GEOMETRY SUMMARY -!

LAMINAR FLOW TURBULENT FLOW WING GEOMETRY Area - Theoretical m2 (ft2) 381 (4,100) 267 (2,870) 12 12 Aspect Ratio 0.3 0.3 Taper Ratio 30 30 Sweep of c/4 deg 0. II 0.137 t/c Mean Flap Area m2 (ft2) 68 (729) 53 (572) VCK Area m2 (ftz) 39 (422) Aileron m2 (ftL) 15 (163) 13 (137) Spoiler Area m2 (ft _) 36 (384) 27 (287) TAIL GEOMETRY Horizontal Tail Area m2 (ft 2) 64 (693) 89 (960) Elevator Area m2 (ft z) 16 (168) 23 (248) Horizontal Tail Length m (in) 27.43 (1,080) 24.38 (960) Horizontal Tail Volume 0.75 1.52 Vertical Tail Area m2 (ft 2) 63 (679) 46 (495) Rudder Area m2 (ft _) 22 (237) 18 (191) Vertical Tail Length m (in) 20.83 (820) 25.4 (I,000) Vertical Tail Volume 0.051 1.23 FUSELAGE GEOMETRY Length m (in) 55.65 (2,191) 55.65 (2,191) 6.02 (237) 6.02 (237) Maximum Diameter m (in_ Wetted Area - Gross m2 (ft_) 907 (9,767) 907 (9,767) Main Cabin Floor Area m2 (ft L) 234 (2,518) 234 (2,518) ENGINES Turbofan Turbofan Type SLST/Engine kg (Ib) 11,054 (24,370) (11,657) (25,700)

Table 5-12

PERFORMANCE CHARACTERISTICS

(STUDY I)

LAMINAR FLOW TURBULENT

WEIGHT

kg (LB)

Takeoff

182,344 (402,000) 187,243 (412,800)

Landing

137,438 (303,000) 129,637 (285,800)

OEW

96,842 (213,500) 85,811 (198,000)

MEW

91,263 (201,200) 84,141 (185,500)

Mission Fuel

54,204 (119,500) 66,179 (145,900)

Payload

31,298 (69,000) 31,298 (69,000) CRUISE Range km (N HI) 9,620 (5,000) (9,620) (5,000) Altitude m 10,670 - 11,890 9,450 - 11,890 Step/Climb (FT) (35,000 - 39,000) (31,000 - 39,000) 0.40 - 0.49 CL 0.50 - 0.60 L/D 24.0 - 25.8 16.7 - 18.2 G/hr/N SFC (AVG) 60 (0.59) 59 (0.58) (LB/LB/HR) 1.16 RangeLFC On/RangeLFC Off Fuel Required kg (LB) 1,190 (2,620) For Suction f. - "<._.

Takeoff Distance m (FT)

2,800 (9,180) 3,050 (10,000)

V2 m/s (KN) 70.5 (137) 74.1 (144) 1.55 2.05 CLT.O.

Flap Angle Degree 5110 Landing Distance m (FT) 1,539 (5,050) 1,265 (4,150) VApproach m/s (KN) 66.9 (130) 58.1 (113) 1.29 2.20 _LA_p_ h Degree 22.5 35/15 Engine thrust is sufficient to allow an initial cruise altitude of 9,450 m (31,000 Ft) for both aircraft. This can be achieved without using LFC on the laminar flow aircraft; with LFC-on the initial cruise altitude is 10,670 m (35,000 Ft).

In accordance with commercial operational rules, a step cruise/climb mission profile and international fuel reserves are assumed. When using this commercial type flight path, the aircraft cruises at a constant altitude until enough fuel has been used to permit cruise at an altitude 1219 M (4,000 Ft) higher; the increases in cruise altitude thus are made in steps of 1219 M (4,000 Ft). Figure 5-16 presents the altitude and operating CL profiles over the mission ranges of the two base aircraft. Climb profiles for the two aircraft are summarized in Figure 5-17.

ORIGINAL PAGE 18 OF POOR QUALITY 3O f_ 2O M =0.8 FUEL FLOW AS_ SUCTION AIRFLOW g/HG/g/s BASED ON q NORMAL 1. CPN UPPER -0.86 t CPN LOWER - 0.10 (LB/HR/LB/SEC) 2. WING Ac/4 = 30 ° 3. PCOMPRESSOR FACE = 0.80 P SURFACE 4. QUPPER/QLOWER = 0.353/0.642 5. EXHAUST VELOCITY = FREE-STREAM VELOCITY 6, BSFC = 0.268 g/HR/W (0.44 LB/HR/HP) , , 111 7 12 13 8 9 10 I L _ I (25) (3O) (35) (4O) 81 _EN 223137 ALTITUDEkm (1000 FT) FIGURE 5-15. PRELIMINARY ESTIMATE OF FUEL FLOW FOR SUCTION SYSTEM PUMPING M CRUISE= (18 RANGE= 9,620 km (5,(XX) N MI) PAYLOAD= 31.t00 kl (69,000 LB) RANGE (1,000 N MI) (1) (2) (m) (4) (S) (o) 14 (46) 13 (42) AL13TUDE ALTITUDE (38) (1000 FT) (i=m) n LAMINARFLOWCONTROL I 9 (3O) Ct. 03 "_ --_-_ I I , I , l , I , 1 , lO 0 2 4 6 II RANGE(1000 m) FIGURE 5-16. AIRCRAFT ALTITUDE PROFILES The "optimization" of each aircraft configuration, as to wing area and thrust selection, is accomplished by a computer mission program (K5JA) which sizes as well as computes the performance. Typical carpet plots are shown in Figures 5-18 through 5-21. (The output is in English units and conversion to metric units in the carpet plots is impractical). See also Section 7.3.2.

Performance limitations, such as takeoff distance, minimum approach speed and thrust may be imposed on the sizing matrix. The configuration selection is made on the basis of minimum takeoff gross weight; costing and pricing of the laminar flow aircraft were not sufficiently developed at this stage to warrant using Direct Operating Costs (D.O.C.).

Results of a preliminary aspect ratio "optimization" study are presented in Figure 5-22. Aspect Ratio 12 is selected for both base cases. See also Sections 7.3.1 and 9.1.6.

5.3.8 Economics - Initial Baseline Aircraft The results of a preliminary comparison of economic aspects of the two initial baseline aircraft is indicated in Table 5-13, This very early assessment of the probable economic trends of the two configurations, LFC and turbulent, does show the LFC to be at least competitive. A comprehensive economic analysis of the final configuration is in Section 6.0.

Table 5-13 PRELIMINARY COSTING- INITIAL BASELINE AIRCRAFT (1976) (DOLLARS IN MILLIONS) LAM INAR TURBULENT AIRPLANE COST 49.4 45.6 - With Engine 43.1 38.5 - Without Engines DIRECT OPERATING COST $ per aircraft-km (NMI) 3.41 (6.31) 3.44 (6.38) $ per seat-km (NMI) .0114 (.0211) .0115 (.0213) OF poOR QLI/kLi3"Y, RANGE=9,620 km (5,000 N MI) PAYLOAD=31,300 kg (69,000 LB) f..

LAMINAR FLOWCONTROL TOGW =182,344kg(402,000LB)2 AR=12 Sw=380.9m2(4100FT ) TURBULENT FLOW TOGW = 187243 kg(412,800LB) =12 Sw = 266.6 m 2 (2870 FT 2) 14- (40) i / (30) / 13- / ,SAME FOR LAMINAR AND / TURBULENT FLOW AIRCRAFT 12-

i (20)

I

i

I 11- I I (to) I tO- I I ,l[

/

I 1 I

/

[

_ 0.6 0.4 0.8 0 0.2 0.6 0.8 1.0 02 0.4 0 MACH NUMBER .............

CL FIGURE'5-17. CLIMB PARAMETERS RANGE = 5000 N MI PAYLOAD = 69,000LB (300 PSGR) ASPECT RATIO =10 ASPECT RATIO =12 ASPECT RATIO =14 \ \ / "_ _ API_OACH SPEED LENGTH " 10,000 FT 0 39O FIGURE 5-18. MISSION SIZING - LAMINAR FLOW CONTROL BASELINE ___ RANGE= 5000 N MI PAYLOAD= 69,000 LB (300 PSGR) i ASPECT RATIO = 10 A OF PO'_ Q_.LI'I'Y i,,,4 ASPECT RATIO = 12 v ASPECT RATIO = 14 ,,, 105 Z ._1 uJ 100 M.

95_ i -- 7.GEN-22934A FIGURE 5-19. MISSION SIZING-LAMINAR FLOW CONTROL BASELINE - FUEL BURNED ,,%_ • ASPECT RATIO ASPECT RATIO ASPECT RATIO A = 10 = 12 = 14 m 4801 ,.,.J T.O. THRUST v 460 I-.

'I"

/

I.IJ 3: 44O o_ U3 O p,,.

U.

EL O L,J _TAKEOFF RUNWAY b¢ ,< 400 LENGTH = 10,000 FT I- ?'GIrN 22|72 FIGURE 5-20. MISSION SIZING - TURBULENT BASELINE - TAKEOFF GROSS WEIGHT OF PO0? ;_,'!;': ASPECTRATIO ASPECTRATIO ASPECTRATIO -- t0 = 12 _- 14 FUEL 140 BURNED ( 1000 LB) 7-GEN 22917 FIGURE 5-21. MISSION SIZING - TURBULENT BASELINE - FUEL BURNED , LAMINAR FLOW CONTROL AIRCRAFT ............... 3,050 m (10,000 FT) TAKEOFF DISTANCE -(13S) ..... TURBULENT FLOW AIRCRAFT -(45) 60- 400.

(125) (4O) 55- FUELBURNED, (ITS) 3S0 tOO0 WINGAREA, (1000 LB) 5O (35) m 2 (105) (tO0 FT =) 300- (30) 45.

2_0- , (25) I (iS) ENGINE THRUST, TAKEOFF I_K) -_ ._(420) KN 110- "'°°"-°-.OO°o., ..... _.= r (24) GROSS WEIGHT, I-"" "-- (1000 LB) i ....... (41o) (lOOO LB) °°"oo..o..__|j: 100" , , : (22) , J(400) 182-[ I 10 11 12 13 14 13 14 10 11 WING kSPECT RATIO 7.GEN-22676-I WING ECT RATIO EFFECT OF ASPECT RATIO ON AIRCRAFT SIZING FIGURE 5-22.

The costing comparison shown in Table 5-13 is based on a 400 aircraft buy depreciated over 14 years. The turbulent aircraft cost estimate appears i_ _.

reasonable using the DC-IO prices as a comparion.

The cost of the laminar aircraft is about 8 percent higher than the turbulent. Breakdown of the cost differentials into the labor and material components shows that the major increases in the laminar flow airplane costs are due to engineering (1.2 percent), manufacturing (2.0 percent) and outside vendor costs such as materials, tooling, special engine installation parts (2.8 percent).

The DOC for the laminar flow aircraft is approximately 1 percent less but at this early stage of the economic analysis, the DOC's are essentially breaking even. It is emphasized that this preliminary costing should be considered only on a comparative basis - not on absolute values.

5.3.9 Study Conclusions - Initial Baseline Aircraft The base case laminar flow aircraft study and the comparison with an equally advanced technology turbulent aircraft, emphasized the following areas needing special attention: o Approach CL - to reduce wing area o Effective Wing Thickness-to reduce wing weight o Optimized Structure - to reduce wing weight o Minimum Weight LFC Panel - to reduce wing weight o Suction System Simplification - to reduce cost and weight o Maintenance - to reduce operating costs The most important issue is a reduction of area for the laminar flow wing and consequent reductions in the overall airplane size and weight; the low approach CL obtainable without a leading edge high lift device, is the critical sizing criterion. Further possible weight reductions are associated with the effective wing thickness, and minimum weight LFC panel design.

Simplicity and efficient integration of the suction system into the aircraft is essential. Maintenance procedures are particularly important in making the LFC aircraft readily acceptable from an operational standpoint.

Particular attention was therefore directed to the following details for the

next improved baseline configurations:

o High Lift System

o Wing Planform Variables

o Extent of Laminarization

o The Number and Location of Propulsion Engines

o Structural Design Concept

o LFC Surface Material

o Structural Concept/Suction System Manifolding Integration

o Suction System Manifolding and the Numberand Location of

Suction Engines

o Integration of Leading Edge Protection

Trade studies on the number of propulsion and suction engines are summarized

in the following pages. Other items are covered under the sections relating

to the various disciplines involved.

6O

5.4 STUDY II FOUR VS THREE PROPULSION ENGINES ,f , The aircraft configurations for this study are shown in Figures 5-23a and b.

The results of the trade study to determine the relative advantages of a four or a three engine main propulsion system arrangement are presented in Tables 5-14a, b and c. Sizing constraints considered in this analysis were: (I) initial cruise altitude, (2) approach speed, and (3) takeoff field length. In both cases the approach speed of 66.9 m/sec (130 KEAS) was the limiting constraint which established the aircraft size. Cost data were based upon 1976 dollars and a 45 cents per gallon fuel price. Subsequent increases in fuel prices, over and above general inflation would increase the economic difference shown between the two configurations without altering the conclusion reached. The differentials between the four engine and the three engine arrrangements favor the three engine arrangement. The three-engine configuration was estimated to be 2.54 percent lower in acquisition cost and 3.62 percent lower in direct operating cost than the four-engine configuration. Thus the three-engine configuration with engines located aft on the fuselage was selected for subsequent LFC aircraft studies.

Further detail on this trade study with respect to the influence of engine noise on engine location may be found in Section 7.2.

o OF PO0?:: i __'::..i _ -= / %, WING HORIZ VERT

i/

ASPECT RATIO 12 5 1 TAPER RATIO 0.3 0.4 0.8 SWEEP 300 30 o 40 o 67.61 m !.., (221.81 _) i

i

lln in nil ,_ (_2 _) 7-GEN-22726 • FIGURE 5-23a. STUDY II - FOUR-ENGINE CONFIGURATION m m ASPECT RATIO 12 5 TAPER RATIO 03 0.4 0.8

4O'

SWEEP 30 ° 300 1221.81 FT) 17.04 m i" .,,.

(55.92 FT) i, 61.6 m (202 FT)' ,-_,.,_,,,-, FIGURE 5'23b_ STIJD-Y H-- THREE-ENGINE cOnFIGuR,_TION Table 5-14a FOUR VERSUS THREE PROPULSION ENGINES DRAG & WEIGHT DIFFERENCES D/q - 3 ENGINES m2 SQ. FT.

Fuselage 0 0 Wing 0 0 Horizontal Tail 0 0 Vertical Tail -0.0576 (-0.62) Propulsion +0.0102 (+0. II) Miscellaneous -0.0037 (-0.04_ TOTAL -0.0511 (-0.55) TOTAL D/q - 3 ENGINES LFC ON 4.312 (46.41) LFC OFF 5.785 (62.27) 6 OEW - 3 ENGINES -886.3kg (-1513 LB) Table 5-14b STUDY II CHARACTERISTICS OF SIZED AIRCRAFT RANGE : 9,260 km (5000 NMI), PAYLOAD = 31,300 kg (69,000 LB), 300 PASSENGERS 4 - ENGINES 3 - ENGINES TAKEOFF GROSS WEIGHT 195,349 (430,670) 193,191 (425,913) km_ (LB) WING AREA (SQ FT) 412.5 (4,440) 412.5 (4,440) THRUST/ENGINE N (LB) 112,540 (25,300) 148,593 (33,405) FUEL BURNED kg (LB) 46,557 (102,641) 45,900 (101,193) TOTAL FUEL kg (LB) 56,273 (124,061) 55,511 (122,381) OPERATOR'S EMPTY WEIGHT kg (LB) 107,778 (237,609) 106,382 (234,532) LANDING WEIGHT 148,791 (328,028) kg (LB) 147,295 (324,729) INITIAL CRUISE ALTITUDE m (FT) 10,670 (35,000) 10,670 (35,000)

Table 5-14c

DIRECT

OPERATING COST SUMMAR Y

STUDY II CONSTANT 1976 DOLLARS ITEM FOUR-ENGINE THREE-ENGINE CONFIGURATION CONFIGURATION AIRFRAME COST $ x 106 47.369 46.612 TOTAL ENGINE COST $ x 6.620 6.008 AIRPLANE COST $ x 106 53.989 52.620 LANDING FEE $/FLT CYCLE 732. 14 724.05 COCKPIT CREW $/FLT CYCLE 4,434.

4,420.39 CABIN CREW $/FLT CYCLE 3,491.

3,491.17 DEPRECIATION/INSURANCE $/FLT-CYCLE 04 11,096.

10,814.67 MAINTENANCE $/FLT-CYCLE 47 6,432.

5,653.45 FUEL $/FLT CYCLE 80 6_893. 6,796.54 TOTAL $/FLT CYCLE 33,079.

31,900.27 DOLLARS/km ($/NMI) 3.57 (6.62) 3.44 (6.38) CENTS/SEAT-km (¢/ASNMI) 1.19 (2.21) 1.15 (2.13) 5.5 STUDY Ill FOUR VS TWO WING-MOUNTED SUCTION ENGINES The results of this trade study discussed in Section 10.6, indicate that two wing-mounted suction engines per airplane can provide adequate suction for the 300 passenger aircraft configuration.

The initial LFC configuration distributed suction along the wing using four suction engines. However, closer examination shows that the integral duct sizing is adequate to permit the required suction flow to be handled by two suction engines, placed at the wing break on each side.

A comparison of two versus four suction engine characteristics is presented in Table 5-15 and the advantages of two suction engines only is summarized as fol 1 ows : o Simpler system o Lower weight and drag o More efficient units Lower initial cost o o Lower operating cost The decision was therefore made to use two suction engines only on further LFC configurations.

Further discussion of this trade study may be found in Section 7.2.

Table 5-15 COMPARISON OF TURBOSHAFT ENGINE ARRANGEMENTS FOR SUCTION SYSTEM 2-ENGINES 4-ENGINES CHARACTERISTIC/ENG I NE Rated (SLS) Power kW (SHP) 895 (1200) 447 (600) Airflow kg/s (LB/SEC) 2.90 (6.4) 1.59 (3.5) Pressure Ratio 12-14 8-10 Weight kg (LB) 115.7 (255) 95.3 (210) Installed BSFC kg/h/N (LB/HR/SHP) at 12,190m (40,000 FT) 0.040 (0,39) 0.043 (0.42) 5.6 STUDY IV FINAL AIRCRAFT CONFIGURATIONS In this study, two improved LFC configurations having different extents of laminar flow are considered. They are then compared with an updated turbulent configuration and the final LFC configuration is selected. The configuration ground rules are given in Table 5-16. Changes relative to the initial LFC configuration, that result from previous trade studies and analysis include: Three propulsion engines instead of four.

Two wing mounted suction engines instead of four.

Updated wing structural design.

Wing leading edge protection system added.

High lift systems improved.

The resulting LFC configurations are illustrated in Figures 5-24a and b.

Figure 5-24a shows the improved LFC configuration with LFC on both wing surfaces to 70 percent chord as before and Figure 5-24b shows the competitive LFC configuration with LFC extending to 85% chord on the upper wing surface only. The updated turbulent configuration is shown in Figure 5-24c.

Detail consideration leading to the selection of these final configurations are discussed in the following paragraphs.

Table 5-16

FINAL CONFIGURATIONS - GROUND RULES

PAYLOAD

299 Passengers at 75 kg (165 LB)/PAX 31,300 kg (59,000 LB) Design Mission 42,980 kg (94,755 LB) MAX 7 Pallets forward, 12 LD3 Containers Aft Cargo Density 160 kg/m 3 (I0 LB/FT 3)

GALLEY SERVICE International standard

1st Class-13 kg (28.5 LB)/PAX Tourist-6kg (13.5 LB)/PAX

PASSENGER SERVICE

1.8 kg (4 LB)/PAX C. G. TRAVEL -2 to 32 Percent MAC - LFC upper surface 0 to 32 Percent MAC - LFC both surfaces 14.5 to 39.5 Percent MAC - Turbulent Items not covered are the same as for the initial baseline, Tables 5-1 and 5-2.

OF POQ_ q'.!._LtTY WING HORIZ VERT / AREA m2(F"r 2) 331 (3560) ¢_,B 11064) 62.0 (667) ASPECT RATIO 10 5 1.1 TAPER RATIO 0.25 0.4 0.7 SWEEP 30 DEG 30 DEG 40 DEG THICKNESS RATIO 0.117 AVG 0.11 0.11

Yq

TAIL VOLUME -- 1.23 0.0646 2Z2m -- I \ (72.9

Fr)

63.0 m ....

• (206.6 FT)_ .I

57.5 m

"r "='"' J

r__J_' 16.7m --_ ---_.-=_. _3 (S4.9 FT) 9 GEN 262lt8 I LFC AIRCRAFT- UPPER AND LOWER WING SURFACE LAMINARIZED TO FIGURE 5-24a.

70-PERCENT CHORD if: WING HORIZ VERT AREA-m2 (FT2) 288 (3100) 80.3 (864) 52.7 (567) ASPECT RATIO 10 5 1.1 TAPER RATIO 0.25 0.4 0.7 SWEEP 30 DEG 30 DEG 40 DEG THICKNESS RATIO 0.108 AVG 0.11 0.11 TAIL VOLUME COEF -- 1.24 0.068 1204.3 FT) 53.7 m (176.1 FT) 62.3 m ____,"__ t_GEN_7 - | FIGURE 5-24b. LFC AIRCRAFT- WING UPPER SURFACE LAMINARIZED TO 85-PERCENT \ 4 ¸ CHORD ORIG'_.,_L i" OF POOR _" _'"' r.

r WING HORIZ VERT // / _/ AREA-- m 2 (FT 2) 260 (2800) 89.6 (964) 47.2 (508) / _/ ASPECT RATIO 0.25 0.35 0.35 TAPER RATIO 10.85 3.95 1.60 _ I__ 1 m SWEEP 40DEG _ \., L_ ......

30DEG 35DEG toz ru) THICKNESS RATIO

o,=,,vo ,o i'

TAIL VOLUME - _.38 0.0,9 __, _ J, 53.1 m (1743 IT)

• -F

__ 16.55 m ....... _ ...................... 5- (_.3IT) _EN._21n FIGURE 5-24c. ADVANCED TURBULENT AIRCRAFT SWEEP 30 DEG C L 0.5 Moo 0.8 0.006 -'*QOQOQ, CONVENTIONAL I"*QQQR'bOIOmQQI.Q.Q==a..=..= I TURBULENT WING b'l'lm'ie''''°''eee=='''°=''=-- t-.ia.M..= 0.004 PROFILE SUCTION TO 85 PERCENT DRAG CHORD ON UPPER SURFACE ONLY COEFFICIENT Imm "m= =mm ,,.m, ,m= =m,= _ mm, (Cd) 0.002 SUCTION TO 70 PERCENT CHORD ON BOTH SURFACES 8O 2O 4O 60 CHORD REYNOLDS NUMBER x 10 .6 IFGEN-_ FIGURE 5-25. EFFECT OF LFC EXTENT ON PROFILE DRAG 5.6.1 LFC Aircraft 5.6.1.1 Aerodynamic Considerations- a. Extend of Lamination This significant trade study was prompted by the fact that the drag reduction achieved with laminar flow to 85 percent chord on the upper surface approaches that obtainable with laminar flow to 70 percent chord on both surfaces. This is shown in Figure 5-25. The study shows that the advantages gained by utilizing LFC on the upper wing surface only, more than compensate for the increase in drag coefficient. These advantages include: 1. Simplification of the LFC system.

2. LFC system weight is reduced.

3. Initial cost is less.

4. Vulnerability of the lower LFC surface to damage from foreign objects thrown up from the runway (FOD) is avoided.

5. The possibility of fuel leakage into the LFC panels and integral ducts is avoided.

Conventional access panels to wing leading and trailing edge systems and fuel tanks can be provided for inspection and maintenance without disturbing any LFC surface.

7. Maintenance costs are reduced.

8. A shield for contamination avoidance can be deployed forward of the wing leading edge and can be retracted into the lower surface when not required.

o The shield can be designed geometrically to function as a high lift device. Wing area can then be reduced and wing loadings become competitive with those of advanced turbulent aircraft.

10. The use of a retractable high lift device allows the safe use of a sharper leading edge on the basic wing. This results in a reduction or possible , LL elimination of suction requirements along the attachment line.

Both LFC configurations were investigated in depth and the results are summarized in the following paragraphs under headings of the technical disciplines involved.

b , Win 9 Planform - The wing planform characteristics are: Sweep angle 30 o Taper ratio .25 Aspect ratio i0 The wing sweepback of angle of 30 o was selected on the basis of wing thickness permissible for an operational CL of .5 to .6 and the practical aspects of associated internal space available for landing gear stowage and suction manifolding installation. With an alternative sweep angle considered of 25 o and a correspondingly thinner airfoil section, gear and duct space limitations were found. The selected wing planform, sweep and taper ratio of .25 allows the use of analytical and wind tunnel test results performed during the Douglas EET-related design work on similar wings.

Aspect ratio of 10 was selected for both the upper surface only and the upper-plus-lower laminarization cases. An extensive aspect ratio study for LFC on both surfaces with wing strength, stiffness, flutter and aeroelastic effects taken into account, is discussed in Sections 7.3 and 9.1.8. Aspect ratios of I0, 12 and 14 were considered. The variation of wing weight with aspect ratio is not definitive between aspect ratios I0 and 12; the curve is essentially flat. Consequently, aspect ratio i0, which gives increased wing depth and shorter span for the ducting was selected for both LFC configurations.

C° Airfoil Section - As stated previously in Section 5.3.2.(b), the supercritical type airfoils selected for the laminar flow aircraft are designed to be shock-free at the design CL, also no significant amount of separation should occur even with turbulent flow on the laminarized portion of the airfoil. More than sufficient buffet margin exists with these shock-free airfoils. Continued "optimization" during the LFC study resulted in the selection of the airfoil thickness as a function of the operating CL, as shown in Figure 5-26. Wing thickness variation was considered in the aircraft sizing program. The selected airfoil thicknesses are as follows: Upper-Surface LFC Onl_, Upper-Plus-Lower-Surface LFC Cruise CL 0.56 0.50 Airfoil Section Shock-free Shock-free Root t/c 12.76 13.82 (DSMA 679 Type) Aero Break t/c 10.08 10.92 (DSMA 691 Type) (40% Span ) Tip t/c (DSMA 691 Type) 10.08 10.92 Avg t/c 10.3 11.7 d. High Lift System - Figure 5-27 illustrates the wing cross-sections for the two extents of laminarization considered. The figure shows that the extent of laminarization is a controlling factor on the ducting/structure integration and on the trailing edge flap installation. The finite thickness at the airfoil trailing edge is required to provide adequate depth for the flap structure. This increased depth is provided by rotating the lower airfoil surface contour about the 60 percent chord point without significantly affecting the airfoil characteristics. The resulting depth of the airfoil at the trailing edge is less than one percent of the chord; consequently any drag contribution is negligible.

©

,'C;,.+: .... -. :++ ;:++++ -, ++ OF POOR QUALITY Moo = 0.8 CHORD REYNOLDS NUMBER=36 x 106 0.7 I I CL = C _ .COS2A 1.17 0.6 AIRPLANE LIFT 0.5 COEFFICIENT (CL) 0.4 O3 8 9 10 11 12 13 14 15 AVERAGE WING THICKNESS -- CHORD RATIO (PERCENT) II-GEN-Z2622 FIGURE 5-26, VARIATION OF DESIGN LIFT COEFFICIENT WITH AIRFOIL THICKNESS LAMINARIZATION TO 70 PERCENT CHORD _" UPPER AND LOWER SURFACE ....

C L MAX-- 2.5 (FLAPS 35°/15 °) ] _+_ , _ _ _ f _ _ r _, ! _ ' LAMINARIZATION TO 85 PERCENT CHORD +" UPPER SURFACE ONLY C L MAX = 3.1 (FLAP 30 °) ,.,,0,,,.

+j FIGURE 5-27. COMPARISON OF LFC WING SECTIONS A highly efficient and powerful high lift system is critical to the design of a competitive LFC aircraft. Takoff distance or approach speed are usually the wing sizing criteria; both of these depend on the aircraft high lift capability. Therefore, in each laminarization case, an effort was made to install the most efficient high lift system consistent with space available and compatibility with the laminar flow installation. Note that in either case, laminar flow is not carried across the flap hinge.

For LFC on both surfaces, Figure 5-27 shows the selected 25 percent chord trailing edge double-slotted flap, with a maximum deflection of 35°/150 . No leading edge device is used since such an arrangement, with its retraction requirements, is not compatible with achieving laminar flow on the lower surfaces.

With LFC on the upper surface only, and suction over a larger percent chord, the trailing edge flap is reduced to 15 percent chord with a single slot, and the maximum deflection is 300 . However, with the lower wing surface not laminarized, a 10 percent chord leading edge shield is usable as both a high lift device and as protection of the leading edge from insect impingement, and can be retracted into the lower surface without affecting the extent of laminar flow.

With either extent of laminarization, the flap mechanism is designed to provide considerable chord extension before angular deflection take place, namely: 15 percent chord extension with the 25 percent chord douhle-slotted flap o 7 percent chord extension with the 15 percent chord single-slotted flap

©

?4 The low speed high lift characteristics of the two flap systems are shown in Figures 5-28 through 5-31.

e. Horizontal Tail Sizing - The horizontal tail sizing scissor plots for the two laminar configurations are presented in Figures 5-32 and 5-33.

Several points of interest may be noted from these plots: The critical sizing criteria for the forward c.g. limits are different for the two aircraft. For the case of laminarization of the upper-surface-only, the nose wheel liftoff is the critical sizing factor. In the case of upper-plus-lower surface laminarization, the critical condition is the trim to 1.4 Vstal 1 in landing approach with ice on the horizontal tail and flaps deflected 35°/15 o .

The landing gear position at 49 percent MAC, in conjunction with the airplane static ground angle of 12 o , establishes the aft c.g.

limit for both aircraft as 30 percent MAC, due to a tip-over limitation.

This tip-over limit precludes effective use of negative stability or even stability limits in establishing the aft c.g. limits. The landing gear is positioned at the most aft location compatible with submerging the retracted gear strut into the airfoil contour. The combination of a supercritical airfoil, a thinner airfoil for the laminar flow requirement, and a wing planform with a minimum "yehudi" (trailing edge extension) at the wing root, imposes this restriction on landing gear placement.

ORIGINAL PAC_ [$ OF POOR QL!ALLTY SINGLE-SLOTTED T.E. FLAP, CF/C-0.15 VCK L.E FLAP, CF/C=0.10 AR=IO 3.0 C L CLAT 13 V_,_ 1.5

Sf

5 10 15 20 25 30 (DEG) ,._E, 22,,_ FIGURE 5-28. LOW-SPEED LIFT CURVES - UPPER SURFACE SUCTION ONLY ,q-*- SINGLE-SLOTTED T.E. FLAP, C F/C =0.15 VCK LG FLAP C F/C = 0.10 AR = 10 L/D 12 0 0.4 0.8 1.2 1.6 2.0 2.4 CL ,.=,_.._ FIGURE 5-29.

LOW-SPEED L/D CHARACTERISTICS -- UPPER SURFACE SUCTION ONLY TWO-ELEMENT T.E. FLAP CF/C =0.25(STOWED) AR=10

ORIGINAL PAG,_ t'_

OF POOR QUALITY 35/15 6 r (DEG) -- I 2.,5/10 15/10 --_" 5/10

._"-_ o

CL 1.0 _--CLAT 1.3V MIN

J

O.5

J

0 5 10 15 20 25 oz (DEG) 8-GEN.22550A FIGURE 5-30. LOW-SPEED LIFT CURVES - BASE CASE - SUCTION ON BOTH SURFACES TWO-ELEMENT T.E. FLAP CF/C=0.25 (STOWED) AR=10 2O

f

_0 _5/10 _F (DEG)

f

15/10 - f

f

L/D 10 25/10 35/15 0.4 0.8 12. 1.6 2.0 2.4 CL .-GEN-22574 h" :-'i FIGURE 5-31. LOW-SPEED L/D CHARACTERISTICS - SUCTION ON BOTH SURFACES ,,,_...;_,::;.-,= PAGE IS OF POOR QUALITY 1.6 I TIPOVER LIMIT /

r

/ LANDING APPROACH 1.4 WITH ICE ON HORIZ TAIL / 5f = 30 °, TRIM 1.4 V S 1.2 FT LIMIT WITH 2% NEGATIVE STATIC MARGIN 1.0 V H 0.8 0.6 I I I MAIN GEAR 0.4 APPROACH 6f = 30 ° , TRIM 1.3 V S O.2 --10 0 10 20 30 40 50 60 CG (PERCENT MAC w) =,E.,,_ FIGURE 5-32. HORIZONTAL TAIL SIZING - UPPER-SURFACE-ONLY WING LAMINARIZATION I TIPOVER LIMIT / I i j _ AFT LIMIT WITH 2% NEGATIVE I STATIC MARGI N I I % % APPROACH WITH % ICE ON HORIZ TAIL 1,0 %% "f = 35/15, TRIM 1,4 V S ; APPROACH VH 0.8 _f " 35/15, TRIM 1.3 V S I i MAIN GEAR I --10 0 10 20 30 40 50 60 B1 _GE N-22301B CG (PERCENT MACw) FIGURE 5-33.

HORIZONTAL TAIL SIZING - UPPER-PLUS-LOWER-SURFACE WING LAMINARIZATION o The specific tail volumes were selected to provide 33.5 percent and 31.5 percent c.g. travel, respectively, for the upper-surface-only and the upper-plus lower surface laminarization cases; see Figures 5-34 and 5-35. These c.g. travels are consistent with the aircraft loading and balance requirement and are compatible with DC-9 and DC-IO characteristics, taking into account the cabin length and differences in wing MAC.

In the case of LFC on the upper-surface-only, the flap deflection was subsequently found to be restricted to 12.4 o due to second segment climb limitation. This flap restriction would allow the forward c.g.

limit to be extended from -3.5 percent to -7 percent increasing the overall travel to 37 percent. Alternatively the tail volume coefficient could be reduced from 1.24 to 1.15 to maintain the 33.5 percent travel with the same c.g. limits.

f, Vertical Tail Sizin 9 - The vertical tail volumes (Vv) shown previously in Figures 5-24a and b, provide a level of directional stability equivalent to that of the DC-9-30 and the DC-I0-30 airplanes. The vertical tail volumes of .068 and .0646 respectively for the upper-surface-only and the upper-plus-lower-surface laminarizations provide a low-speed C of .0028. V is not n_ mcg critical since the moment arm of the outboard engine thrust is relatively small.

Lateral Control - A brief analysis of the lateral control effectiveness go of the upper-surface-only configuration was made using the following simplified procedure.

The lateral control effectiveness of the LFC aircraft was estimated based on DC-IO-30 low-speed aileron and spoiler data. Adjustments were made using a series of ratios, to account for differences in the wing area affected by the control surface, in the rolling moment arm of the control surface, in the lift curve slope of the wing, and in the

OF : " ;....... :'

24 FIRST-CLASS AND 275 ECONOMY-CLASS PASSENGERS AT 75 kg (165 LB) EACH WEIGHT, 1000 kl (1000 LB) 24 FIRST-CLASS AND 275 ECONOMY-CLASS PASSENGERS AT 165 POUNDS EACH 75 kg (166 LB) EACH MTOW ,,o-,_, .... /',_ .... _,o .... ;,_ .... ;'oi,, ....

.,. III/

i 60 ,

i '= <=.o, r\\_ _\\\\\\\\\\\\\ \t;wt i_, tIit //1 / /I/11

,,o '<:!I , I!1///I

100 . i (-24) (-20) (-16) (-12) (-8) (-4) 0 (4) (8) I I I I I I I I I -30 -25 -20 -15 -10 -5 0 5 10 MOMENT,kI-cm x 10 = (LB-IN. x 10 _ ) "_"-_"' FIGURE 5-35. LFC AIRCRAFT BALANCE DIAGRAM - UPPER AND LOWER WING SURFACES LAMINARIZED TO 70-PERCENT CHORD effective section angle-of-attack change due to the control surface deflection. This resulted in the following equation, where (as') is a function of the height and chordwise location of the deflected spoiler trailing edge as indicated by Figure 5-36.

polte,r

( s'J,r t,,'\ Lr'r" Lr.c

This approach indicates a 22-percent increase in effectiveness (at maximum control deflection) of the LFC lateral control system over that of the DC-I0-30 in the flaps-up configuration, and 13 percent increase in the landing-flap configuration.

Table 5-17 lists the contributions of each of the major components of the lateral control systems of the two aircraft at maximum control deflection, as well as the damping in roll derivative (Clp) A The higher value of( Clp)A shown for the LFC is attributed to the relatively higher aspect ratio and lower sweep angle of the LFC wing.

A comparison of the maximum low-speed rolling performance of the two 0_;_'_7 - ......

OF pC_C,;.-_ Q'U.-%LITy aircraft is also shown.

Included are the wing tip helix angle Pb and the roll rate (p) at an arbitrary speed of 77 m/s (150 knots).

The rolling performance of the LFC aircraft is comparable with that of the DC-IO-30 and, by inference, the lateral control system proposed for the LFC aircraft should be adequate.

Table 5-17 LATERAL CONTROL CONTRIBUTIONS COMPARISON DC-I0-30 LFC AIRCRAFT Flaps Up Flaps 50 ° Flaps Up Flaps 30 ° Control Surface Inboard Spoiler (_ CR) 0.00428 0.01092 0.01885 0.03260 0.00896 0.00878 ....

Inboard Aileron (A CR) Outboard Spoiler (A CR) 0.01948 0.04973 0.02563 0.05022 Outboard Aileron (A CR) 0.01610 0.02195 0.01504 0.02050 TOTAL 0.04882 0.09138 0.05952 0.10332 Dampin_ in Roll ( )A Per Degree CLp -0.00790 -0.00920 Rolling Performance 0.108 0.202 0.113 0.196 Wing Tip Helix Angle (Pb/2V) (Radian) Roll Rate at 77m/s (150 Kn) 18.9 35.4 18.6 32.3 p.(Degree/Second)

5.6.1.2 Win 9 Structure

Brief descriptions of the structural characteristics of the two LFC wings are presented. Further detailed description and analysis may be found in Section 9.0.

The wing structures of the two LFC configurations are illustrated in Figures 5-37 and 5-38. Complete wing sections were shown on Figure 5-27. The LFC suction panels are supported by the external stiffness of the main wing box. This creates integral ducting which is used to transfer the suction airflow to the dry bay above the suction pump nacelle. Figure 5-39 shows that holes must be provided in the main box skin panels, as illustrated further in Figures 5-40 and 5-41, in order to transfer the airflow to the suction pump.

The holes required through primary structure are similar for both LFC configurations.

The main wing box is of graphite epoxy composite construction and the LFC panels are of fiberglass with a porous metal surface. Figure 5-42 shows the cross-section of a typical LFC panel that has a calendered woven stainless steel porous surface, manufactured under the trade name Dynapore. The layers are described as follows: Layer 1: 80 x 700 + 80 x 80 diffusion bonded Dynapore .305mm (.012 in) thick porous surface Layer 2: Perforated 'S' glass laminate 1.4 mm (.055 in) thick This layer provides support for the surface.

Layer 3: This was subsequently eliminated to avoid moisture entrapment within the perforations of Layer 2, thus avoiding a possible cause of laminar separation due to freezing, also reducing the time required to restore full porosity to the surface layers after exposure to moisture or contamination avoidance liquid.

_'[ v, B3 0.20

x$ I

m= 0.8 C 0.16 - 0.12 -

A_ S

(rad) 0.08 - 0.04 - 0 I I I I 0 0.04 0.08 0.12 0.16 0.20 h s m C 83-GEN-23276 FIGURE 5-36. SPOILER LIFT EFFECTIVENESS AT LOW SPEEDS fPOROUS PANEL I i_,. 'i' : L . L I_--_ i : :_ _ ,_ I I ! --_ ' --,' _',' ¢ ,, / I )_ ' \ I TYPICAL WING RIB \_,---FUEL BREAK AERO / AA DRAIN HOLES POROUS PANEL FRONT SPAR PLANE'--" I y..._A_j_-_ :_ • SPANWISE COLLECTION • CONVERGENT STRINGERS • SINGLE SUCTION ENGINE

ACCE= ___' I

• SUCTION TO 0.70 C • EXTERNAL STIFFENED -_" _-'_' " A l SUCTION PUMP SUPPORT STRUCTURE ,REAR _. SUCTION PUMP SPAR PLANE _I_ _ '.._, t- FIGURE 5-37. STRUCTURAL ARRANGEMENT - SUCTION THROUGH BOTH SURFACES O.B5 C #._-- _ U POROUS PANEL !

"_"_"'_ REAR SPAR PLANE (0.67 C) ORIGINAL " _'_ OF POOR QUALITY "--="= FIGURE 5-38. TYPICAL WING RIB - SUCTION THROUGH UPPER SURFACE ONLY TO 0.85C

_------ _:. -_ - ]. I_ _ I,_,::_ .'1 " _1 :_*L--=_

LAMINARIZATION TO 70 PERCENT CHORD UPPER AND LOWER SURFACE LAMINARIZATION TO 85 PERCENT CHORD UPPER SURFACE ONLY

C

FIGURE 5-39. SUCTION SYSTEM MANIFOLDING INTEGRATION OF PO_,', _.J .... ,. • VIEW LOOKING DOWN AT UPPER SURFACE (GLOVE PANEL NOT SHOWN) I_GIEN-22776 FIGURE 5-40. STRUCTURAL ARRANGEMENT- UPPER WING BOX PANEL AT _ OF SUCTION PUMP SYSTEM

C

C°C / C_,J LOOKING DOWN AT LOWER SURFACE DRY BAY BULKHEAD I.GE=N-Z2T_ FIGURE 5-41. STRUCTURAL ARRANGEMENT SHOWING LOWER SURFACE AND FUEL BULKHEAD AT C L OF SUCTION PUMP SYSTEM Layer 4: 'E' glass laminated web, .5mm (.02 in) thick. This is now sewn directly to Layer 2 so that Layer 2 becomes part of the base sandwich panel.

Layer 5: 'E' glass back face, thickness .5mm (.02 in) minimum, depends on test results and buckling characteristics.

Substitution of an electron beam (EB) perforated titanium surface for the Dynapore outer layer resulted in improved structural and damage resistance properties. The titanium sheet thickness is .63 mm (.025 in) and the perforations are .063 mm (.0025 in) diameter. After performing satisfactorily as an LFC surface during wind tunnel testing, EB perforated titanium was selected as the LFC panel surface for the final configuration.

The wing leading edge box is where the structural, aerodynamics, suction and environmental problems must all be integrated very compactly. Typical leading edge sections for both LFC configurations are shown in Figures 5-43 and 5-44. The fiberglass corrugations support the surface between nose ribs and provide integral ducting for the suction air. Integration of the environmental systems is covered in the following paragraphs.

5.6.1.3 Wing Leading Edge Protection This system is described fully in Section 11.0. For LFC on both wing surfaces, the liquid systems must provide complete protection against both icing and insect impingement. Integration of the LFC suction and liquid systems is difficult, particularly if suction is required along the attachment line. One possibility that needs to be investigated is using the same porous surface for both suction air and liquid systems. The purging system and valves required would increase complexity and porosity requirements at the surface would need to be matched for both systems.

The problem is eased considerably with LFC on the upper surface only. A shield is deployed to protect the leading edge region as shown in Figure 5-44. With a large enough shield, the liquid systems would be needed only for ice protection. If the shield is not large enough, there is a tendency for very I t STR EAMWlSE FLOW _ 80 x 700/80 x 80 DYNAPORE SURFACE

I

I I _"E" GLASS LAMINATE POROUS GLASS SHEET "E" GLASS

SEWN TO WEB NODES / t_ '_

LAM I NAT E _cE_743-___ _ FIGURE 5-42. LFC GLOVE PANEL LAYER IDENTIFICATION SUCTION C_K_TS INSECT \ PROTECTION \ \\\ FLUID .._..._..,._.__ INSECT AND ICE /

\

PROTECTION FLUID _ _ INSECT PROTECTION FLUID B-GE N.22389 FIGURE 5-43.

LEADING EDGE DESIGN - SUCTION ON BOTH SURFACES small insects to be deflected with the airflow between the shield and the wing and to impinge on the lower leading edge surface. Large insects might not be deflected by the airflow and could contact the upper surface with a glancing impact. Consequently a liquid system is shown for insect protection in addition to the shield, but flight testing may show that this is not necessary. An alternative method of applying protection liquid to the leading edge is to use the shield as a spray-bar as shown in Figure 5-45.

Retraction of the shield into the confined space available near the wing tip can be a problem. One solution is to move the front spar further aft outboard and to use a mechanism similar to that shown in Figure 5-46.

5.6.1.4 Suction/Ductin_ System The laminarized planform area, for the upper-surface-only case is illustrated in Figure 5-47. The boundaries of the laminarized areas are practical ones as noted below: o A 10 o wedge from the wing-fuselage intersection over the upper wing surface o The landing gear retraction area.

o The 80 percent span, 15 percent chord trailing edge flap area.

o The 20 percent span, 25 percent chord aileron.

o The area outboard of the wing tip closing bulkhead.

The manifolding at the dry bay area (approximately 40 percent span) for both laminarization cases was shown previously in Figure 5-39. With the suction ducting required further aft in the upper-surface-only case, the suction engine must also be mounted further aft. Detailed technical discussion of the suction ducting, manifolding, and suction engine is included in Section 10.0 of this report.

Additionaldrawings of the two suction ducting systems are shown in Figures 5-48 and 5-49. The ducting and manifolding is considerably simplified for the upper-surface-only case.

. r, r, OF F_t;_< < , _ : r /---'INSECT PROTECTIONFLUID _" 7//v / __ / I,> ." \_ 7_ / -_..._ ,/_-- _ _ > --

,'/- _ A_ -____--_+---_/ --

;_- '.. /"/ \ /' _'-" _,-_----'- _'---,_

\ <j

o GE h 223_2 FIGURE 5-44. LEADING EDGE DESIGN - UPPER SURFACE SUCTION ONLY

? ,/'

,_ FLUID SPRAY NOZZLE g GEN 21896A FIGURE 5-45. SPRAY CONCEPT

F.S.-_ 25PERCENT CHORDJ

83,.G EN. 23273 FIGURE 5-46. OUTBOARD WING SHIELD

Ci

LFC SUCTION AREA OVER WING PLANFORM 8-GiEN-22892 FIGURE 5-47. WING STRUCTURES/SUCTION DUCTING INTEGRATION - UPPER SURFACE ONLY LAMINARIZED O_i_ o,_., "_ "3 _ _ _ ,_ OF PUO_ QU,,.LII f DUCTING FROM UPPER SURFACE DUCTING FROM LOWER SURFACE =_.._ FIGURE 5-48.

INTEGRATION OF SUCTION DUCTING IN DRY BAY AREA - UPPER AND LOWER SURFACE LAMINARIZATION TO 70-PERCENT CHORD FIGURE 5-49.

INTEGRATION OF SUCTION DUCTING IN DRY BAY AREA - LAMINARIZATION OF UPPER SURFACE ONLY TO 85-PERCENT CHORD 5.6.1.5 Weights and Performance - LFC Aircraft The two LFC configurations shown in Figures 5-24a and b were sized using the K5JA computer program and sizing matrices as described previously under 5.3.7. The study ground rules remained as in Table 5-1. The interior layout, as shown in Figure 5-1, and operational items Table 5-18, were common to both. A comparative breakdown of LFC ducting system weights is presented in Table 5- 19.

From the Weight Summary Table 5-20 it can be seen that the upper-surface-only configuration results in a lower takeoff and empty weight and requires less fuel. Table 5-21 shows that the upper-surface-only configuration has a smaller wing area and smaller engine thrust required. Although it has a lower lift-drag ratio, it burns less fuel due to its lower weight. It also has a lower approach speed. There can be no doubt that the "upper-surface-only" configuration is superior. In addition to its better all-round performance_ it has all of the advantages listed previously under 5.6.1.

Table 5-18 WEIGHT BREAKDOWN OF OPERATIONAL ITEMS (SAME FOR LAMINAR AND TURBULENT AIRCRAFT) KILOGRAMS POUNDS (510) Cockpit Crew 59O

(1,3oo)

Cabin Crew

141 (31o)

Crew Baggage and Flight Kits

(2 4)

Oil

(5oo)

Unusable Fuel 1,994 (4,397) Food, Galley Service Equipment and Bev.

(1,196) Passenger Service Equipment (1,574) Potable Water

(200)

Lavatory Fluids

(1,47o)

Escape Slides/Rafts

(506)

Life Vests

(1,75o)

Pallets Containers (3,240) 7,810 17,217 OPERATIONAL ITEMS (TOTAL) Table 5-19 LAMINAR FLOW CONTROL SUCTION/DUCTING SYSTEM WEIGHTS LFC LFC UPPER SURFACE ONLY BOTH SURFACES 188,660 (415,930) Takeoff Gross Weight kg (LB) 183,400 (404,320) m (SQ FT) 288 (3,100) 331 (3,560) Wing Area 18,532 (40,855) 20,950 (46,186) Total Wing Weight kg (LB) (Including LFC)

LFC Suction/Ducting System kg (LB) 2,790 (6,150) 4,55o (lO,030)

699 (I ,540) 690 (1,520)

Suction/Ducting System kg (LB) Ducting-Porous Panel 2,091 (4,610) 3,860 (8,510) LE Weight kg (LB) Collector Ducts Suction/Ducting System Weight 15.1 21.7 Percent Total Wing Weight 1.5 2.4 Percent Takeoff Gross Weight Weight/Wing Plan Area kg/m 2 (LB/FT 2 ) 9.67 (1.98) 13.77 (2.82) Table 5-20 WEIGHT SUMMARY LFC AIRCRAFT COMPARISON LFC UPPER & LOWER LFC UPPER SURFACE SURFACE ONLY kg (LB) kg (LB) Wing 20,950

(46,186) 18,532 (40,855)

Horizontal Tail 2,076 (4,577) 1,735 (3,824) Vertical Tail 1,882 (4,149) 1,733 (3,820) Fuselage 18,434 (40,640) 18,312 (40,370) Landing Gear 7,604 (16,763) 7,377 (16,264) Propulsion System 11,330 (24,979) 10,881 (23,988) Fuel System (1,892) 803 (1,770) Flight Controls and Hydraulics 3,222 (7,103) 2,800 (6,172) Auxiliary Power Unit (1,080) 490 (1,080) Inst rument s 794 (1,750) 794 (1,750) Air Conditioning and Pneumatics 2,075 (4,574) 2,075 (4,574) Electrical 2,327 (5,130) 2,327

(5,13o)

Avionics (Including AFCS) 1,427 (3,146) 1,427 (3,146) Furnishings 16,400 (36,156) 16,400 (36,156) Anti-lce 194 (428) 169 (372) Auxiliary Gear

(62) 28

(62)

Manufacturer's Empty Weight 90,090 (198,615) 85,880 (189,333) Operational Items 7,810 (17,217) 7,810 (17,217) Operational Empty Weight 97,900 (215,832) 93,689 (206,550) Payload 31,298 (69,000) 31,298 (69,000) Zero Fuel Weight 129,198 (284,832 ) 124,987 (275,550) Fuel 59,465 (131,098) 58,409 (128,770) Maximum Design Takeoff Weight 188,663 (415,930 ) 183,396 (404,320) Table 5-21 LFC AIRCRAFT CHARACTERISTICS COMPARISON _k_- ' 5000 N MI RANGE 69,000 POUNDS PAYLOAD UPPER & LOWER SURFACES UPPER SURFACE LFC LFC TO 70 PERCENT C TO 85 PERCENT C POWER PLANT 3 ADVANCED TURBOFANS 3 ADVANCED TURBOFANS kn (LB) SLS Thrust/Engine 145.4 (32,690) 139.8 (31,430) WING Area m2(SQ FT) 331 (3,560) 288 (3,100) Sweepback, c/4 DEG 30 30 AR I0 I0 Taper Ratio 0.25 0.25 Airfoil t/cAV G 11.7 10.3 WEIGHT TOGW kg (LB) 188,663 (415,930) 183,396 (404,320) OEW kg (LB) 97,899 (215,830) 93,690 (206,550) Fuel Burned kg (LB) 49,745 (109,670) 49,260 (108,600) Fuel Reserves kg (LB) 9,709 (21,405) 9,147 (20,165) CRUISE CL 0.5 0.56 L/D 23.1 22.2 m/s (KN) 66.9 (130) 64 (124.5) VADproach Takeoff Field Length m (FT) 2,632 (8,635) 2,615 (8,580) The economic analysis in Section 6.0 will show that the "upper-surface-only" case also has lower initial cost and lower operating costs. The LFC configuration with suction on the upper wing surface only was therefore finally selected and compared with the turbulent configuration.

5.6.2 Turbulent Aircraft Comparison The final turbulent aircraft was shown previously in Figure 5-24c. The scissors plot used for horizontal tail sizing is shown in Figure 5-50. As in Study I, the critical forward limit for tail sizing is established by trim requirements at 9@ 1.4 Vstal I with full flaps and ice on the tail. The 21 percent c.g. travel obtained with a horizontal tail volume coefficient of 1.38 is adequate with f / °c wing mounted engines. The reduction from 1.52 used in Study I is due to the changes in wing aspect ratio, taper ratio, and airfoil thickness. The latter necessitated a movement of the main gear forward in order to obtain the depth required for housing the main gear.

Figure 5-51 shows the sizing matrix. The increase in takeoff gross weight from 187,243 kg (412,800 LB) for Study I to 191,853 kg (422,964 LB) is primarily due to an increase of operational items (cargo containers and pallets).

Table 5-22 compares the turbulent aircraft characteristics with those of the selected upper-surface-only LFC aircraft.

The LFC aircraft has a higher operational empty weight and wing area but due to its greatly improved lift/drag ratio, it burns 18.2 percent less fuel and the takeoff weight, engine thrust and field length required are lower.

9?

OF pO0_ QQp-.L|'r_( NEUTRAL POINTS _VFc 2.0 0.724 _ _ M= 0.70 2%NEG M = 0.60 _ m_,C_lJ/ MARGIN ESTIMATED AFT M = 0.56 1,8 TIPOVER LIMIT.--_ 1.6 FORWARD LIMIT TRIMMED AT

';;V,z ,%

VH PLANFORM /_EST,MATEOA.

1.4

r _ S, EER,NGU_,T

I _ NOSEWHEEL

/I ;o" x %_W%OLiM,T

.- o._,?EO A.

0.8 - --, _J 0.6 I I ; 10 20 30 40 50 60 70 CENTER OF GRAVITY, CG (PERCENT MAC) BZ _7.EN.2230_ FIGURE 5-50. ESTIMATED CENTER-OF-GRAVITY LIMITS - FINAL TURBULENT CONFIGURATION ASPECT RATIO = 10.85 M CRUISE = 0.80 1000 kg (1000 LB) 200- (440)- 3_e E#GII_IE e._ _soF_

195 - (430)

TAKEOFF GROSS WEIGHT (420) 190 -'_-_1t"_ TAKEOFF RUNWAYLENGTH - tO,ON FT ,: ,j2:, :2:,;EE _Lj,::: 2E "J"°"'" _,,U.E,, (410) I_dEl_ 2¢_42 • 1 t';' FIGURE 5-51. TURBULENT AIRCRAFT SIZING Table 5-22 AIRCRAFT CHARACTERISTICS COMPARISON TURBULENT VS. LAMINAR 5000 N MI RANGE 69,000 POUNDS PAYLOAD UPPER SURFACE LFC TURBULENT TO 85 PERCENT C 3 ADVANCED TURBOFANS POWER PLANT 3 ADVANCED TURBOFANS SLS Thrust/Engine kNI(LB) 147.9 (33,240) 139.8 (31,430) WING Area m2(SQ FT) 260 (2,800) 288 (3,100) Sweepback, c/4 DEG 30 30 AR 10.85 10 Taper Ratio 0.25 0.25 Airfoil t/cAV G 12.7 10.3 WEIGHT TOGW kg (LB) 191,854 (422,965) 183,396 (404,320) OEW kg (LB) 91,401 (201,505) 93,690 (206,550) Fuel Burned kg (LB) 60,217 132,755 49,260 (108,600) Fue7 Reserves kg (LB) 8,936 (19,700) 9,147 (20,165) CRUISE CL 0.58 0.56 L/D 17.5 22.2 m/s (KN) 63.5 (123.5) 64 (124.5) VApproach m (FT) 3,048 (I0,000) 2,615 (8,580) Takeoff Field Length The economic analysis, Section 6.0 will show that even using the study ground rule of only 12¢ per liter (45¢ per gallon) the LFC aircraft has a lower direct operating cost (DOC). At a more realistic cost of 26_ per liter ($i per gallon) the DOC for the LFC aircraft would be 6 percent less than for the advanced turbulent aircraft.

5.7 OPERATIONAL CONSIDERATIONS

5.7.1 FAA Regulations For commercial aircraft to obtain the fullest possible advantage in terms of drag reduction and fuel savings from the use of laminar flow technolog_ it may be necessary that the Federal Aviation Regulations be modified. Two areas of particular concern are: o Reserve fuel requirements; o Dispatch with all or part of the laminar flow control system inoperative.

When new technology is incorporated into new commercial aircraft, the Federal Aviation Administration often issues "Special Conditions" that outline new requirements for the certification of aircraft using this new technology.

Such action is likely to occur with the introduction of laminar flow control.

An area of special concern to the FAA will be fuel reserves if all or a portion of the laminar flow control system becomes inoperative.

Under present FAA regulations it is very likely that the fuel reserves would have to be determined assuming that the laminar flow control system failed at the most critical time. The FAA probably would not allow the increased fuel consumption resulting from the loss of laminar flow to be taken out of existing fuel reserves. However, for flights over land and with adequate airports available enroute, a modified "re-clearance" procedure could be established. This procedure would permit full advantage to be taken of the fuel savings provided by a fully operational laminar flow control system. In the unlikely event that the system failed enroute on a long range mission, a landing could be made, if necessary, short of the final destination in order to replenish the fuel before continuing to the final destination.

For long over-water operations, the extent that "re-clearance" could be used would depend on the location of suitable airports along the route.

I OO Failure of the laminar flow system does not make the aircraft inoperable or unsafe in any way. Commercial airplanes incorporating laminar flow control would be designed so that they could be dispatched with all or part of the system inoperative. To achieve this capability the manufacturer and the FAA would need to work in close cooperation commencing with the initial design of the airplane and its systems. This dispatch capability may require special conditions to Federal Aviation requirements.

5.7.2 Airline Comments Discussions have been held with both United Airlines at San Francisco and Flying Tigers Airlines at Los Angeles International Airport. Douglas has agreements with both airlines for consulting on this LFC contract. Both airlines represent large carriers which differ in the principal emphasis of their operation. United is representative of the large domestic and overseas commercial passenger business, and Flying Tigers is concerned with transporting cargo over their long range routes. The long range routes are particularly compatible with LFC.

The discussions included a review of the Douglas laminar and turbulent aircraft Configuration drawings and performance. Both LFC concepts and operational aspects were reviewed. The comments of the two airlines, relative to LFC, are consistent. Both airlines felt that with the amount of fuel that could be saved, the acceptability of an LFC aircraft is not dependent on its having a lower DOC than conventional aircraft. When evaluated at the average trip length of the consulted airlines operations, the DOC of the laminar aircraft was considered to be acceptable. The airlines are looking to LFC as both a fuel saving measure and as a hedge against limited fuel allocations.

In the case of United Airlines, the cost of an LFC aircraft would be depreciated over the assumed lifetime of the aircraft. They saw no logical reason for considering LFC as a special case requiring an accelerated rate of depreciation. However, Flying Tigers initially viewed the cost of LFC as a modification cost and, therefore, felt that this cost should he depreciated over a shorter time period, such as five years.

Throughout the LFC aircraft configuration and integration study work, the airlines have reacted positively towards laminar flow. The following summaries of airline comments best express their positive attitude whenever they were consulted.

AIRLINE'S COMMENTS -- CONCEPT United and Flying Tigers - June 1977 Both Airlines View LFC Favorably LFC is an Attractive Hedge Against Fuel Allocation LFC is Acceptable at Same DOC as Conventional DOC's Must be Evaluated for Average Trip Length- Not at Design Range Depreciation Over Life of Aircraft--United Depreciation of LFC Over 3 to 5 Years--Flying Tigers High Interest in Porous Materials --Honeycomb Unsatisfactory in Landing Gear Area--United --Honeycomb Unsatisfactory--Flying Tigers --Isogrid Unacceptable--Flying Tigers AIRLINE'S COMMENTS--OPERATIONAL United and Flying Tigers--June 1977 o Douglas Operating Assumptions are Conservative o LFC Suction System Should Have Same Reliability As Main Propulsion System o LFC System Must Not Be Required For Dispatch o FAA Rules are Subject To Review For LFC Application o Maintenance of Surface Cleanliness is a Major Concern AIRLINES' COMMENTS United and Flying Tigers--May 1978 LFC AFFORDS MAJOR ADVANTAGES o Fuel Reduction o Hedge Against Fuel Allocation UPPER-SURFACE-ONLY LAMINARIZATION UNQUESTIONABLY PREFERRED Maintenance Ease and Efficiency Elimination of Leakage Problem Elimination of Under-Wing Surface Frost Problem FUEL SAVING ADVANTAGES DUE TO LFC FAR OUTWEIGH o System Maintenance Costs o Added Cleaning Required for LFC Aircraft T#, AIRLINES' COMMENTS United and Flying Tigers--Sept 1979 Fuel Availability is an Important Consideration Reduction in DOC Due to LFC More Than Adequate For Serious Consideration by Airlines Suggested Additional Evaluations Be Made Using o Shorter Stage Lengths/Lower Utilization o 1980 Economics Validator Aircraft A Necessary Step Airline Interest in LFC is Becoming Serious as the Program Advances and Fuel Prices Increase

APPENDIX 5-A DRAG ESTIMATION CRUISE DRAG The estimated cruise drag for the laminar and turbulent baseline aircraft can be broken into three parts: the parasite, the induced, and the compressibility drags, as indicated in the following equation: f CL CD + + A CDc : Sw _ _R-e Parasite Induced Compressibility Drag Drag Drag Parasite Drag The form drag is estimated using established form factors from Douglas and Hoerner data. The skin friction coefficients for the turbulent components are based on the method of Van Driest as adjusted for surface roughness by Clutter. A sand-grain roughness level of .024mm (0.00095 in.) is used, which is representative of typical transport aircraft. For the wing of the laminar aircraft without LFC, the sand grain roughness is assumed to be zero. With LFC, the equivalent skin friction drag (wake drag) and the suction quantity are assumed to vary with Reynolds number as shown in Figure 5-AI. These results are based on a preliminary airfoil analysis conducted prior to the start of the base case selection process and are being refined as part of the detailed airfoil design process.

The resulting parasite drag breakdown for the laminar and turbulent initial base case aircraft of Study I is shown in Table 5-A1. The drag due to the fuselage canopy and upsweep, wing twist, and control surface gaps are ESTIMATED FLAT PLATE WAKE AND SUCTION QUANTITY = 30 DEC, TRANSITION AT X/C = 0.7, BOTH SURFACES M® = 0.80, ^c/4 0.003 0.002 - EQUIVALEN_ C D CDw -- BASED ON BOUNDARY LAYER MOMENTUM LOSS w AND C AT TRAILING EDGE q C 0 -- BASED ON THE MINIMUM SUCTION REQUIREMENT 0.001 FOR LFC WING _FlClENT, C I I I I I I I I I I I I I I J I J 107 10 a 10 6 CHORD REYNOLDS NUMBER 113-GE N,Z_28 uc594 c_rol FIGURE 5-A1. PARASITE DRAG - LAMINAR AND TURBULENT ORIGINAL PACE re,,..

OF POOR QUALITY

©

1 06 individually accounted for. The miscellaneous drag accounts for items such as o steps, vents, etc., and is equal to 7.8 percent of the total friction, form, and roughness drag (average for Douglas transports) for the turbulent base case. Due to the cleaner wing required for LFC, this factor is estimated to be equal to 5.4 percent for the laminar base case. An additional 1.2 percent is included for potential interference drag.

TABLE 5-A1 LFC INITIAL BASE CASE AIRCRAFT PARASITE DRAG Turbulent Laminar Base Case Base Case With LFC Without LFC Friction, Form Roughness m2 (Ft 2) m2 (Ft 2) m2 (Ft 2 ) Fuselage 1.649 (17.75) 1.649 (17.75) Wing 1.647 (17.73) 1.013 (10.90) 2.246 (24.18) Flap Fairing .102 (1.10) .182 (1.96) Horizontal .421 (4.53) .396 (4.26) Vertical .261 (2.81) .346 (3.73) Nacelles and Pylons .593 (6.38) .619 (6.66) Subtotal 4.673 (50.30) 4.205 (45.26) 5.438 (58.54) Canopy .010 (.11) .010 (.II) Upsweep .017 (.18) .017 (.18) Twi st .025 (.27) .027 (.29) Gaps

.018 (.19) .022 (.24)

Miscellaneous .370 (3.98) .232 (2.50) Interference

.061 (.66) .054 (.58)

TOTAL 5.174 (55.69) 4.567 (49.16) 5.801 (62.44) Induced Drag Efficiency Factor e .819 .922 .832 '., . "71_ 1 o?

Induced Dra 9 The induced drag is based on the Giesing vortex lattice lifting surface program for inviscid flow plus a factor, based on Douglas flight test data, to account for the variation of parasite drag with lift for the fuselage and wing, due to viscous effects. For the laminar base case, it is assumed that with LFC operating the viscous contribution of the wing to induced drag is equal to zero. The resulting efficiency factors (e) for the base case aircraft are shown in Table 5-AI.

Compressibility Dra_ The fuselage compressibility drag is calculated using the current Douglas method based on isolated fuselage wind tunnel data. For the turbulent base case, the wing compressibility drag, which is a function of CL and Mach number, is calculated using Douglas design charts based on 2-D wind tunnel tests of advanced airfoils. Due to the shock-free design of the laminar base case wing, the compressibility drag is assumed to be zero. Trim drag is assumed to be equal to 25% of the total aircraft compressibility drag, based on current Douglas studies of reduced stability level configurations.

©

6.0 ECONOMICS ANALYSIS This section of the report presents the results of the cost analysis generated for the LFC concepts and the turbulent baselines and the aproach that is taken to derive the cost data. Included is a brief narrative of the derivation of the cost data and the ground rules, guidelines and assumptions used. This section includes the acquisition costs for both the turbulent and LFC configurations as they occur in time, along with the operating costs.

6.1 AIRPLANE COST ESTIMATING APPROACH The cost estimates are used to evalute the LFC concept compared to a turbulent baseline. These estimates are also used to examine the sensitivity of operating cost to fuel price. Proven cost analysis techniques are used to provide a basis for the evaluation. Estimates are derived using a systematic approach to predicting cost behavior in the future, on the basis of current capability and the expected advancements in technology. However, the relationship of technology to cost behavior is more subtle than can be expressed by continuous functions and trend analysis methods. The airplanes in this study are configured partly with current technology design concepts, materials and manufacturing methods; these cost estimates are derived using Douglas standard techniques for advanced design studies. The advanced technology elements require the use of judgement based on scientific and practical experience and background. These judgments are used to modify the existing engineering analysis techniques to reflect the anticipated cost impact of the advanced technology. Experience in cost estimating for advanced technology studies coupled with prototype production and tracking of actual costs (labor and materials) versus the estimates, formed the basis for judgment.

Estimates for the conventional segments of the aircraft are derived using discrete type estimating techniques that parallel the industrial engineering methodology supplemented by parametric cost estimating models.

@

No attempt is madeto establish or justify a market size, a particular market, the number of manufacturers and/or pricing strategies. An arbitrary production quantity of 400 aircraft and a single manufacturer is selected and used in this analysis. The market size could be larger with more than one manufacturer, but this is not relevant to the outcome of the cost comparison between LFC and turbulent designs. It is assumed that an incentive of a 20 percent profit with a production of 400 aircraft is sufficient to attract the manufacture. This establishes the quantity selected and the price level achieved. The methodology incorporates a computation to account for the interest or cost of money with respect to financing the project.

Common guidelines that are applied to each configuration evaluated are presented in Table 6-1. The DOC factors and coefficients used are shown in Table 6-2.

Table 6-1 GENERAL COSTING GROUND RULES -'-T ROM Costing Level of Estimating 1976 Dollars Used Throughout 14-Year Aircraft Life 5000 Hr-Per-Year Utilization 400 Aircraft Production/Single Manufacturer 45-Cent-Per-Gallon Fuel* 0 Modified 1967 ATA DOC Equations Used Addition of Landing Fees & Cabin Attendants Factors and Coefficients Based on Douglas Experience With Operators * Effect of Fuel Cost Variation from $0.45 to $2.50 Per Gallon Included in the Study

Table 6-2

DOC FACTORS ANDCOEFFICIENTS

Landing Fee ($) 1.70

Cabin Crew ($) 25.48

Cockpit Crew ($) 278.62

Deprec iati on (Yr) 14

Insurance Rate (%) 1.5

Residual Value (%) 10

Spares (%) 15

Labor Rate Per Hour ($) 9.23

Labor Burden (%) 180

Fuel Per Gallon ($) 0.45

6.2 COSTSUMMARY

The cost analysis includes the initial baseline configurations of Study I

(Section 5.3), the three and four engined configurations of Study II

(Section 5.4), and the final configurations of Study IV (Section 5.6). A

summaryof significant aircraft characteristics considered in the cost

analysis is presented in Table 6.3.

A top level summaryof the cost data for the configurations examined in this

study is shown in Table 6-4. The summaryresults contain only the flyaway

costs and direct operating costs, which are the significant measures. Flyaway

cost is defined as the cumulative average of the total quantity produced.

Estimates developed for this study are to be considered only as Budgetary and

C

OF P©OF_ QUALIi_

d _ m

000_ ,---_ OJ r_ N z z o rY I n, Z QJ _ IJ.

I-- O'9 C.O p--- OOh I_ -P- _o I.iJ I--- 0 c'_

;I-'

c-- c- "r" O 0 r_o'_ 0 0 _r.._

°'ii

C c'- ..-J --J II II A Z_-_ %/ .-.I ! I I, LL ILl Z Z ,--J--J A A n,,,, _._J C_'_._.i _..-_ c_n _,.i Z Z • ,j ,,_.., v' _.._ I I I I _' i,-- ._1 Lid 1.1.1

C

I-- Z I'-- OF POOR Qu_ __, i Y z 2o _iI. _ _ ,-q I'_

I_,1 "-

>. O'_ ,-I ('M _1" ,-' A _. I _ _

-5

>.

I:I: w:E • . _ l_c_ _ _ ZZZ o _ ZZZ (J O _ _ ZZ _w lii _Z

8 _ _ _ oo

_ _ Z II II li li li

C_

Planning costs intended for relative rather than absolute use, in consonance with the primary purpose of such costs to be used in the development and comparison of candidate concepts.

The configurations analyzed are presented chronologically with time as the study proceeded. The final three configurations of Study IV are those on which attention should be focused; they incorporate the most recent updated technical parameters, and resultant costs. Comparisons should be limited to those within each particular study. The final LFC configurations are distinguished by the extent to which the LFC is applied to the wing - i°e., upper and lower surface, and upper surface only.

The acquisition and DOC costs of the finally selected LFC configuration with upper surface suction only, and the turbulent aircraft of Study IV are compared in greater detail in Tables 6-5 and 6-6.

C:_:z:=:i; ";, .;. i. ;,, _;_ OF PC;C,.__.,_ .'-,,..:iY D-- Z L_J (J LJJ I-- I,I I .-J .-J_ _,.-J,-C_,d,._ ,d ,=En_ r-.r--_o,-,_>.Qo _J oo

II

F-I-- ILl(j ¸ {:El, UJ..J I, I, c- O °p.

i--- p- °f..

O ,=,, r',,, Oh L_ C:) "_t Z (" E:_ °P" I-_ _C M::) ,_t',-_ _C, CO l (/) O U '_E r...

I--- p- P =E O P L_ (.& Z U !

I--- r_ ___-I E (/}oh _U E:) ,.._ _J .2 z_ _ I _ Ill q- e- Cr (j-l_ E f..

U {.- c- (IJ O _E 0p..

IlJ U E_ % e- e-- C_ --A ,,=, I'-- r" V) (/) I-- °r.- Q _ UJ I-- .1_ e- "_ _j e" E e" -i-- {:: e- _'- O" _'} °v- L _ C/ °_-- ,r- L • I-) O_ I--- ._ -r- I.iJ I::---_ =E .1:: ('- L E: O I_-.- O'_ U e- 4_ =E I-- _. 1.1. LIJ I.-- LaJl--- ¢f_ ¢v'_.L_l--._ I--

Table 6-6

DOC COMPARISON

Constant 1976 Dollars

LAM I NAR

TURBULENT (USO)

Landing Fee

Cabin Crew Cost

3,443 3,336

Cockpit Crew Cost 4,297 4,215

Depreciation/Insurance 9,426 8,949

Maintenance

5,830 5,634

Labor

(3,130) (2,790)

Materials

(2,700) (2,844)

Fuel * 7,294

8,916

Dollars/Flight 30,978 31,770

Dollars/Aircraft Km (NMI) 3.34 (6.19)

3.43 (6.35)

Cents/Seat Km (NMI)

1.12 (2.07) 1.15 (2.125) * Assuming Fuel Cost = 12_/liter (45_/gallon) 6.3 SENSITIVITY OF DOC COMPARISONS TO FUEL PRICE f The fuel price of $O.12/liter ($O.45/gallon) was stated as a ground rule in the Study Contract NAS1-14632; however, this fuel price became obsolete during the study due to the fuel shortage and associated rapid increase in fuel costs. Consequently, sensitivity to fuel costs, varying from $0.12 to $0.66 per liter ($0.45 to $2.50 per gallon), on the relative DOC advantages of the final LFC aircraft over the turbulent aircraft were investigated and the results are shown in Figures 6-I and 6-2.

As shown in Figure 6-1, over the life of the aircraft, the LFC aircraft is estimated to save approximately $45 million over the comparable turbulent aircraft, assuming an average fuel price of $0.50 per liter ($1.80 per gallon). This is equivalent to the price of the airplane. Even at $0.12 per liter ($O.45/gallon), the saving would have been $12.5 million.

In terms of comparative percent DOCreduction, Figure 6-2 shows an 8 percent

advantage in DOC for the LFC aircraft over the comparable turbulent aircraft at fuel prices predicted for 1981 to 1984. Also shown on Figure 6-2, is the 2 percent advantage in DOC for LFC on the upper surface only compared with LFC on both wing surfaces.

The DOC reduction of approximately 8 percent in favor of the LFC aircraft in the 1990 time period is significant and warrants continued development of the LFC concept.

ORIGJI_JAL PAGEiS

OF POOR QUALITY

LFCus O RELATIVE TO TURBULENT AIRCRAFT

t [ i I

ASSUMPTIONS: I4-YEAR LIFE, 5000 FLIGHT HOURS/YEAR I I I I FUEL SAVING: 24381,480 GALLONS/AIRCRAFT EQUIVALENT TO A FREE AIRPLANE _ " / FUEL COST I /F, SAVINGS 40 I_......_STI M AT E D SPREAD PER AIRCRAFT F 1990 FUEL PRICES 3O ($ MILLIONS) LF__CSTUDY GROUND RULE 0.1 0.2 0.3 0.4 0.5 0.6 0.7 r 1 I | ) (1.0) (1.5) (2.0) (2.5) (0.5) FUEL PRICE $ PER LITER (GALLON) FIGURE 6-1. EFFECT OF FUEL PRICE ON FUEL COST SAVINGS LFC RELATIVE TO TURBULENT AIRCRAFT LFC (USO) ESTIMATED 1981-1984 FUEL PRICE -'-7 • i DOC 8 REDUCTION '- LFC (U+L) I (LFC RELATIVE TO TURBULENT AIRCRAFT) 4 (PERCENT) STUDY GROUND RULE

l

0.6 0.1 0.2 0.3 0.4 0.5 L i I i J (0.5) (1.0) (1.5) (2.0) (2.5) FUEL PRICE $ PER LITER (GALLON) FIGURE 6-2. REDUCTION IN DOC AS A FUNCTION OF FUEL COST 7.0 AERODYNAMIC DESIGN AND ANALYSIS 7.1 AIRFOIL DESIGN AND SUCTION REQUIREMENTS The purpose of this study was to develop a series of airfoils that would exhibit aerodynamic characteristics compatible with those necessary for successful laminar boundary layer flow control. In addition, it was required to determine the sensitivity of these aerodynamic characteristics to variations in Reynolds number, airfoil thickness and the extent of laminarization. These data were needed to support configuration and suction system design tasks of the overall study.

The procedure used was to develop a series of airfoils having varying thickness-chord ratios and then determine the suction flow quantities and wake drag characteristics of each airfoil as a function of Reynolds number and chordwise extent of laminar flow. Two sets of airfoils were designed corresponding to the normal sections of 25 degree and 30 degree swept wings.

Since it is unlikely that laminar flow can be maintained behind the shock downstream of the supersonic flow region on a conventional supercritical z ¸ • • airfoil, the LFC airfoils were designed to remain shock-free. Furthermore, -4.. ¸ from operational considerations it is important to prevent buffet in the event that laminar flow is lost during flight. This condition requires that aft pressure recovery gradients satisfy separation criteria for turbulent boundary layer flow on the laminarized portion of the airfoil. While the likelihood of suction system failure is remote, the possibility of partial loss of laminar flow due to environmental disturbances must be considered.

7.1.1 Design Guidelines for LFC Airfoils A two-dimensional infinite swept wing design procedure was used to develop LFC compatible airfoils. The upper surface pressure distribution, normal to the leading edge, was similar for all airfoils. The upper surface pressure peak was constrained to limit the maximum local Mach number to 1.1 near the leading

edge, with gradual supersonic compression to a local Mach number of 1.02 in

the vicinity of 65 percent chord. Over the aft portion of the airfoil the

adverse gradient was limited so that separation would not occur if laminar

flow were interrupted or lost.

The lower surface pressure profile was varied to generate airfoils of

different relative thickness and corresponding design section lift

coeffi cients.

A slightly favorable pressure gradient was maintained from the leading edge to b5 percent chord on the lower surface with the adverse gradient near the trailing edge constrained to preserve attached flow with fully turbulent boundary layer.

7.1.2 Airfoil Anal>,sis Airfoil profiles were developed for the specified pressure distributions using the Tranen code (Reference 7.4-1) which is an inverse transonic analytical procedure. This method is an extension of the 2-D Garabedian airfoil analysis method. Boundary layer analyses and suction requirements for the various airfoils were determined using the Cebeci boundary layer program of Reference 7.4-2. A specialized version of the program was developed to compute suction velocities necessary to satisfy boundary layer stability criteria.

Laminar flow control airfoils based upon the foregoing criteria and methods were developed for a free stream Mach number of 0.8. Two wing sweep angles were considered, 25 degrees and 30 degrees with corresponding normal Mach numbers of 0.725 and 0.70, respectively. Pressure distributions and resulting airfoil geometry for representative cases are shown in Figures 7-I and 7-2.

For the 25-degree swept wing, the design section lift coefficients are 0.73 for the 11.3 percent thick airfoil and 0.54 for the 14.3 percent thick airfoil. Similarly_ for the 30-degree swept wing the design section lift coefficients are 0.78 and 0.60 for the 12.8 percent thick and 15.8 percent thick airfoils, respectively.

OF POOi_ Q:SJ_LI-I'Y Moo= 0.8, A= 25 DEG -1.0 UPPERSURFACI_S -0.5 t/c = 113% LOWERSURFACES

C

Cp 0.0 t/c = 14.3% I 0.5 1.0 20 40 60 80 100 PERCENT CHORD 7-G EN-22_)45 FIGURE 7-1. TYPICAL LFC WING CHORDWISE PRESSURE DISTRIBUTIONS AND AIRFOIL SHAPES - 25-DEGREE SWEEP

C

M==0.80, A=30DEG UPPER SURFACES -1.0 tic = 12_%

\

-0.5 LOWERSURFACES Cp tic = 15.8% 0.0 tic = 121_r _J \___ 0.5 _J 1.0 0 20 40 60 80 100 PERCENT CHORD 7-GEN-2264B FIGURE 7-2. TYPICAL LFC WING CHORDWISE PRESSURE DISTRIBUTIONS AND AIRFOIL SHAPES - 30-DEGREE SWEEP Estimated variation of airplane lift coefficient with average streamwise wing thickness is summarized in Figure 7-3. The relationship between two-dimensional design lift coefficient, C1, and airplane lift coefficient, CL, is given by: CL = C1 cos2_.

1.17 where.A.is the wing sweep angle and the empirical factor 1.17 compensates for spanwise load distribution and airplane trim effects. For comparison, the corresponding variation of wing CL and average thickness for a series of advanced supercritical airfoils, on a 3U-degree swept wing, is included in Fi Qure 7-3.

7.1.3 Suction Requirements Suction requirements for the laminar flow airfoils were based on the Cebeci boundary layer analysis for a swept wing (Reference 7.4-3) used in conjunction with the laminar boundary layer empirical stability criteria developed during the Northrop X-21 program (Reference 7.4-4). The analytical procedure and empirical stability relations were combined into the method and computer program described in Reference 7.4-2. A subsequent modification of the computer program provided an interactive mode through a remote graphic terminal. This modification greatly expedited the solution of suction requirement s.

The boundary layer stability criteria developed during the X-21 program are listed below: (1) Attachment line instability-momentum thickness Reynolds number, R , at e 0 the leading edge dividing streamline.

Re9 0.4044 sin.A.l.e ) _-" _00 i ' ., r,o.

OF PO0;,_ c_2.L;,,':, :'CJ OF POOR Qb,4.L_iY where: R - _J'_9--c c _ ' U cross-flow velocity in boundary layer, normal n to streamline at edge of boundary layer at chordwise station.

(2) Tollmien-Schlichting instability -- Momentum thickness Reynolds number required in presence of streamwise disturbance.

2 USe Re _< .6 - 10602 _ y2 wall

us ]3

_) sUse ; R - whe re: e9 W _ total velocity at edge of boundary layer at Us e chordwise station.

-

es

streamwise momentum thickness at chordwise station (3) Cross-flow instability -- cross-flow Reynolds number required in the presence of a component of boundary layer flow as a consequence of wing sweep (sheared pressure gradients).

U

/ )

< 1.8 F|57 - 0.72 _--2(Un nmax Ren - L _ (y / Y0.1)2 U nmax YO.I where: R : en _,_ maximum cross-flow velocity.

Unmax Yo.I = y at u n : 0.1 Unmax , reference dimension of boundary layer cross-flow velocity profile at chordwise station.

Determination of the design suction for a given geometry and pressure distribution was accomplished by computing the suction necessary to satisfy the appropriate stability criteria above. The procedure consisted of a "marching" solution, beginning at the attachment line and progressing downstream. At each chordwise station_the critical instability was identified and the corresponding suction velocity requirement was calculated using an iterative procedure based on an estimated initial value. The results from this procedure are limited in that there is no interaction by which down stream conditions are allowed to affect upstream requirements.

Typical results showing the upper and lower surface suction velocity distributions are presented in Figures 7-4 and 7-5, respectively. These data are for a 30 degree swept wing at M = 0.80 and CL = 0.502 with three values of wing chord, representative of the root, mid-span, and tip chords of a tapered wing. The different chord lengths change the incident Reynolds number, Rc spanwise. This significantly affects suction requirements in the adverse pressure gradient region over the aft third of the airfoil. Also the chordwise "scale" affects the attachment line suction requirement as the leading edge radius and attachment line Reynolds number vary. The most significant effects of chordwise dimension occur at the leading edge !

(attachment line) and at the aft pressure recovery region. Strong cross-flow instability conditions are characteristic of the latter re qion.

7.1.4 Variation of Suction with Airfoil and Flight Parameters Based on suction velocity distributions for the various airfoils and flight conditions_a total suction flow coefficient can be determined. The suction flow coefficient, CQ, is obtained by integration of the required suction velocity distribution to the chordwise extent to which laminar flow is maintained. Thus, x/c "_ OF POOF_ _UA=RV Using this definition, suction flow coefficients were determined as a function of several operational and configuration parameters.

OF POOR _U;.L_',"/ CHORD REYNOLDS NUMBER = 36 x 106 M=, = 0.8 30-DEGREE SWEEP 0.8 I _-_ J := CL = C t • COS2A 0.7 i tb,,_ _@,,,., i i I "'='O_,=._.,V//.,_,. = AIRPLANE 0.6 LIFT COEFFICIENT

_:._',_ ""'" _._."t _ i I

(CL) 0.5 0.4 be,i D I

0.3 I i

8 9 10 11 12 13 14 15 16 AVERAGE WING THICKNESS--CHORD RATIO (PERCENT) 7-GEN-22657-1 FIGURE 7-3. VARIATION OF DESIGN LIFT COEFFICIENT WITH AIRFOIL THICKNESS .f Moo--0.8 C L = 0.502 30-DEG SWEPT WING ALTITUDE= 10,970m (36,000 FT) m/s FT/SEC .-. --0"5 1 (--1"6)0, (--1.4) I >--'-°'I , c,o.,, I I I I I /I _.

m53m (19FT) CHORD x ;ATrACHMENTLINEVALUES/ |

_-o._ (_1o)1 x_(_rr) CHORO o 7-1 I / m

I

(--0.8)_-10.7m(35 FT)CHORD • j- / 5_CHORD

o (o]

10 20 30 40 50 60 70 80 90 ARC LENGTH (PERCENT CHORD) 7-GEN- 22678"1 FIGURE 7-4. TYPICAL SUCTION DISTRIBUTIONS - UPPER WING SURFACE The variation of suction coefficient with airplane design CL for a 30 degree swept wing with laminarization to 7[}percent chord is given in Figure 7-6. As the lift coefficient increases, the lower surface pressure distribution is modified to alleviate the lower surtace aft pressure gradient and reduce the suction required on the lower surface. Since the airfoil design criteria require the upper surface pressure profile to remain essentially unchanged with change in lift, the upper surface suction requirement also is unchangea with lift coefficient.

An alternative presentation of suction flow coefficient, as a function of airfoil thickness (in percent chord), is shown in Figure 7-7 and 7-8 for 30 degree and 25 degree swept wings. These data are for suction applied to 70 percent chord and show the same characteristic variation with section lift. The lower surface suction flow coefficient increases with increased airfoil thickness as expected, since airfoil thickness varies inversely with the design section lift coefficient. The variation of CQ with airfoil thickness, or CL, is greater for the 25 degree swept wing than for the 30 degree swept wing.

Suction requirements were also computed for a representative 30 degree swept airfoil as a function of chord Reynolds number. Chord Reynolds number was varied by changing unit Reynolds number with fixed chord and by changing chord length for fixed unit Reynolds number. The results are presented in Figure 7-9. Two cases are shown: one with suction to 70 percent chord and one with suction to _5 percent chord. It is evident that the adverse gradients and consequent increased suction requirement (Figures 7-4 and 7-5) result in the total suction flow quantity required to almost double as laminar flow is extended from 70 to 85 percent chord on both surfaces. Included in Figure 7-9 for comparison is the classic assumption of suction varying inversely withthe square root of chord Reynolds number. Such an assumption is optimistic with respect to the effect of increasing Reynolds number. The representative airfoil used has a 12.8 percent normal thickness and normal section lift coefficient of 0.783.

Moo=0.8, CL-0.502 30-DEG SWEPT WING ALTITUDE = 10,970m (36,000 FT) (-1.4) l_m (6 F:T) cHoRD -0.4 / \ (-1.2) / / \ \ (-i.o) -0.3 / p , 5.8m (19 FT) CHORD (-o.8) SUCTION

I[/

VELOCITY 10.7m (35 FT) CHORD_ LOWER SURFACE (V w ) -0.2 (-o.6) (-0.4) -0.1 (-0.2) !

(o) 80 90 100 FPS 10 20 30 40 50 70 m/s 81GEN 2262b ARC LENGTH (PERCENT CHORD) FIGURE 7-5. TYPICAL SUCTION DISTRIBUTIONS - LOWER WING SURFACE I Moo=0.8 30-DEG SWEPT WING SUCTION TO 70% CHORD 0.0005 0.0004 SUCTION 0.0003 FLOW COEFFICIENT 0.0002 (CQ) UPPER SURFACE L.'-.

m mm=m'mn n E n ! m _ i bm_ q 0.0001

"_o'-w'_'._-_c _ •

03 0.4 0.5 0.6 AIRPLANE LIFT COEFFICIENT (C L) 7-GEN,22664 FIGURE 7-6.

VARIATION OF LFC SUCTION COEFFICIENT WITH DESIGN CL OF POOR Q_;Ai_ITY / 0.0005 I i ! 1 1 I I Mo_=0.8 CHORD Re = 36 x 106 30 ° SWEPT WING

I

SUCTION TO 70% CHORD I 0.0004 SUCTION 0.0003 BOTH SURFACES_ FLOW COEFFICIENT =_ (CQ) 0.0002 UPPER SURFACE ONLY 7 "E, m,n.I _jwm 0.0001 I'-LOWER SURFACE ONLY 8 9 10 11 12 13 14 15 16 AIRFOIL THICKNESS-CHORD RATIO (PERCENT) BI_C_EN.22315 FIGURE 7-7. VARIATION OF SUCTION REQUIRED FOR LFC WITH THICKNESS-CHORD RATIO - 30-DEGREE SWEEP Moo=0.8 CHORD R e =36 x 10 6 25 ° SWEPT WING 0.0005 I t SUCTION TO 70% CHORD 0.0004 SUCTION 0.0003 FLOW COEFFICIENT BOTH SURFACES/-7.

(CQ) 0.0002 .-_ sUPPER.SURFACE ./.. .... ,,.,_ --..: 0.0001 LOWER SURFACE 0 I I 8 9 10 11 12 13 14 15 16 AIRFOIL THICKNESS-CHORD RATIO (PERCENT) 8) -GEN223|4 VARIATION OF SUCTION REQUIRED FOR LFC WITH THICKNESS-CHORD FIGURE 7-8.

RATIO - 25-DEGREE SWEEP Airfoil profile drag (wake drag) was calculated for the LFC sections and was found to have only a very small variation with thickness ratio and sweep at the design CL. This is due to the similarity of the design pressure di stributi ons.

Variation of profile drag for the representative airfoil used in the preceding discussion, as a function of Reynolds number is shown in Figure 7-10.

Extending the suction from 70 percent to 85 percent chord reduces the profile drag by approximately one half. However, as noted previously, this extension of laminarization requires a severe increase in suction required. This situation suggests a very practical alternative in which suction is applied on the upper surface only back to 85 percent chord. Profile drag for such a case is included in Figure 7-10. Drag for upper surface suction to 85 percent, without LFC on the lower surface, is only slightly higher than the drag for both surfaces laminar to 70 percent chord.

LFC aircraft configuration studies (see section 5.6.1) showed that using upper surface suction only, to 85 percent chord, resulted in an overall lighter and more efficient aircraft. Greater effective structural depth and a high lift device at the leading edge of the wing are possible with suction only on the upper surface. A lighter wing weight results, which more than compensates for the slightly greater profile drag. In addition this configuration enhances wing accessibility for fueling and maintenance and significantly alleviates requirements for the environmental systems.

7.1.5 Effect of Loss of LFC Aerodynamic consequences of the loss of LFC were investigated by computing the effect such loss would have on airfoil characteristics. The results are given in Figure 7-11. At constant angle of attack, the section lift coefficient is reduced by 20 percent from 0.783 to 0.025. In order to maintain a constant lift coefficient the angle of attack must be increased by approximately 0.8 degree, from 1.137 degrees to 1.927 degrees. A very significant effect is that with LFC off and constant lift a shock wave forms at 35 percent chord.

OF PGO_ _' ,_' _TY M = 0.80, A -- 30 DEG, C L -" 0.50 -O.OOO7 T 1 1 I Z_ CONSTCHORD=5J|2 m (19.1FT) , , 10 s _"_ _ 0 CONST ,-, - 0.0006 O' iNIT Re =11"86x v i- -0.0005 Z SUCTION TO 85% CHORD "_ _.

uJ BOTHSURFACES -0.0004 nl O -0.0003 Z ...........,.. C'_ -o.00o2 SUCTION TO 70% CHORD CQo< BOTH SURFACES m -0.0001 I0 20 30 40 50 60 70 80 CHORD REYNOLDS NUMBER x 10 -6 7-GEN.22654 -!

VARIATION OF LFC TOTAL SUCTION COEFFICIENT WITH REYNOLDS FIGURE 7-9.

NUMBER 30-DEG SWEPT WING Moo = 0.8 C L -" 0.502 0.005 I I UNIT Re CONST= 1.86 x 106 CHORD CONST = 5.82 m (19.15 FT) 0.004 PROFILE 0.003 DRAG COEFFICIENT m_ m __ ..- SUCTION T 85% CHORD (C d ) 0.002 CN UPPER SURFACE ONL O SUCTION T 70% CHORD 0.001 v 10 20 30 40 50 60 70 80 CHORD REYNOLDS NUMBER x 10 -6 7-GtrN- 22663-1 FIGURE 7-10. VARIATION OF PROFILE DRAG WITH REYNOLDS NUMBER OF POG'_ QC;I_,LiTY -1.5 SYMBOL COMMENT (:z C_ 1.137 0.783 WITH LFC -1.0 1.927 0.783 WITHOUT LFC 1.137 0.625 WITHOUT LFC

\

-0.5 Cp

\

0.0 =

\

=NORMAL- MACH NUMBER = 0.7 0.5 1.0 0 20 40 60 80 100 PERCENT CHORD FIGURE 7-11. EFFECT OF LOSS OF LFC ON AIRFOIL PRESSURE DISTRIBUTIONS -1.5 SYMBOL a C_ LFC SF 1.137 0.783 ON 0 -1.0 1.927 0.783 OFF 0 0.99 0.783 OFF 1.50 -0.5 C P

\

0.0

p' \

0.5 NORMAL MACH NUMBER = 0.7 1.0 0 20 40 60 80 PERCENT CHORD FIGURE7-12. EFFECT OF TRAILING EDGE FLAP ON AIRFOIL PRESSURE DISTR I BUTIONS 132 r_RECEDiNG PAGE. _LANt& NOT FILMED 7.1.6 Off Design Conditions Off-design operation of the representative LFC airfoil was investigated by i calculating the pressure distribution and suction requirements for lift coefficients above and below the design lift coefficient. The resulting upper surface pressures and suction velocities are presented in Figures 7-13 and 7-14. When the lift coefficient (angle of attack) is reduced:the pressure peak is also reduced with a similar but more favorable gradient using slightly reOuced suction. On the other hand, a small increase in lift coefficient raises the local Mach number, above the design value of 1.1, causing a shock to form near 40 percent chord with possible loss of laminar flow downstream of the shock.

Such velocities in the region of the shock indicate a substantial increase in required suction for the increased lift off-design condition. The local suction values are not considered to be quantitatively correct because there are no reliable methods for predicting suction requirements with a shock present on the airfoil.

7.1.7 LFC Airfoil Design Study Summary The results of the Airfoil Design study provide the following conclusions: 0 Existing analytical aerodynamic design techniques are readily applicable to the design of LFC compatible airfoils having shock-free supercritical flow with a turbulent boundary layer.

LFC compatible airfoils are significantly thinner than comparable supercritical airfoils.

Extending LFC beyond 70 percent chord to 85 percent reduces profile (wake) drag by 50 percent while the required suction flow i s doubled.

Trailing edge, small chord trim flaps can provide a ready means of adjusting and maintaining section lift if LFC is lost.

Operation of the LFC compatible airfoil at off design conditions does not cause any problem at reduced lift coefficients.

However, at higher than design lift conditions shock waves appear with a consequent loss of LFC aft of the shock.

P_.,G;: ISI _ ORIGINAL '_ _- OF POOR QUALITY Moo = 0.8, A= 30 DEG DESIGN C L = 0.502, I 1 C L = 0.438 (NORMAL M =0.7) -- C L= 0.502 = 11 C L = 0.556 -1.2 / -1.0 _ .0.8 ( PRESSURE COEFFICIENT -0.6 ' _'X (Cp) -0.4 -0.20 0.2 20 40 60 80 100 PERCENT CHORD 7-GEN-22656 FIGURE 7-13. CHANGE OF AIRFOIL PRESSURE DISTRIBUTION WITH LIFT COEFFICIENT CHORD REYNOLDS NUMBER - 35.6 x 10 6 mls (FT/SEC') --0.6 (--2.0) UPP'ER SURFACE

m]

Me=- 0.8 A=30 °

I

--0.5 C L "-- 0.438 (--z.6) CL =0.502 C L =0.556 i >: -0.4 I- i (--z2) O II -0.3 II I.M II (-0.8) Z _o -0.2 i- (-0.4)/_ ¢n -0.1 D

\

(o)

0 lO

20 30 40 50 60 70 80 9O 7-GEN-22.675- ] ARC LENGTH (PERCENT CHORD) FIGURE 7-14. EFFECT OF CHANGING LIFT COEFFICIENT ON LFC SUCTION VELOCITY DISTRIBUTION 7.2 NUMBER AND LOCATION OF PROPULSION AND SUCTION ENGINES An investigation was conducted to guide propulsion system configuration development of a baseline LFC aircraft, lhree principal items were considered L in this investigation.

Effect of engine acoustic environment on laminar flow.

Number and location of propulsion engines.

Number and location of suction engines.

7.2.1 Effect of Engine Noise on Allowable Engine Location The first phase of the investigation focused on the influence of engine noise on the aerodynamic surfaces where it would be desirable to apply LFC . Near field noise for three candidate engine cycles were estimated. Since difference in overall sound pressure levels (OASPL) were not large enough to affect engine location, only one engine cycle, the Energy Efficient Engine (E 3) type, was selected as the reference for acoustic characteristics and installed engine performance. Allowable acoustic disturbance criteria were developed from X-21A data and alternative engine locations were evaluated. An acoustic map showing contours of OASPL in terms of dB relative to .02mPa (.0002 dyne/cm 2) for the E3 engine is presented in Figure 7-15.

The interaction of discrete noise frequencies with resonance conditions within the boundary layer should be considered in determining the location of noise induced transition. Unfortunately, detail frequency noise levels are not known for such advanced engines and analytical methods are as yet inadequate to solve the problem, thus a method based on overall sound pressure level was used to assess the possibilities of achieving laminar flow relative to powerpl ant location.

An estimate of the allowable acoustic environment was made for a standard day flight condition of 0.8 Mach number at 10,670 m (35,000 feet) altitude. This estimate was based upon X-21A criteria presented in Reference 7.4-5. These criteria were determined in terms of the equivalent sound pressure level, as a function of chord Reynolds number as shown in Figure 7-16. The equivalent sound pressure level is given by the relation: / Where: _ _ = the ratio of the root mean square disturbance to the freestream velocity.

the freestream Mach number.

Moo = the ratio of specific heats the freestream pressure Pa

P:=

Included in Figure 7-16 is a curve showing the X-21A criteria increased by 6dB which was suggested as a result of X-21A flight test experience.

It was also determined during this investigation that at a fixed Mach number, the variation of Reynolds number and ambient pressure at flight levels from 9,140 m to 12,190 m (30,000 to 4L),UO0 feet) is such that the allowable noise level at a given distance from the leading edge is essentially independent of altitude.

The regions affected by engine acoustic environment were estimated for both wing mounted and aft-fuselage mounted engine configurations. By super-position of the engine acoustic field and the allowable noise levels on the aircraft wing planform, an assessment of the extent of detrimental engine noise effects was made.

In the case of the wing mounted engines most of the wing is subjected to an acoustic environment which exceeds the allowable sound pressure levels. The affected area is indicated by the shaded region in Figure 7-17. Since the region shown does not assume any benefit due to shielding of the upper surface by the wing itself, it is probable that the amount of laminar flow achievable would be larger. However, a 10 dB reduction in engine sound pressure level would not increase the laminar flow area appreciably.

;[ ORIGINAL PAGE |_ OF POOR QUALITY Mo_ -0.8 ALT = 10,670 m (35,000 FT)

/

r - - (--60) (--40) (--20_ ( 0 ) (20) (40) (60) (80) (IO0,(FT) I 1 I i J 1 ] I I I J --20 --15 --10 --5 0 5 10 15 20 25 30 m 7-GEN-22790_l DISTANCE FROM ENGINE INLET FIGURE 7-15. NEAR-FIELD ACOUSTIC ENVIRONMENT DUE TO ONE ENGINE Mo_- 0.8 ALT = 10,670 m (35,000 FT) %%%%%%% ,% ALLOWABLE SPL (dB)

N

- _ N%N%%%%% I I I I I I I (40)(50) (FT) (3 (4) (5) (6)(7)(8)(9)(10) (20) (30) I I I I I I I I I I I 15 m 2 3 4 5 678910 1.0 7-GEN-22780" 1 DISTANCE FROM LEADING EDGE FIGURE 7-16. ALLOWABLE ACOUSTIC ENVIRONMENT In contrast to the preceding case with wing mounted engines, aft-fuselage mounted engines were found to be compatible with achieving laminar flow on the wing. The areas affected by the engine acoustic environment are exclusively on the tail surfaces as shown in Figure 7-18. Engine induced sound pressure levels for this case were increased 3dB to account for the dual-engine pod concept shown. Most of the vertical tail is not amenable to laminarization so application of LFC on the vertical tail was not recommended. Laminar flow could be established over a significant portion of the horizontal stabilizer.

However, unlike the wing which operates over a limited range of lift coefficients in cruise, the horizontal tail lift may vary from positive to negative as center-of-gravity location varies. The complication of a suction system for the horizontal tail, considered along with the limited amount of laminar flow which may be obtained and the inaccessibility of the tail for inspection and cleaning, approx. 15.2 m (50-feet) above ground, resulted in the recommendation to forego LFC on the horizontal tail.

From the investigation of engine acoustic environment effects on laminar flow_ it is evident that the engines on an LFC aircraft should be located on the aft fuselage. It is also indicated that tail surfaces may not effectively utilize LFC due to effects of the engine acoustic field.

7.2.2 Number of Propulsion Engines A study was conducted to evaluate the performance and economic tradeoffs between three and four engine configurations for the baseline LFC aircraft.

The aft-fuselage location for propulsion engines was dictated by the need to minimize exposure of the LFC wing to the detrimental engine acoustic environment, as established in Section 7.2.1.

The procedure used to select the number of propulsion engines consisted of sizing both the three engine and four engine LFC aircraft configurations to minimize the takeoff gross weight. This sizing was based upon the design mission and ground rules defined in Sections 4.0 and 5.0. After sizing each ORiC!I.. .. ".....

-- ENGINE NOISE ALLOWABLE NOISE (6 dB ABOVE X-21A CRITERIA) SHADED AREA IS REGION WHERE ACOUSTIC ENVIRONMENT MAY CAUSE TRANSITION

J .ll

m _ _ 7-C, EN-22793 FIGURE 7-17.

EFFECT OF ENGINE-INDUCED ENVIRONMENT ON EXTENT OF LFC (WING- MOUNTED ENGINES)

/-/7 _,. ,.,._

///

-- ,_ow,.__o.s_ / /77- _

(6 dB ABOVE X-21A CRITERIA) / //___ /_7 J ACOUSTIC ENVIRONMENT MAY /[ ///i t_.___._..-J--/ CAUSE TRANSITION _I__ / 7-GEN-22665 FIGURE 7-18.

EFFECT OF ENGINE-INDUCED ACOUSTIC ENVIRONMENT-ON EXTENT OF LFC (FUSELAGE-MOUNTED ENGINES) configuration, direct operating costs were determined. The configuration with three engines which had the lower operating cost was selected for subsequent LFC aircraft. See Section 5.4 for details.

7.2.3 Number and Location of Suction Enqines Investigation of suction system characteristics was conducted to develop overall design, control, and operational requirements. Details of this study and analysis work are described in Section 10.6. A summary of results and conclusions is given in Section 5.5.

The results showed that one suction engine per side, located outboard of the wing planform break was the most suitable arrangement. This configuration is satisfactory for suction on both wing surfaces as well as for suction on the upper surface only.

7.3 ASPECT RATIO, THRUST, AND WING LOADING ANALYSIS Analysis and configuration development tasks were carried out to evaluate performance parameters and select the aspect ratio, engine size, and wing area for LFC aircraft utilizing two concepts of suction distribution on the wing.

The basic concept considered suction on both upper and lower wing surfaces to 70 percent chord, while the alternative concept used suction on the upper surface only back to 05 percent chord. Rationale for this alternative concept is based on the comparison of profile drag coefficients together with significant structural and operational advantages of using suction on only the upper surface of the wing. See Section 5.6.1.

The general procedure used in this analysis consisted of: (i) establishing aerodynamic and structural weight characteristics in parametric form as functions of aspect ratio, wing area, and engine size, (2) calculation of takeoff gross weight (i.e., aircraft size) required for the design mission, and (3) determining operational limits on takeoff field length, initial cruise altitude, and approach speed. Finally, the direct operating cost was estimated for the minimum takeoff gross weight aircraft.

7.3.1 Wing Configuration Selection of the wing planform for this study was based on concurrent transport design work and wind tunnel test data from the Energy Efficient Transport program. See also Section 5.6.1.1. Wing planforms for aspect ratios of 10, 12, and 14 are shown in Figure 7-19. In each case quarter-chord sweep is 30 degrees and taper ratio is 0.25 for the basic trapezoidal wing. A trailing edge extension, with a trailing edge sweep-back of 8 degrees, is located between the side of the fuselage and 40 percent semi-span station.

Design lift coefficient is O.5O for each aspect ratio and the average thickness is 11.7 percent chord. The thickness-chord ratio distribution is tabulated below: Semispan Station Thickness-Chord Ratio Percent Percent Side of Fuselage 13.82 32.6 11.87 40.0 10.92 80.0 10.92 lO0.O 10.92 Twist distributions for the 1.0g and rigid reference conditions are shown in Figure 7-20.

The general configuration used in this study was the three engine airplane selected as under Section 7.2.2. Fuselage size was fixed as previously, to be compatible with the 30U passenger, 31,298 Kg (69,000 Ib) payload, requirement for the 9260 Km (bOUO n.mi.) design mission. Tail areas were initially sized for the nominal wing area. Thereafter the tail volume coefficients were held constant as wing area was varied.

High lift systems were adapted to accommodate each of the two LFC concepts.

For the basic concept a 25 percent chord, two element, trailing edge flap was used. A sketch of the flap is shown in Figure 7-21. Because porous suction surfaces extend below the leading edge on the basic LFC wing, leading edge high lift devices were not feasible with suction on both surfaces. The high lift system at the trailing edge for the alternative LFC wing is severe]y constrained due to the extension of suction to 85 percent chord. Thus the flap was limited to a single slotted configuration having 15 percent chord.

However, with LFC on the upper surface only a leading edge device can be used which compensates for the smaller trailing edge flap. The leading edge device also acts as a shield to alleviate insect contamination. The latter high lift system is shown in Figure 7-22.

Weight characteristics for the analysis considered strength, stiffness, and flutter criteria. For the LFC wings it was found that mainly bending stiffness requirements for roll control established the wing structure/weight.

Methods and results of the strength and stiftness analyses are presented in Section 9.1.3. Aeroelastic analyses of the LFC wings were conducted to evaluate the elastic-wing rolling moment due to aileron deflection at varying wing stiffness values. This was done at maximum level flight cruise speed, where it was required that the elastic-wing roll capability be at least 25 percent of the rigid-wing rolling effectiveness in order to assure the desirable level of roll control at high speed. Figure 7-23 illustrates the variation of elastic-wing roll capability with increasing wing bending stiffness for aspect ratio 12. Increased torsional stiffness was also investigated, however, with a 30 degree swept wing the dominant aeroelastic rolling effects are due to wing bending. The resulting wing bending S w = 334.5 m 2 (3600 SQ FT),AC/4= 30 DEG = AR = 12 AR -14 / # I / I I / I / t I / I /

,£

8-G EN-22640-1 FIGURE 7-19. ASPECT RATIO SIZING WING PLANFORMS 7.0 __ _ 14 6.0 5.0 '_"'"_.,,_-RIGID 4.0 i ZL L (DEG) 3.0 1D II ALLASPECTRATIOS -''_'_ 2.0

\

1.0 0 10 20 30 40 50 60 70 80 90 100 PERCENT SEMISPAN 8J -GEN 22365 FIGURE 7-20.

EFFECT OF ASPECT RATIO ON ANGLE OF ZERO LIFT ORi_'IAL PAGE 1| OF POOR Q!._ALll"hf ___.___ FLAPSTOWED FLAP DEFLECTED "'_,,_ 8 F =35/15 DEG _'\ FIGURE 7-21. HIGH-LIFT SYSTEM - LFC BASE CASE - SUCTION BOTH SURFACES TO 70 PERCENT CHORD t.

SUCTION TO 85%c--_ v _l SINGLE-SLOT TRAILING EDGE FLAP ""v% LEADING EDGE FLAP AND SHIELD _F = 30 DEG CF/C = 0.10 CF/C =0.15 B-GEN-22554A-) _" FIGURE 7-22. HIGH-LIFT SYSTEM - UPPER SURFACE SUCTION ONLY

stiffness, and corresponding wing weight factors, are shown as functions of

wing aspect ratio in Figure 7-24. Although the weight factor does not

increase as rapidly as the bending stiffness factor the weight penalty

incurred with higher aspect ratios is clearly evident.

7.3.2 Aircraft Sizing - Base Case Sizing matrices were constructed for each aspect ratio to graphically show the interrelation between wing area, engine thrust, and takeoff gross weight for the design mission. Mission constraints were then superimposed on the matrix to determine the configuration size parameters for the minimum takeoff gross weight aircraft which meets mission requirements and operational constraints.

Figure 7-25 is an example of a sizing matrix for an earlier LFC configuration. Initial cruise altitude limits of 9449 m (31,000 ft) and 10,668 m (35,0t_0 tt) are shown along with the 66.9 m/s (130 KEAS) approach speea cutoff. In this instance the 3,(]48 m (10,000 ft.) takeoff distance limit did not appear within the limits of the matrix. The design points corresponding to the two initial cruise altitude limits are indicated by the symbols at the intersection with the approach speed limit. Two initial cruise altitudes were considered in order to evaluate the penalty for selecting the higher altitude 10,670 m (35,0(J0 tt) where the likelihood of encountering ice crystals is greatly reduced.

Aircraft configuration and performance parameters are summarized, as functions of aspect ratio, in Figures 7-26a & b for the LFC aircraft with suction on both surfaces corresponding with Figure 7-25. The higher initial cruise altitude requires a slightly larger wing area, larger engines and higher takeoff gross weight. However, fuel burned and takeoff field length are reduced with the larger wing.

Based upon the analytical results and practical operational factors the baseline aircraft was sized to meet the 10,670 m (35,000 ft) initial cruise altitude and have an aspect ratio of 10. Low speed lift curves and L/D characteristics were shown previously in Figures 5-30 and 5-31, and discussed in Section 5.0.1.1.c.

OF PO£:_ _:_;;;LI-i'Y' AR = 12 M = 0.85 h - 9,144m (30,000 FT) 1.0 E/R C_Sa O.5 REQUIRED I I I I I I I 1 I I I I 1.5 EISTI FFNESS 1.0 2.0 2.5 EISTRENGTH 8-G EN-22642-] FIGURE7-23. ELASTIC WING BENDING STIFFNESS CRITERIA 3.0 2.5 RELATIVE WING STIFFNESS 2.0 AND WEIGHT 1.5 WING WTSTIF FNES S WING WI'STRENG _ ==_.==.

1.0 9 10 11 12 13 14 ASPECT RATIO 8-GEN-22620 FIGURE 7-24. ELASTIC WING BENDING STI FFNESS AND WEIGHT FACTORS ORIGi:,';.:L_- |-",_\C: _. [G OF POOR QUALITY

ii" , *

(1OOO LB) 1000 kg (420) 190 - iODESIGN POINTS (410) 185 - \ TAKEOFF n _ ENGINE SIZE-FNTO (4O0) WING AREA kn (1000 LB) GROSS 180 - _'_.,_ _ m2 (SO FT) 138 (31} WEIGHT 111_(26}"'_"_'_.,_,_600) 334 _._ /-- _ i • _,_ 120 (27).,_._-_.., _ _(30) 133_ / (390) C "_---_' "\ _125 (28) .-.--'_129 (29) -- _ _ APPROACH SPEED_"'_k"_ _ _ ",,,,,_(34uu) - _ --_j,<..._ -- _ (35,000 FT) (380) INITIAL CRUISE ALTITUDE /_, ,450m (31,000 FT) 170 - (370) 8_GEN-22631 • ] FIGURE 7-25. MISSION SIZING MATRIX FOR AR=10 -- SUCTION ON BOTH SURFACES 1 47 ORIGINAL PAGE iTS.

OF POOR QUALITY jf • INITIALCRUISE ALTITUDENOTED • RANGE=9260 km (5000 N MI) • PAYLOAD=31,3(X)kg (69,0(X) LB) • APPROACH SPEED = 67 m/s ( 130 KEAS) kN 1000 LB km (1000 FT) 3.4 17e _ i _''_ I('m) ENGINE 35.000(10,668m) I (3!000 Fl'! (11) TAKEOFF 3.0 SIZE FNTO

169 ] I (38)

110) FIELD 160 __ (36) LENGTH 2.7 ...9i,_o_ . _,is,ooo F'r_ _ (91 m2 (100 FT2) 2.4 ($) 350 ,_ 138) 330 ,_ 1361 t/c WING AREA (%) 310 7 _ _,_ • (34) (CL=O.5) 290 (32) 10 11 12 13 14 11 12 13 14 ASPECTRATIO ._EN_263_1 ASPECTRATIO FIGURE 7-26A. ASPECT RATIO SIZING SUMMARY I --SUCTION ON BOTH SURFACES

F

INITIALCRUISE ALTITUDE NOTED • RANGE = 9260km (5000NMI) 1000 Iq (1000 LB) • PAYLOAD= 31,300 kg (69,000 LB) 47 , , (105) • APPROACHSPEED = 67 m,,,_(130 KEAS) FUEL 45 '_"=50" (il 00° it) --_ (100) (1000 LB) 1oookg BURNED (430)

! Tor=o5 i

43 1951 (*,2O) ¢/ASkm (¢/ASM) 186 (410) TAKEOFF 1.19 I 12.21 GROSS 181 (40O) 1.13 (lo,r_m) / (2.1) _v DIRECT WEIGHT OPERATING (390) (31,000 FT) L...-..

COST (38O) 1D3 (9,_9m) (19) 037 I 11_1 168 (370) 14 10 11 12 13 14 10 11 12 13 ASPIECTRATIO ,-_EN-.63_-_ ASPECT RATIO FIGURE 7-26B. ASPECT RATIO SIZING SUMMARY II - SUCTION ON BOTH SURFACES .;.

7.3.3 Aircraft Sizing - Upper Surface Suction Only The aspect ratio and sizing study for 1_he alternative concept, utilizing upper surface suction only IUSSO), benefited from the trends established in the preceding base case. Sizing matrices were constructed for aspect ratios 10 and 12 and only the 35,UUO ft initial cruise altitude condition was evaluated. The sizing matrix for aspect ratio 10 is presented in Figure 7-27. In this case the initial cruise altitude became the critical sizing condition.

Performance parameters for the USSO configuration for the two aspect ratio points evaluated are shown in Figures 7-28a & b. These points are superimposed on the summary plots for the preceding basic case. Considering these results and the preceding rationale, an aspect ratio of 10 was also selected for the USSO aircraft. See also Section 5.6.1.1.a. The configuration and performance parameters for the alternative LFC configurations at this stage are compared in Table 7-i. Low speed aerodynamic characteristics, lift curves and L/D ratio, were shown previously in Figures 5-28 and 5-29, respectively.

From the results of the aspect ratio, thrust, wing loading analysis it was recommended that the LFC concept utilizing suction to 85 percent chord on the upper wing surface only, be used for further LFC aircraft development. It should be emphasized that the comparisons were biased in favor of the initial LFC concept because no allowance was made for lower acquisition and maintenance costs that should be credited to the upper surface suction only case.

Later comparisons of updated alternative LFC configurations were presented in Section 5.6.1.5, see Tables 5-20 and 5-21. These showed even more advantages for the USSO configuration.

UPPER SURFACE SUCTION ONLY f 185 ' (410)i APPROACH SPEED = 130 KEAS (67 MPS) .-_ \ p- (400)1-- ENGINE SIZE FNTO kN (1000 LB) _,,.,,_1 U,I

,,, 1,21

WING AREA .o.__ ,,_,_/ .,_,_/ _ _,/"" ' (390)1-- "_ m z (SO FT) _._. \'_"Y ._7.,/ .\%,o_,f ,_ n,, __IAL CRUISE ALTITUDE (380) 1- L_ \ 10,670m (35,000 FT) _"--TAKEOFF RUNWAY LENGTH 3,050m (10,000 FT) (370)!- 165 - 1000 kg (360) (1OOO LB) • DESIGN POINT FIGURE 7-27. MISSION SIZING MATRIX FOR AR=10 - UPPER SURFACE SUCTION ONLY \ \ F:_GT r3 ORIG_Nf_L OF POOR Q u,_:._l-fY * RANGE=926Okm (5000 N MI) -- SUCTION ON BOTH SURFACES • PAYLOAD=31298 kll (69_X) LB) • SUCTION ON UPPER SURFACE ONLY . APPROACH SPEED=67 raps (130 KEAS) • INITIAL CRUISE ALTITUDE = 10E:68 m (35D00 FT) 187 (42) 3.O (10) ENGINE SIZE FNTO 178'.

L,,.._.__ _ (40) (1000 LB) (9) TAKEOFF 2.7 (38) kN 169 FIELD LENGTH 2.4 _ 353 (S) (38) (36) 2.1 335 (7) j 1000 m (1000 FT)

J

(34) (100 FT 2 ) 14 WING 316 AREA tic (32) (%) 12 ,_ q m 2 279 (C L= 0-53)1_,, (30) 10 11 12 13 14 10 11 12 13 14 ASPECT RATIO ASPECT RATIO 8-G EN-22638- ;' FIGURE 7-28A. ASPECT RATIO SIZING SUMMARY COMPARISON I SUCTION ON BOTH SURFACES • SUCTION ON UPPER SURFACE ONLY • RANGE = 9260 iorn(5000 N MI) • INITIAL CRUISE ALTITUDE = 10,668 m (35,000 FT) (IO(X) LIB) • PAYLOAD= 31,298 kg (69,000 LB) 1000 kg (105) APPROACH SPEED=67 mps (130 KEAS) 471p 1000 kll (1000 LB) FUEL J _- (100) 195 (430) BURNED 45 - (95) 19o /" (42o) 43 c/kin (¢/ASM)

/

TAKEOFF 186 (410) 1.19 (2.2) GROSS / WEIGHT 181 (400) "_ DIRECT 1.13 (2.1) J OPERATING 177 / (390) COST 1.O8 (2.0) q ( 172 (380) 1.O3 (1.9) 168 (370) 0.97 (1JS) 10 11 12 13 14 10 11 12 13 14 ASPECT RATIO ASPECT RATIO 8Z-GEN.Z2471 *NO ALLOWANCE FOR LOWER INITIAL AND MAINTENANCECOSTSFOR UPPERSURFACEONLY FIGURE 7-28B.

ASPECT RATIO SIZING SUMMARY COMPARISON II Table 7-1 Configuration and Performance Parameter Summary Sized Ai rcraft RANGE - 9,260 Km (5,000 NMI), PAYLOAD 27,215 Kg (60,000 LB) Suction Sucti on Both Surfaces Upper Surface Only to 70% C To 85% C i0 i0 Aspect Ratio Initial Cruise Altitude m Ift) lO,b7U (35,000) I0,670a (35,000) Wing Area m2 (ft 2)

311 (3,34_) (3,1oo)

.503 .504 Cruise CL 11.7 10.80 (t/c )ave 90,115 (198,670) 87,965 (193,930) Operating Wt. Empty kg (Ib) Takeoff Gross Weight kg (Ib) 176,275 (38_,620) 174,905 (385,600) 63 (121.7) Ola (130) Approach Speed m/s (KEAS) Takeoff Field m (ft) 2,667 (8,750) 2,466 (8,U90) 13,608 (3U,O00)

13,5 9(29,980)

Thrust/Engine kg (Ib) Total Fuel Burned kg (Ib) 45,917 (101.230) 46,847 (103,280) 96_ (2,134) 1,054 (2,323) Suction Engine Fuel kg (lb) D.O.C. ¢/ASkm (NMI)b 1,065 (1,973) 1,059c( 1.962 ) Critical sizing factor Fuel 45¢/Gai 1 on Reduction of maintenance cost with one surface only not taken into account.

7.4 REFERENCES - AERODYNAMICS 7.4-1 Tranen, T.L.: Analysis and Design of Transonic Airfoils.

Report No. MDC A3760, December 1975.

7.4-2 Cebeci, T.; Kaups, K.; Ramsey, J.; and Schimke, S.: A Two Point Finite Difference Boundary-Layer Method for Incompressible and Compressible Two-Dimensional and Axisymmetric Laminar and Turbulent Flows Incluaing Infinite-Swept Wing. Douglas Report MDC J7986, September 1978.

7.4-3 Cebeci, T.: Calculation of Three-Dimensional Boundary Layers.

I. Swept Infinite Cylinders Second Small Cross Flow. AIAA Journal, Vol. 12, p. 779.

7.4-4 Nenni, J.P. and Gluyas, G.L.: Aerodynamic Design and Analysis of an LFC Surface. Astronautics and Aeronatics, July 1966, pp. 52-57.

f 7.4-5 Eichelkrout, W.L.: LFC Aircraft Design Data: Section A-14 Ice Protection Systems. Northrop Report NOR-67-136, April 1967.

/ .

!

C., 8.0 AERODYNAMIC TESIING Aerodynamic tests conducted during the course of this contract consisted of two substantive test programs which are summarized in this section.

The first series of tests were conducted to evaluate the relative aerodynamic smoothness, under conditions of applied suction, of candidate porous LFC surface materials. A second model test program was carried out to demonstrate that suction through a porous surface can sustain laminar boundary layer flow in a representative swept-wing flow situation, where cross-flow instability is a dominant cause of transition. Both test programs were conducted in the Douglas Low Speed Wind Tunnel.

8.1 AERODYNAMIC SMOOTHNESS TESTING A test program was established to evaluate a variety of candidate surfaces with respect to suitability for application to laminar flow control. Tests were conducted to determine how inherent roughness (surface textures) and porosity characteristics affect boundary layer transition and the ability to maintain laminar flow.

6.1.1 Model Description and Installation The model consisted of a two-dimensional panel incorporating a removable frame for mounting the porous test specimens. The panel had a total chord of 3050 mm (120 inches) and a thickness of 57 mm (2.Z5 inches) as shown in the installation diagram and photo, Figures 8-i and 8-2.

The flat panel was constructed using stock-size square tubing to frame honeycomb areas covered with sheet aluminum. The leading edge section was a 10.2 degree wedge shape. The trailing edge incorporated a full span flap sect ion.

PRECEDING PAGE E_LANK NOT FILMED ,4 '

ORiGinAL PACE T_

OF POORQU,,:_.LtTY

METER suCTION Ji P HINGE LINE I h I - _ DIFFUSER 1190 mm 1370 mm !(47 IN.)

POROUS (54 IN.)

SPECIMEN

___L____ __L_

280 mm PLAN __F_____ (11 IN.)--/_ 965 mm J (38 IN.) I -+5 /-PIVOT _/--- FLAP ____ _ DIFFUSER (2.25 IN.) __.10 o --TEST SECTION 3050 mm--___ (120 IN.) _.o,_-_ooo_,, SPECIMENS FIGURE 8-'1. WIND TUNNEL MODEL TEST FOR LAMINAR FLOW WITH POROUS SURFACE FIGURE 8-2. WIND TUNNEL MODEL INSTALLATION The radiused nose shape and tangential flat surfaces of the flap formed a trailing edge angle of 11.5 degrees with a trailing edge thickness of 3.18 mm (.125 in). The flap hingeline was located at 90 percent of the panel chord and the flap was manually adjustable to + 10 degrees deflection.

The panel could be rotated from the horizontal plane, pivoting at 50 percent of the chord length and was manually adjustable to + 5 degrees of incidence to the tunnel stream. A seal was provided between the ends of the panel and the tunnel walls to prevent leakage and to allow rotational motion of the panel.

8.1.2 Test Specimens Various 279 x 432 mm (11 x 17 in) effective suction area specimens of porous material were installed in the top surface of the panel between 20 and 30 percent chord. A plenum under the specimen allowed suction through the specimen for removal of air from the boundary layer. A removable cover plate in the bottom surface of the panel provided access for installation and adjustment of interchangeable specimens. The plenum was manifolded to provide an even distribution of suction flow.

Listed below are the surface panel specimens that were tested. Other materials were considered but only those tested are included here. Detail descriptions of the various surface materials and substructures are given in Sections 9.2.1 and 9.2.2.

1. Solid aluminum flat plate - reference surface for transition location without suction 2. Metallized Doweave 3. Sintered fiber metal on Doweave .

P 4. Micro perforated plate (MPP) on Doweave Lockcore 5. Dowea ve 6. Porous Strips 7. Slotted aluminum 8. Pe forated titanium 9. 50 x 2bO Dynapore on Isogrid 10. 50 x 250 Dynapore on Honeycomb 11. 80 x 7UO Dynapore on Honeycomb 12. 80 x 700 Dynapore (closed) on Honeycomb 13. _0 x 700 Dynapore (open) on Honeycomb 14. 80 x 700 Dynapore (modified) on Honeycomb 8.1.3 Testing Procedure Due to tne large blockage of this model in the tunnel it was necessary to determine the reference (actual) dynamic pressure in terms of the nominal standard tunnel reference pressures. This was done by relating the nominal tunnel dynamic pressure to that measured on a pitot static tube mounted on the tunnel center line at 30 percent chord and midway between the panel surface and tunnel ceiling. The resulting calibration showed that the actual tunnel dynamic pressure, with the model installed, was approximately 87.5 percent of the nondnal value for the open tunnel.

Chordwise pressures were measured along the centerline of the model using a length of strip-a-tube attached to the upper surface of the model. The angle of attackand trailing edge flap deflection were adjusted until the stagnation r point was on the upper surface and the pressure gradient was slightly favorable-to-neutral over the forward 80 percent of the chord. As indicated in Figure 8-3, a small negative incidence and flap deflection were required.

©

The solid flat plate was tested first to establish the transition location as a function of chord Reynolds number. This value was then used as a reference to compare the transition location for each of the various porous panels. The reference transition location is shown in Figure 8-4. Testing was conducted with one and three screens installed in the tunnel settling chamber in order to evaluate the effect of the tunnel turbulence level. With three screens in place, the maximum tunnel dynamic pressure obtainable was limited to 1.436 kP (30 lb/ft2). Most of the testing was therefore done with only a single screen instal led.

Transition location was identified by means of a hot-film sensor probe giving a signal which was displayed on an oscilloscope and projected audibly from a speaker. Typical visual display results are shown in Figure 8-5. Transition was identified by judging, insofar as possible, a 50-50 distribution of laminar and turbulent signals in the oscillograph trace. Measurements based upon the visual signal were more consistent than those using only an audio reference. In several instances, Tollmein-Schlichting waves were detected.

The resulting signal exhibited a regularity as indicated in Figure 8-5(B) and a distinct high pitched humming sound could be heard on the audio output.

This phenomenon is related to the most amplified frequency in the initial transition process, which persisted over a sufficient region to be identified with the hot film sensor.

Each porous specimen was tested by varying the suction flow rate from zero to a maximum corresponding to a flow coefficient, CQ, value of approximately 0.005. Transition location was determined for several values of CQ at Reynolds numbers varying from approximately 5.0 x 106 to 11 x 106 .

Reynolds number variation was accomplished by changing tunnel speed (dynamic pressure). Transition location was considered to be the aft-most point where the boundary layer was observed to change from laminar to turbulent flow.

Generally_ transition occurred along a slightly irregular spanwise line which was the result of many turbulent wedges finally merging.

ORIGINAL Pi:,::._::: [$ OF POOR _UALITY

f

\, RUN o .._r_ _ o 16 -OA --9.6 0.16 o 19 - 0.4 --11D 22 - 03 --11D 0 25 - 0.2 --11D PRESSURE COEFFICIENT 0.08 Cp

o

I I -0.08 _ J L 1.0 0.8 0 0.2 0.4 0.1 CHORDWISE LOCATION, X/C 81 _;E N-22394 FIGURE 8-3. WIND TUNNEL MODEL - SOLID SURFACE PRESSURE DISTRIBUTION 0.9 0.8 0.7 0.6 0 3 SCREENS TRANSITION 0.5 A 1 SCREEN LOCATION X/C 0.4 ,..% 0.3 0.2 0.1 0 2 4 6 8 10 12 CHORD REYNOLDS NUMBER (10 e) Bl .GEN-22:lg0 FIGURE 8-4. TRANSITION LOCATION ON WIND TUNNEL MODEL - SOLID PLATE SURFACE

_--- 8.1.4 Test Results

Typical results of transition surveys are shown in Figure 8-6 which gives transition location (XTR/C) versus suction coefficient (CQ) at different chord Reynolds numbers for one surface. Solid symbols denote the reference transition location on the smooth solid surface at the corresponding chord Reynolds number. It should be noted that CQ is presented on a logarithmic scale. The zero suction and reference transition locations are plotted along the vertical axis.

Comparison of the zero suction and reference transition locations indicates the combined effect of the porous surface texture (roughness) and any inflow/outflow that might be present. Since the pressure gradient is essentially zero in the region of the porous panel, the principal cause in this decrement is assumed to be the surface roughness. The condition of aerodynamic smoothness is defined as occurring when the transition location, with suction applied, is downstream of the reference transition location at the same Reynolds number.

Extension of transition farther downstream as suction is increased provides an indication of the effectiveness of the suction and the amount of suction flow required to maintain laminar flow in the boundary layer. A brief review of results is given below.

Metallized Doweave - The Doweave Specimen did not achieve aerodynamic smoothness at any level of applied suction. It was considered unsatisfactory.

Sintered Fibermetal on Doweave - This specimen achieved some extension of laminar flow with moderate suction (CQ > .0001). However, due to mediocre aerodynamic performance combined with difficult structural features, this material was not considered further.

r Microperforated Plate IMPP) on Lockcore - The specimen had a wavy surface originating in the Lockcore truss pattern of the substructure. The results

ORIGINAL PA,:_ _S

OF POOR QUALITY" (A) LAMINAR (B) TOLLMIEN-SCHLICHTING (C) TRANSITION (D) TURBULENT ............

FIGURE 8-5. BOUNDARY-LAYER SIGNATURES - PROBE SIGNAL DISPLAYED ON OSCI LLOSCOPE DYNAPORE HONEYCOMB SANDWICH SOLID SYMBOLS DENOTE SOLID SURFACE TRANSITION LOCATION 0.9 0.8 C 0.7 TRANSITION LOCATION 0.6 X /C TR 0.5 CHORD REYNOLDS NO. Rc 0.4 Q 5.2 7.4 0.3 [] 10.0 11.7 0.2 0 0.0001 0.001 0.01 SUCTION COEFFICIENT C Q FIGURE 8-6, TYPICAL WIND TUNNEL RESULTS - TRANSITION LOCATION with suction applied indicated that a satisfactory level of laminar flow could be attained, however, significantly higher suction flow was necessary compared with other specimens tested.

Doweave This specimen was similar to the metallized Doweave and was unsatisfactory. Although it achieved aerodynamic smoothness at moderate suction values it was subject to complete loss of laminar flow with higher suction flow.

Porous Strips - The porous strip specimen consisted of a Dynapore surface over honeycomb sandwich with nonporous adhesive bonding the Dynapore to the honeycomb. Four porous strips 3.3 mm (0.13 in) wide, were located on the panel. These strips were separated by 72.9 (2.87 in) wide nonporous strips purposely blocked with adhesive. Test results for this specimen indicated that it was effectively smooth with moderate suction coefficients, although at higher suction flows the transition distance was reduced significantly. This was attributed to deflection of the panel surface due to the high pressure differential across the panel, rather than to over-suction. As configured %.¸¸¸¸ without adequate surface support, this specimen was considered unsatisfactory, however, subsequent tests of the swept wing model with a more substantial sublayer showed very favorable result with porous strips.

Slotted Aluminum - A slotted specimen was included for comparative purposes.

This panel had 0.127 mm (0.005 in) wide slots spaced 76.2 mm (3.0 in) apart.

Thus the slots were arranged at approximately the same spacing as the preceding porous strips. Performance of the slotted specimen was satisfactory. Transition was delayed to almost 0.8 chord at moderate suction coefficient (CQ = 0.0017). However, this specimen was sensitive to over suction flows, which caused transition.

Perforate(tTitanium - Perforated titanium specimens were constructed from r .b35mm (0.025 in) titanium sheet which had been perforated using an electron beam technique. Holes were nominally .102 mm (0.004 in) diameter and were spaced 1.02 mm (0.04 in) apart in an equilateral triangular pattern. Ribs were provided to support the panel within the frame and minimize distortion of [ the surface.

,p Aerodynamic performance of the perforated titanium was satisfactory.

Transition was moved farther aft with increasing suction to the maximum suction applied. It was noted in several instances that small foreign particles were caught and held on the surface by the suction through the perforation. The turbulent wedge resulting from the particle disturbance persisted downstream. This phenomenon was alleviated on subsequent testing of the perforated titanium leading edge on the swept wing model by having the perforations elongated to approximately twice the nominal hole diameter. The elongated hole was presumed to be less likely to entrap passing particles in the ai rstream.

Later development in electron beam perforation technology has produced high quality perforated titanium with a nominal hole diameter of .0635 mm (0.0025 in). This advance in the state-of-the-art appears to have eliminated the contamination problem.

Dynapore - Several Dynapore specimens were tested. The coarser textured 50 x 250 Dynapore surface material was bonded to Isogrid and honeycomb supporting substructures while the 80 x 700 Dynapore material was tested on honeycomb substructure only. Both 50 x 250 Dynapore specimens performed satisfactorily. The 80 x 700 Dynapore specimen with honeycomb substructure had quite irregular porosity due to excessive bonding adhesive. This condition very likely contributed to the susceptibility of the panel to oversuction.

Testing of Dynapore specimens was concentrated subsequently on those having the 80 x 7UO surface material. The finer weave of this material provided a smoother surface. However, the finer weave also results in lower strength and stiffness of the porous surface. In order to stiffen the 80 x 700 Dynapore surface, a sublayer of 80 x 80 Dynapore was fusion bonded to the basic surface. _'Later, since the surface was still more open than desired and to further increase surface strength and stiffness, a perforated fiberglass sublayer (40 percent open) was added beneath the Dynapore layers. The perforated fiberglass sublayer was divided into three spanwise segments with each segment having different diameter holes 3.18, 4.76, and 6.35 mm, (i/8, 3/16, and 1/4 in), respectively. This was done to determine whether the size of the holes in the sublayer would have any effect upon laminarization of the boundary layer flow. The surface performed satisfactorily and no effect of hole size in the sublayer was detected.

A brief evaluation of the effects of surface damage was made on the last Dynapore specimen tested. Several severe depressions were made in the surface using a spherical tool. in each case a turbulent wedge occurred behind the depression. Laminar flow was restored by simply filling the depressions with tunnel wax and smoothing the surface.

It was determined from tests of these specimens that: (1) Dyanpore is a satisfactory porous surface material for laminar flow control, and (2) within the range of hole diameters tested, there was no observable effect of sublayer hole diameter on the laminar flow or transition location.

A summary of these results is presented in Figure _-7. Comparative effectiveness for several candidate LFC surfaces is shown for a chord Reynolds number of 8.8 x 106 . As a result of these tests 7it was concluded that only the 60 x 700 Dynapore and the perforated titanium should be considered further as practical surface materials for laminar flow control using distributed suction through a porous surface. At this point in the program it appeared that Dynapore offered the most promise of successful LFC application because of the limit on the smallest hole diameter available from the electron beam perforation process. Subsequent improvements in the perforation process and the results of subsequent swept-wing model test described in Section _.2 following, have changed this position significantly and perforated titanium is now the preferred LFC surface.

_.2 SWEPT WINGWINDTUNNEL TESTS

The second test program conducted as part of the LFC contract involved testing

a swept-wing model in the Douglas Long Beach Wind Tunnel facility. This test program was a cooperative effort in which Douglas IRAD resources provided for design and construction of the basic models and NASA contract funds supported testing and data analysis.

The swept-wing model test was directed toward the primary objective of demonstrating the ability to sustain laminar flow, using suction through a porous surface, under representative full-scale swept-wing flight conditions including practical treatment of surface anomolies such as panel joints. In particular, it was important to demonstrate laminar flow control in the presence of instabilities which are prevalent on a swept wing; namely, cross-flow instability and those resulting from the flow along the attachment line. The basic test provided the demonstration desired first with an all Dynapore LFC surface and then with a perforated titanium leading edge surface. Subsequent testing of the leading edge insert panels was done to evaluate the performance of improved perforated titanium surface material relative to the alternative Dynapore surface. This evaluation was done using an improved substructure which did not require sewing of the fiber glass cloth during fabrication.

Secondary objectives of the swept wing model test included: (1) evaluation of suction requirements relative to analytical criteria, (2) investigation of effects of surface anomolies, panel joints, etc., and (3) development of fabrication techniques applicable to full scale surface panel construction. A description and summary of the swept wing wind tunnel model design is presented in the following paragraphs: 8.2.1 Model Design in order to' provide representative swept wing flow for testing the laminar flow control surface, a large chord model was required. This necessitated the /. i use of special design procedures because of the limited size of the Douglas

Low Speed Wind Tunnel test section which was .965 x 1.372 m (38 x 54 in). The

design procedure used to develop the model profile and sidewall fairings is

outlined below: .

The desired pressure distribution was selected from previous airfoil design work for the LFC aircraft configuration. This pressure distribution is shown in Figure 8-8. Although the design pressure distribution shown is for Mach 0.8, this upper surface pressure distribution was selected so that the appropriate pressure profile would exist throughout the LFC test region on the upper surface of the model.

.

Development of the airfoil shape for the model was accomplished using 2-D methods (Tranen, Ref. 7.4-1, and Neumann, Ref. 8.3-i) to obtain the desired upper surface pressure distribution in the presence of the test section floor and ceiling. Numerous iterations were necessary to achieve the required normal profile shown in Figure 8-9. The resulting normal pressure distribution (Figure 8-10) illustrates the strong effects

C

of the tunnel wall restraint and the compromise imposed on the lower surface pressure distribution. It is obvious that the resulting airfoil is relatively thick and causes tunnel blockage of approximately 28 percent.

Fortunately, the distortion of the lower surface pressure distribution is such that the section lift and resulting structural loads on the model and support structure are reduced.

.

Sweep angle - Originally, variation of sweep angle was considered in order to achieve an attachment line Reynolds number (Ro) greater than 100.

However, the length of the tunnel test section precluded consideration of sweep angles greater than 30 degrees. Reducing sweep would necessitate further increases in airfoil thickness to obtain the increase nose radius required. This was deemed advisable since the airfoil was already inordinately thick. The predicted value of Re for this profile at 30 degrees sweep was approximately 90 and it was possible that contamination from the tunnel wall OF POOR QUALITY' 0.7 TRANSITION LOCATION XTR/C 0.6 %7 _7 80 x 700 DYNAPORE 0.5 0 50 x 250 DYNAPORE • _ SMOOT_H PLATE [] PERFORATED ....

TITANIUM SLOTTED ALUMINUM 0.4 t> METALLIZED DOWEAVE <> SINTERED FIBERMETAL n ON DOWEAVE 0.0001 0.001 0.01 SUCTION COEFFICIENT CQ 8-DP-80055 Q FIGURE 8-7. COMPARATIVE EFFECTIVENESS OF LAMINAR FLOW CONTROL SURFACES ° :..

M = 0.8, C L = 512 30-DEGREE SWEPT WING -1.2 - MN = 0.7 NORMAL Cp = _'P -1.0

P=

-0.8 -0.6 -0.4 -0.2 I NORMALC r 0 10 20 30 40 50 66 T0 80 90 %_00 0.4 X-PERCENT CHORD _ 0.6 0.8 TYPICAL AIRFOIL SHAPE Z-PERCENT I I t I 70 CHORD 39 40 50 6_ 90 100 'IJkGIE N-2,5,2,,_3 -10 X-PERCENT CHORD FIGURE 8-8. SWEPT-WING MODEL TEST - REPRESENTATIVE CHORDWISE PRESSURE DISTRIBUTION I ORIGINAL PAGE _'S' OF POOR QUALITY' TUNNEL CEILING TUNNEL FLOOR tI-GEN-2¢_237 FIGURE 8-9. SWEPT-WING MODEL - AIRFOIL PROFILE IN WIND TUNNEL --1.5 -1.0 NORMAL -0.5 PRESSURE COEFFICIENT 0.5 0 20 40 60 80 100 PERCENT CHORD _GF4_.m FIGURE 8.10.

SWEPT-WING MODEL - CHORDWlSE PRESSURE DISTRIBUTION OF AIRFOIL fT. "P INSTALLED IN TEST SECTION boundary layer would be sufficient to excite boundary layer instability along the attachment line; it was therefore decided to proceed with the model design using the profile as defined with a sweep angle of 30 degrees.

Suction Requirement - Suction velocities for the swept wing model were determined using the X-21 boundary layer stability criteria. Estimated suction requirements for a nominal dynamic pressure 2.15 kPa (45 psf) are presented in Figure 8-11. C'ross flow instability establishes the suction level required over the forward 18 percent chord of the model and over the pressure recovery region aft of approximately 65 percent chord. In between the forward and aft steep pressure gradient regions, the suction is determined by the streamwise Tollmein-Schlichting instability criterion.

.

The first step in the development of the tunnel sidewall fairings was to calculate streamline traces for the infinite yawed wing profile, in the presence of the tunnel floor and ceiling, at several vertical stations.

After comparison of the streamlines, the one passing through the station 0.005c above the crest of the upper surface was selected as the reference streamline. This streamline is shown in Figure _-12. The three-dimensional analytical method (Ref. _.3-i) and computing program (Ref. 8.3-2) were used for these computations.

Verification of the significant influence of the tunnel wall contour is shown in figure _-13 where the pressure distributions on the upper surface of the model are shown for these spanwise stations. Three-dimensional Neumann calculations were made for the swept wing in the wind tunnel with both straight sidewalls, and with the sidewalls contoured two-dimensionally corresponding to the reference streamline. With straight sidewalls, at the inboard station the pressure peak is suppressed while at the outboard station it is accentuated. The distortion of the pressure profile would be intolerable without contoured walls in the test section.

Practical contouring of the test section sidewalls involved further compromise as illustrated in Figure 8-14. Obviously, the reference streamline, extended two-dimensionally from tunnel floor to ceiling and when translated to be S _ V _ = 58.5m/s (192 FPS) ORIGI_w,_-,L ._ , _, A = 30 DEG -1.5 OF POOR QUALI'IY O0000QO00000 0 0 0 O0 O -1.0 0 000 O 0 O0 .0 -0.5 CP N 0 I I I I I I I I I l 0.5 1.0 -0.09 -0.25 -0.20 -0.06 !ROSS- _ TOLLMEIN-SCH LICHTING__IICROSS-FLOW Vw -0.15 m/s FT/SEC -0.04 -0.10 -0.02 -0.05 c.,T,cA.

' ' ' I i I I I I I I 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 X/C _E. z_7. ] FIGURE 8-11.

SWEPT-WING MODEL - UPPER SURFACE PRESSURE DISTRIBUTION AND SUCTION VELOCITY DGE DGE . -- -- L -- __ nL , 1.1,, ___,, i ' 9-GEN-_I_36 FIGURE 8-12.

SWEPT-WING MODEL TEST - REFERENCE STREAMLINE ON UPPER SURFACE-INFINITE YAWED WING "-0.81 f...

-0.( Cp _0 ,'( -0.,_ 0o o= 0 DEG o STRAIGHT WALL WAVY WALL FIGURE 8-13. SWEPT-WING MODEL TEST - EFFECT OF TUNNEL SIDEWALLS ON PRESSURE DISTRIBUTION - WITH SIDEWALLS SHAPED TO MATCH REFERENCE STREAMLINE STA 0 STA 120 t" TEST SECTION "_I

__: __._._..__._=_-__ - --

F INBOARD SIDEWALL FAIRING L

----V _-_--_- \ _._'_._NC_

.......... _oo- A

---* , .____ • _--- --er-

OUTBOARD SIDEWALL FAIRINGS 9._EN 2¢_, 7 3 FIGURE 8-14. SWEPT-WING MODEL TEST - SIDEWALL FAIRING DEVELOPMENT tangential to the test section sidewalls, cannot be reconciled with the tunnel ,/ contraction upstream and the diffuser downstream. Hence, the sidewall fairings indicated by the cross-hatched areas in Figure B-14 were faired using experienced and intuitive judgment. As shown later in Section 8.2.3, the results were satisfactory.

8.2.2 Model Description and Installation

a)

The basic model, as noted previously, was a two-dimensional, thirty degree, swept wing. The chord was 1.8 m (6 ft) normal and 2.11 m 16.93 ft) streamwise; the thickness/normal chord ratio was 0.1504.

Leading-edge and upper-surface panels were removable as illustrated in Figure B-15. 15-percent chord, simple trailing edge trim flaps were provided as a means of adjusting the pressure level without having to pitch the entire model. The trim flaps were in three segments so that differential setting could be used to adjust the flow and provide a uniform spanwise pressure distribution. This adjustment was used to compensate for the compromise involved in the design of the sidewall fairings. The basic model structure was built up from aluminum spars and ribs as pictured in Figure 8-16. Aluminum and fiberglass sheetmaterial was used for the non-removable sections of the model surface.

The LFC test surface extended from below the leading edge at 0.036 chord to 0.70 chord on the upper surface. The juncture between the leading edge panel and the upper surface panel was located at the model front spar (0.18 chord). Non-porous panels were installed for initial testing of the model so that chordwise and spanwise pressure distributions could be obtained and the natural transition of the boundary layer determined.

This was used to provide a reference of transition location on a smooth surface and to measure and adjust the surface pressures on the model to the required levels. The Dynapore surface was not amenable to the instarllation of static pressure orifices and its surface texture, plus its porosi1_y, could not provide a reliable reference for boundary layer transition.

OF POOR QUALITY _S PANEL (DYNAPORE) FORE AND AFT SECTIONS _STRUCTURE 9-GEN._N_,276 FIGURE 8-15. SWEPT-WING MODEL - COMPONENTS S _o , FIGURE 8-16. SWEPT-WING MODEL - BASIC STRUCTURE ,f -_ ( The porous surface used initially for testing laminar flow control was Dynapore having an 60 x 700 outer layer diffusion bonded to an 80 x 80 sublayer. This was supported by a fluted fiber-glass substructure which also formed spanwise plenums for the suction system. Due to the high porosity of the Dynapore selected for ease of liquid clearance from the surface, an additional porous fiber-glass metering layer was inserted between the Dynapore and the supporting sublayer. The latter consisted of punched fiber-glass having 6.35 mm (0.25 in) diameter holes distributed to provide a 40-percent open sheet and provided support for the porous surface between the flutes nodes. There were 21 flutes in the leading edge panel and 16 wider flutes in the upper surface panel. On the leading edge panel the nominal distance between node lines (i.e., flute width at the surface) was approximately 17.8 mm (0.70 in). For the panel over the mainbox this dimension was approximately 57 mm (2.25 in). The panel structure buildup is illustrated in Figure 8-17 and a photograph of the Uynapore porous panels is presented in Figure 8-18. The latter shows the suction tubes extending beyond the ends of each plenum. Figure 8-19 shows the basic model during fitting of the upper surface panel.

The alternative leading edge panel utilized a Similar substructure.

However, the surface material was 0.635 mm (0.025 in) perforated titanium. Perforation of the titanium was by an electron beam process which provided elongated holes, nominally 0.102 x .203 mm (O.O(J4 by 0.008 in), spaced 1.27 mm (0.050 in) along rows. A close-up photograph of the perforated titanium leading edge surface is shown in figure 8-20 and the titanium surfaced leading edge panel is pictured in figure 8-21.

It should be noted that in the construction of the basic model and LFC porous panels, structural joints, welded seams, and the like were to production standards rather than using laboratory or research quality methods. This was in keeping with the primary objective of the test, that V was intended to demonstrate LFC under conditions as realistic as possible. A description of the structural design and development of the LFC surface is presented in Section 9.9.

OF PO0,_ Q;,..;,;:.L! _'V

f ELECTRON BEAM PERFORATED DYNAPORE ALTERNATIVE _ .

7 //f f / • PERFORATED ER IMPACT AND FLOW CONTROL SHEET NONPOROUS AREAS STRIP PLAN OF WT MODEL 8]-GEN 223_7-1 FIGURE 8-17. WIND TUNNEL MODEL - POROUS PANEL CONFIGURATIONS :IGURE 8-18. SWEPT-WING WIND TUNNEL WITH DYNAPORE POROUS SURFACE \.

t76

ORIGI_',AL PAG_ _r_

OF POOR QUALITY

FIGURE 8-19. SWEPT-WING MODEL WITH UPPER POROUS PANEL INSTALLED

\

FIGURE 8-20. PERFORATED TITANIUM SURFACE NEAR FRONT SPAR JOINT SHOWING NONPOROUS REGION AND WELDED JOINT i f Installation of the model with the nonporous panels in the Douglas Low Speed Wind Tunnel is shown in Figures 8-22 and 5-23. These photographs enlDhasize the size of the model relative to the wind tunnel test section.

The test section sidewall fairings were constructed of thin aluminum sheet supported by wooden ribs which were placed horizontally along the sidewalls. The sidewall fairings were fitted into place after installation of the basic model. A plexiglass ceiling was provided with small ports for inserting the transition probe, Figure _-22 shows the sidewall window used for observation and probing of the leading edge attachment line region.

The suction for the test was provided by an electrically driven vacuum pump. A schematic of the suction system is shown in Figure 8-24 and a photograph of the installed suction systems is shown in Figure 8-25.

Individual plenums were connected to the secondary manifolds by 12.7 mm (I/2 inch) plastic tubing. These manifolds allowed rearrangement of the individual suction tubes as testing required. This configuration allowed

C[

plenums with generally the same surface pressure and suction requirement to be grouped together and controlled by a single valve. It was not within the available resources of the test program to provide a separate flow meter for each plenum. Twelve Meriam laminar flow elements (meters) were used and the flow was controlled by 12 simple gate valves. The 12 suction flow channels were then connected to the primary manifold which was connected to the suction source through a large gate valve.

b)

Leadin 9 edge insert model modification The nonporous leading edge panel was modified to support leading edge insert segments. A sketch of the model with a leading edge insert is shown in Figure 8-26. Figure 8-27 is a photograph of the modified leading edge panel with the insert removed and internal suction lines exposed.

The insert leading edge configuration was selected in order to: (i) allow the leading edge insert segments to be changed without having to remove

OF POORQJALITY

\ FIGURE 8-21. PERFORATED TITANIUM LEADING EDGE Wlc SHOWING SUCTION AND STATIC LINES

i

FIGURE 8-22. SWEPT-WING MODEL TEST - MODEL INSTALLATION OFd:.7_:: _;._" ' . :- OF POOR QL;_,.!._,', FIGURE 8-23. SWEPT-WING MODEL TEST - UPPER SURFACE AT LEADING EDGE /f:" SUCTION PLENUMS POROUS SURFACE PLENUM PRESSURES PLENUM SUCTION LINES SECONDARY MANIFOLDS LAMINAR FLOWMETER: FINE ADJUSTMENT VA COARSE ADJUSTMENT VAt TO SUCTION SOURCE PRIMARY MANIFOLD FIGURE 8-24. SWEPT-WING MODEL TEST - SUCTION SYSTEM SCHEMATIC OF POOR QUALIT'Y FIGURE 8-25. WIND TUNNEL MODEL WITH SUCTION SYSTEM NORMAL _.. OF INSERT-_ NONPOROUSLAND-_ _EAMWISE _-__ POROUSSURFACE - _ - s6s..ts AND PRESSURE TUBES B ) -GIEN.22;IIg6 FIGURE 8-26. ARRANGEMENT OF LFC PANEL INSERT IN WING LEADING EDGE the model trom the tunnel, (2) extract suction more realistically part way along the plenums rather than at the ends, and (3) reduce the cost of additional test specimens.

Two insert panels were constructed using an improved substructure which provided a porous strip LFC suction surface. A sketch of the substructure is shown in Figure _-28. The outermost sublayer formed the bonding surfaces for the porous surface sheet and acted as a baffle between the porous surface and the plenum. Holes 4.76 mm (3/16 in) diameter were drilled in the baffle at approximately 19 mm (3/4 in) pitch.

The first porous surface tested was 80 x 700/8U x 80 Dynapore diffusion bonded to a 0.305 mm (0.012 in) stainless steel sheet perforated with 0.254 mm (0.01 in) diameter holes at 2.54 mm (0.10 in) pitch in a square pattern. The second porous surface tested was the improved electron beam perforated titanium. This titanium surface material was 0.635 mm (0.025 in) thick perforated with 0.0635 mm (0.0U25 in) diameter holes spaced 0.813 mm (0.032 in) apart in a square pattern. A photograph of the

(

perforated titanium insert installed in the tunnel is shown in Figure 8-29.

8.2.3 Model Instrumentation Instrumentation for the swept wing model test was conventional and relatively simple. The nonporous reference surface contained three chordwise rows of static pressure orifices, one row on the tunnel centerline and a row on each side, 381 mm (15 in) from the centerline (static pressure orifices could not be installed conveniently with the porous surfaces). Just aft of the 70-percent chord station, three total pressure rakes each with three tubes, were located similarly on the centerline and 381 mm (15 in) on either side spanwise. The total pressure tubes were at 2.54, 5. U8 and 10.16 mm (0.1, 0.2, and 0.4 ip) above the model surface. These dimensions were selected to be within the turbulent boundary layer and mostly above laminar boundary layer at the 70-percent chord station. For the leading edge insert panel tests, nonporous leading edge surfaces were available at each end of the porous L: OF POOR Q_,-,L_FY" ''_' ° / FIGURE 8-27. WIND TUNNEL MODEL - LEADING EDGE INSERT REMOVED /_ PERFORATED TITANIUM OR DYNAPORE _._L_//.I v_ JU.CT,O. W,T. PLE.U.

\ "_J_"_ IJ S/ _ FLOW HOLES __ _/ _-" THROUGH BAFFLE PLENUMS 81-GEN-22399 FIGURE 8-28.

LEADING EDGE INSERT SUBSTRUCTURE CONFIGURATION AND SUCTION OUTLET TUBING

C_ ¸

insert panels. These were used to provide a row of static pressure holes located 3_.i mm (1.5 in) inboard and outboard from the ends of the insert panels. Thirteen pressures were measured in each row.

During testing of the porous surfaces, plenum pressures were measured in each plenum, with suction on and off. All of the surface, plenum, and rake total pressures were measured on manometer boards and recorded photographically.

To obtain flow quantities through each flowmeter, measurement of the upstream pressure and the pressure differential for each flowmeter was necessary.

These pressures were measured with a transducer via a scan, valve using a digital voltmeter. A schematic layout of the flow data acquisition is presented in Figure 8-30.

The transition location was determined by ear using a simple medical stethoscope connected to a fine total head probe with its opening held within the boundary layer. Originally a hot film sensor was planned for this purpose, however, the problen_ of aligning and maneuvering the hot film probe over the F curved surface proved to be impractical.

8.2.4. Checkout and Initial Calibration with Nonporous Panels - The first phase of the swept wing test program involved checkout of the basic model installed in the tunnel with the nonporous panels in place. The objective of this phase was to identify and establish the required flow conditions for testing the LFC porous panels with suction applied. The test procedure and results for the nonporous surface is outlined below: .

Flow Separation Check - Model upper surface and sidewall fairings were tufted to check for expansion separation. With the relatively large diffusion of the flow in the aft portion of the test section, there was concern that separation might occur. The initial runs, up to maximum r dynamic pressure, confirmed that no separation existed for the test configuration.

OF POOR QUALITY o LFC SWEPT-WING WITH PERFORATED TITANIUM II_SERT IN DOUGLAS FIGURE 8-29.

LONG BEACH WIND TUNNEL LFC SWEPT WING LENuMS SUCTION LINES SECONDARY MANIFOLD MERIAM .,_ I F SCANIVALVE SYSTEM LAMINAR FLOW ELEMENT J y / SUCTION /,_ SOURCE f MAIN MANIFOLD --_

S

TRANSDUCERS I DIGITAL READOUT J 8) -GEN-2239!

FIGURE 8-30. SCHEMATIC LAYOUT OF DATA ACQUISITION Reference Tunnel Velocity - Due to the exceptionally large amount of blockage with this model, it was necessary to establish the reference dynamic pressure for the test. This was accomplished by locating a pitot static tube in the tunnel ahead of the model on the centerline reference streamline. At the corresponding streamline station the velocity ratio, V/V_ , based upon the analytical (3-D Neumann) value, was 0.9634 (Figure B-31). This correction was then applied in determination of the tunnel velocity calibration shown in Figure _-32.

.

Reference Pressure L)istribution - The reference pressure distribution for the test was established by adjusting the three trimmer flaps along the model trailing edge. The flaps were first deflected systematically so that the upper surface pressure distribution closely matched the desired pressure distribution (Figure 8-33). Spanwise uniformity was then achieved by differential setting of the trailing edge flap segments. The required flap positions were, nominally.

Inboard I degree t.e. up Center 2 degrees t.e. up Outboard I degree t.e. down The resulting pressure distributions at the three spanwise stations are compared in Figure _-34.

It should be noted that the model pressure distribution had a slightly steeper adverse gradient relative to the design value. This was considered acceptable since the steeper gradient imposes a more severe condition for possible crossflow instability.

.

Attachment Line Location - Centerline chordline pressures, near the airfoil leading edge, were used to locate the attachment line as indicated in Figure 6-35. In addition flow visualization (Figure _-36) verified the attachment line location to be 10.16 mm (0.4 in) (S/C = 0.006) above the leading edge. The resulting value of the attachment line Reynolds number (Rs) was approximately 65 for the maximum test velocity, l his value is significantly below the nominal critical value of R 8 - 100, and indicated an absence of instability along the attachment line for the test conditions.

,-._-, _ • r. r"t ORhah,:_.L P/..-._i_.,,,, OF POOR QUALITY' 0.98 % 0.96 I V/v=_ 0.9634 I 0.94 0.92 0.90 V 0.88 Voo 0.86 PffOT-STATIC TUBE POSITION 0.84 0.82 0.8O 0.78 0.76 34.0 34.1 342 34.3 34.4 34.5 34.6 34.7 STREAMWISE COORDINATE FIGURE 8-31. LFC THEORETICAL STREAMLINE - 14.22 mm (0.56 IN.) ABOVE MODEL SURFACE ON TUNNEL G L /f'_ L (fps) m/s (250) 0 (() (10) (20) (30) (40) (50) (60) (LB/FT 2) I I I I I J 0 500 1000 1500 2000 2500 3000 Pa 8L -GEN-22308 DYNAMIC HEAD -- q NOMINAL FIGURE 8-32. TUNNEL VELOCITY CALIBRATION - MODEL INSTALLED IN TUNNEL ORIGIN;:L r.,,,, ........

OF POOR QUALITY' --1.0 • oO _)o o 0 _o --0.8 [] o o o o n 0 D 4.6 o n o 4.4 CP s [] O DESIGN 2-D INFINITE bWVEPT WING IN o 4-02 "-"-'-IIV-- TUNNEL WITH FLOOR AND CEILING {] O REPRESENTATIVE 3-D SWEPT WING IN c TUNNEL WITH FLOOR AND CEILING c l,.J AND REFERENCE STREAMLINE WAVY r TUNNEL SIDEWALLS.

• EXPERIMENTAL LB498A (WIA) RUN NO. 25 r J J ! I I 0.4 1.0

o o 0.2 0.4 0.6 0J_

S/C @I-GEN-223_ FIGURE 8-33. STREAMWISE PRESSURE COEFFICIENT VERSUS SURFACE STATION

.1.o

C ] P 0 0 381 mm (15 IN.) OUTBOARD 0.5 [] CENTERLINE I A 381 mm (15 IN.) INBOARD 1.0 0.10 0.20 030 0.40 0.50 0.60 0.70 0.80 X/C 9"GEN-_A FIGURE 8-34. SWEPT-WING MODEL TEST - SPANWlSE PRESSURE DISTRIBUTION COMPARISON f; _;\ ,-,,_:. [3 OF POOR QUALITY (200)!

6O (160) (120) (8O) V.

A'n'ACHMENT LINE m/s 20 tin (40) (0) --10 (-4O) --20 (--80) (0.00) (0.04) (0.08) (0.12) (0.16) (020) (02.4) (FT) I I I I I I I I 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 m SURFACE DISTANCE FROM LEADING EDGE .,oE,2,_., FIGURE 8-35. R_ EVALUATION - SURFACE NORMAL VELOCITY VERSUS SURFACE DISTANCE \ THE ATTACHMENT LINE POSITION IS APPROXIMATELY 0.4 IN. ABOVE LEADING EDGE VELOCITY qNOM -- 50 PSI FIGURE 8-36. NONPOROUS SWEPT-WING - BASELINE FLOW VISUALIZATION WITH Ti 02 OIL MIX During the test, a .b35 mm (O.U2b in) thick boundary layer tripping strip was placed transversely across the flow along the inboard leading edge in an attempt to induce boundary layer instability and cause transition along the attachment line. This technique failed to cause any transition other than streamwise.

Reference Transition Location - Boundary layer transition was located on the nonporous upper surface pane] for several test conditions. Transition was determined using a small total head probe attached to a stethoscope.

This technique was far more satisfactory than the hot film sensor used in the aerodynamic smoothness testing. The total head probe was: (I) less sensitive to orientation in the airflow, (2) not susceptible to damage while probing, and (3) much more adaptable to probing the curved surface with limited access from ports in the tunnel ceiling.

The natural transition location for the smooth nonporous upper surface is shown in Figure b-37 as a function of Reynolds number. A check was made using one screen in the setting chamber; this reduced tunnel turbulence and resulted in the transition location shifting aft by approximately 8 percent chord. Since the purpose of the test was to demonstrate LFC with realistic adverse conditions, the screen was not used for any of the testing with suction. This also allowed the tunnel velocity to be increased.

For the primary test condition at a nominal tunnel qnom of 2.394 kPa (50 psf), which is equivalent to a Reynolds number of 9 x lO 6, the natural transition on the nonporous upper surface was located at 8 percent chord.

.

Surface Flow Visualization - As a result of streamwise flow of the excess titanium oxide/oil mixture being used for attachment line flow visualization, the upper surface streamline was traced to beyond 7U percent chord as pictured in Figure 8-38. The surface oil streaks were scaled from che photos and are compared with the reference streamline that was used to develop the sidewall fairings (Figure _-39). Although these streamlines are not directly comparable, the flow curvature is representati ve.

ORiG!_',J,_L "" ' "'-' _'-" OF POOR QUALITY _ 1 SCREEN INSTALLED

2O

TRANSITION

LOCATION

PERCENT

CHORD

_ (NO SCREENS)

.,7, \

(NO SCREENS)

i 5 7 9

R N (MILLIONS) g-GE _ _,3t0 A FIGURE 8-37. SWEPT-WING MODEL TEST - NATURAL TRANSITION LOCATION ORIGt_L F?_GE _S OF POOR QUALITY -= --1 ° 02' ' = -2 ° 04; lp - 0 ° 57; _fc _fs (_ _ 0 DEG TRANSITION FREE = FIGURE 8-38. EFFECT OF CROSS-FLOW ON STREAMLINES WITH NONPOROUS SURFACE ON SWEPT 8.2.5 Calibration and Testin 9 with Porous Surfaces Testing of the porous surface panels to determine; (i) effectiveness in maintaining laminar flow, (2) the amount of suction required, and (3) effects of surface and suction anomalies, is reviewed in the following paragraphs.

Zo Surface Check and Calibration - Prior to installing each porous panel, porosity was checked to determine the porosity of the surface associated with each plenum. This was done by applying suction to each plenum and recording the flow rate as a function of the pressure differential across the surface. During calibration of the basic Dynapore porous panels, it was detected that the resin used for the plenum walls was, in fact, porous. It was therefore necessary to apply additional sealing material in the spaces between the plenums and on the inner panel surface so that leakage through the plenum walls and between adjacent plenums was eliminated insofar as was possible. A routine leakage check was conducted on all subsequent panels.

f Calibration data were reduced to provide a porosity reference value in terms of flow rate, in standard cubic feet per minute (SCFM), at 670.3 Pa (14 psf) pressure differential. Porosity distributions for the basic model porous panels are shown in Figures 8-40 and 8-41. The latter figure shows an expanded porosity comparison for the Dynapore and perforated titanium leading edge panels. Variations in local porosity of the basic surface material together with possible remaining leakage between adjacent plenums and localized internal blockages introduced during fabrication account for the resulting porosity distributions.

Porosity distributions for the Extension Test Dynapore and perforated titanium leading edge insert panels are presented in Figure 8-42.

2. Basi_Model Test Results, Extent of Laminar Flow and Suction Requirements lhe basic swept wing model was tested first with all Dynapore upper surface panels. Suction was varied chordwise to generally minimize the suction flow needed to maintain laminar flow. The extent of laminar flow OF PO0]:_ _i" :.:;-, o REFERENCE STREAMLINE -- APPROX15 mm (0_ IN.) ABOVESURFACE • " SURFACE STREAMLINE -- PHOTO J9-11064 × SURFACESTREAMLINE -- PHOTOJ9-11061 (FIGURE 8-38) O,_G EN L_o28,2 A FIGURE 8-39. SWEPT-WING MODEL TEST - COMPARISON OF REFERENCE STREAMLINE AND SURFACE FLOW VISUALIZATION t (SCFM/FT 2 ) m3/min/m 2

(30)i _ \\ --

,.o _1- [C'" ........

J1 II,qr T- , :. ,, | AVG = 0.4635 m /rain I Ii " ill : | J 0 10)' I,, , ":- ' ' I ' ' ' 0.8 0.9 1.0 -:-:-0.1 0 0.1 03 0.4 0.5 0.6 0.7 0"2 i S/C FRONT SPAR JOINT SPANWlSE 8] _EN-22392 BUTT WELD FIGURE 8-40. POROUS SURFACE POROSITY DISTRIBUTION I

OF POOR QUAUTV

f DYNAPORE

.P

........ PERFORATED TITANIUM m ..J EB HOLES 0,102 mm x 0.203 mrn (0.004 x 0.008 IN.)

!-4 m (SCFM/FT 2) G.

o4 (20) O_ m3/min/m 2 O 5.0- POROSITY TARGET_ (15) 4.0- MJ 3.0 - (10) p,, 2.0- O (5)1 iu -== =ul M.

0.1- l I I I 0.08 0.10 0.12 0.14 0.16 0.18 020 0 0.02 0.04 0.06 S/C 81 _EN-22393 FIGURE 8-41. LEADING-EDGE POROSITY DISTRIBUTION • AVERAGE POROSITY OF PLENUM -- FLOW RATE AT 0_-----------------_32 Pa(14 LB/FT 2) QAV Ap • PLENUM POROUS AREA (SCFM/FT 2) (120] m3/min/m 2 35 - 30 -( 100] 25 - (80] POROSITY 20 - (6o_I- OAV/Ap 15- NON,POROUS I"---! TITANIUM PRIOR TO BONDING TO MODEL_,_ I : 0_041 s/c =_ (40)1-- 10 i- ! IPi_ F PERFORATED TITANIUM _ Ii i (20)1--

r-4-J"-"------I ,_ ..... J_.,.. _.,.--,-::,._\ II

L--._- - ---I _ NONPOROUS -- _",--- --

I I

I I i; 0.039 sic

iI-

I,I , I L I J I I J I z HI

(0) I --0.02 0 0.02 0.04 0.06 0.08 0.10 0.120 0.140 0.160 0.180 0.20 SIC SURFACE LOCATION OF PLENUM NODES MEASURED FROM LEADING EDGE OF MODEL 61 _GEN.22;q_ FIGURE 8J,2. LEADING-EDGE INSERT POROSITY DISTRIBUTION _- obtained with the Dynapore LFC surface is illustrated in Figure 8-43.

Application of suction extended the boundary layer transition location to beyond the end of the porous surface at 70 percent chord by as much as 5 to I0 percent of the chord. The extent of laminar flow obtained with the perforated titanium leading edge was essentially the same.

Suction velocity distributions for the basic model, with both Dynapore and perforated titanium leading edges, are shown in Figure 8-44. The stepwise distribution is due to the grouping of several plenums to a single manifold.

Included in this figure is the predicted suction velocity required according to X-21 criteria. Suction velocities for the Dynapore surface panels agree quite well with predicted values.

Increased suction ahead of, and behind, the non-porous front spar joint is required. The locaI increase in suction is approximately equivalent to what would have been required for a porous surface in the non-porous region.

The overall suction level with the perforated titanium leading edge is higher than with the Dynapore leading edge. However, it should be noted that at the time of this test the titanium was only considered as an alternative surface material and the suction values were not refined to the same degree as they were for the Dynapore. The principal objective at this time was to determine whether possible disturbances, originating at the perforations in the thin laminar boundary layer near the leading edge, would cause premature transition. Even with over-suction in the leading-edge region, laminar flow was obtained for the same extent of chord as for the Dynapore.

t Simulated System Malfunction - During testing with the all Dynapore Surface, an investigation was made of the effect on laminarization of system failures in the form of simulated suction line venting to ambient or blockage at the most critical location. The line feeding plenum number 7 (s/c = 0.025), which was in the crossflow critical region, was selected. Venting #7 moved transition forward to 73 percent chord. Blockage of the same line caused oversuction in other areas but had no detrimental effect upon the extent of laminarization. The conclusions can be drawn that venting of a suction line is more critical to LFC operation than blockage and that blockage would cause an increase in the suction source loading.

OF PO'.:'" " _' __

\

(a) SUCTION ON TRANSITION- SUCTION OFF PARTIAL VENTING THROUGH SYSTEM OF B.

DUE TO SUCTION TRANSITION - NONPOROUS SUR FACE FLOW [_ 1.!

(a) TRANSITION - SUCTION OFF 2.( SURFACE VENTED THROUGH MANIFOLD (a) TOTAL HEAD PROBES (b) NONPOROUS SIDEWALL WEDGES 10.16 mm (0.40 IN.) ABOVE SURFACE $O-G E1_'-22126 -4 FIGURE 8-43. SWEPT-WING MODEL TEST - EFFECT OF SUCTION ON TRANSITION qNOM -- 2394 Pa (50 PSF) m/s 2.0 fps I- " w 24 m ¢,L CROSSFLOW CROSSFLO _ -I " _

>= z.s "" "'t [R'T'CAL _ ':__ ]2.0sCR'T'CA_\162a m <

(0.4) PERFORATEDT, LEAD,NG EDGE 1.8 2

_> =--_ z.2

'°"lla 1

(0.2) Ei "--7 x.o <

: i [_ o.8

i ..... o.6 8 <

_._ 0.05 " O CROSSFLOW CRIT. -4 (01) 0.2 u_ 0 q uu_....____....¢_ -__---_---J 0 DO 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 ] S/C ATTACH M E NT LINE POSITION ,,_._,,,o J • FIGURE 8-44. SWEPT-WING WIND TUNNEL MODEL - SUCTION VELOCITY DISTRIBUTIONS f - .

Leading Edge Insert Test Results - The results obtained with the leading edge insert panels are presented in Figures 8-45, 8-46, and 8-47. Because the porous leading edge inserts were limited spanwise, the extent of laminar flow with suction on the insert was reduced by the nominal spread of turbulence from the insert extremities. In this case the prime objective was comparison of the leading edge surface materials with regard to LFC suitability. In spite of its limited span, it was possible to demonstrate laminar flow to beyond 60 percent chord with the new perforated titanium surface using suction to 50 percent chord over the main box regions. The testing effort was concentrated on the improved titanium surface because of its quality and preferred structural characteristics. The increased suction applied ahead of the spar joint for the perforated titanium leading edge insert, shown in Figure 8-47, was computed to compensate for the suction required across the non-porous region at the front spar joint. The resulting laminar flow, to 65 percent chord (Figure 8-46), corresponds to the suction distribution indicated. A similar result would be expected for a Dynapore leading edge insert.

8.2.6 Boundary Layer Stability Analysis Stability analysis was made of the wind tunnel test results, using the advanced boundary layer SALLY II computer code for comparison. Stability analysis was applied both with and without suction at a nominal tunnel dynamic pressure of 2394 Pa (50 Ib/ft2), which corresponds to a freestream velocity of 66.08 m/s (216.8 ft/sec), and a chord Reynolds number of 8.87 x 106 . The boundary layer development was calculated with the experimentally determined pressure distri- bution to avoid oscillations in the boundary-#ayer solutions. All stability analysis was done using the envelope method option, that is, the amplification ratios to be integrated were maximized for each selected physical disturbance frequency.

OR?G'p'+I OF PO0_,' \

\

1.5 ITION - SUCTION ON 1!

NONPOROUS LEADING EDGE 1 END PIECES TRANSITION - NONPOROUS SURFACE 1,( F LOW [_ 1.!

(a) TRANSITION - SUCTION OFF SURFACE VENTED THROUGH MANIFOLD (b) NONPOROUS SIDEWALL (a) TOTAL HEAD PROBES 80-GEN.22176 -3 10.16 mm (0.40 IN.) ABOVE SURFACE WEDGES FIGURE 8-45.

SWEPT-WING MODEL TEST - EFFECT OF SUCTION ON TRANSITION AT qNOM = 2,394 Pa (50 PSF) - DYNAPORE L.E. INSERT \

\

1.5 (a) SUCTION ON 1!

NONPOROUS LEADING EDGE 1 END PIECES LAMINARIZATION OF BL DUE TO \ a) TRANSITION - NONPOROUS 0-'= SURFACE 1.( F LOW [_ 1.!

TRANSITION - SUCTION OFF SURFACE VENTED THROUGH MANIFOLD SYSTEM (a) TOTAL HEAD PROBES (b) NONPOROUS SIDEWALL 10.16 mm (0.40 IN.) ABOVE SURFACE WEDGES 60-GEN-22126 -2 FIGURE 8-46. SWEPT-WING MODEL TEST - EFFECT OF SUCTION ON TRANSITION AT L' qNOM = 2394 Pa (50 PSF) - PERFORATED TITANIUM LEADING EDGE INSERT* OI_iGtf',;,-_L PAGE ig OF POOR QUALITY fps L O X-21: CROSSFLOW CRITICAL m/s (0.5) & X-21: TOLLMEIN-SCHLICHTING CRITICAL 0.15 START OF NONPOROUS JOINT O W1D2: EB PERFORATED TITANIUM - OPTIMIZED 3= WITH ALLOWANCE FOR NONPOROUS FS JOINT (0.4) _.---0-----_ _I'W1DI: DYNAPORE INSERT t 0.10 - O ,-J (0.3) / _...._ z / (0.21 ..

O m I-- x.21 CRITE.IA 0.05 ggmD@@ ¢/} 0 0 - _ , , , r I I | , 0.02 0.04 0.06 0,08 0.10 0.12 0.14 0.16 0.18 0.20 S/C (MEASURED FROM LEADING EDGE) 81 "C,E N-Z2459 FIGURE 8-47. SWEPT-WING MODEL - LEADING EDGE INSERT SUCTION VELOCITY DISTRIBUTION (a) No Suction This case corresponds to tests on the model with non-porous surface. The measured location of the start of transition (assumed to be where the frequency of turbulent bursts recorded by the stethoscope is about once every two seconds) is at x/c = 0.08. Since cross-flow instability was suspected of being the transition triggering mechanism, a range of frequencies from 0 to 1500 Hz was considered. The most amplified frequency computed was at 750 Hz, which gave an amplification N-factor of 8.3 at the start of transition. The zero frequency case gave the smallest amplification, indicated by the N-factor of 7.3. The computed range of N-factors is considerably lower than the average N of i0 obtained in low turbulence wind tunnels. The probable reason for this discrepancy is the turbulence level, which in the absence of screens, is about 0.5 percent.

Another interesting feature of the calculated results is that the maximum N-factor for the cross flow barely exceeds 11 before the flow enters the strong adverse pressure gradient on the rear of the airfoil. In view of this, and also considering the fact that the two-dimensional disturbances first

IT: start to grow at about x/c = 0.1, it is possible to visualize transition

occurring at x/c = 0.2 or even further aft if this wing were tested under low turbulence conditions. Thus, the relatively high turbulence level may be viewed as a contributing factor to generating a flow which is unstable to cross-flow disturbances, as required by the test objectives.

(b) Porous Surface Suction As noted earlier, the main difference between the test conditions and calculations is in the smoothing of the experimental suction distribution before boundary-layer calculations. This was necessary because the large discontinuous changes in the suction velocities from one suction chamber to another during the test, when used in boundary-layer calculations, produced an oscillatory solution which made the stability analysis meaningless. In order to eliminate oscillations, a new suction velocity distribution was constructed from the smoothed integrated suction mass flow disdribution. Again, the flow was tested for cross-flow instability in the range of frequencies from 0 to 1500 Hz. On the forward part of the wing the most amplified frequency was 500 Hz, giving an N-factor of 7.75 at x/c = 0.176. The lower frequencies are somewhat less an_lified but their chordwise extent reaches an x/c of 0.25 before a stable region is encountered. The maximum amplification factors calculated are less than the values which are considered to be the upper bounds for obtaining laminar flow with suction. Even lower values should be expected on account of the high turbulence level in the wind tunnel which forced the suction to higher levels than anticipated. Reamplification of disturbances on the rear portion of the wing was found only for zero frequency, starting at x/c = 0.696 but the N-factor grew only to 0.67 at the last station calculated. The last station calculated was at x/c = 0.088, or just one step past the end of suction region at x/c = 0.666 because the abrupt termination of suction caused nonconvergent boundary-layer solutions beyond x/c = 0.70. Thus the N-factors corresponding to the measured transition location at x/c = 0.75 could not be calculated.

A search for amplified waves in the two-dimensional disturbance mode was carried out in the frequency range of IUUO to 4000 Hz. Here no amplified disturbances were detected except at the very last station calculated for all frequencies below 3U00 Hz. The absence of calculated disturbances in the presence of suction is again probably due to the turbulence level in the wind tunnel which caused considerably higher suction quantities to be applied during tests than were originally estimated.

_.2.7 Conclusions - Wind Tunnel Testing - The wind tunnel testing has demonstrated dramatically the aerodynamic practicality of achieving laminar flow control using distributed suction on a representative wing section with a 30 degrees sweep angle and at an Re per unit length close to flight conditions.

The results of this test also show that a production quality surface and structural considerations do not preclude establishing and sustaining a laminar b(Sundary layer flow.

The test data indicates that the convenient and relatively simple X-21 boundary layer stability criteria provide a useful means of estimating suction requi rement s.

The latest progress in electron beam perforation technology has provided a very good surface material for practical laminar flow control.

OR,G'_ : ":' [_ OF POOR (,?J,';LITY :°3 RE=ERENCES - AERODYNAMIC TESTING ._ss _ ' : Review of _n_eqral-Equation Techniques for Solvi_ _.° _c':er:t_al F]o_ Problems w_Lr, Emphasis on tne Surface-Source MethoC.

S(,,_,_ter ,_e:r..;:,ss in App!iec Mechanics and Engineering, 5(i975) :.. ::-Z :_edm_:, L_.'<.- .:i. Tn_ee_Dimension_ 1 L_f:ing Potential ;low Program.

,'_._r_'l._..,,.. :'cuc_aS P, epor:. MDC JSZc2/0!. September i_74 t_/¸_'_ _ 9.0 STRUCTURES .r.

9.1 WING BOX DESIGN For structural efficiency, the LFC surface and suction requirements were inte- grated with structural design to obtain the maximum effective structural depth within the aerodynamic envelope. The initial aerodynamic definition was for suction back to 70 percent chord on both upper and lower surfaces.

9.1.1 General Requirements Design Criteria: o Manuever load factor = 2.5g o Design life = 60,000 hours (crack free). Fatigue scatter factor = 2.

o Advanced composite material usage and structural technology compatible with 1990-95.

Structural Requirements: Structure to provide stiff continuous load paths at or near the outer surface_ to be fatigue, fracture and damage resistant, environmentally durable and easy to maintain. The requirement for suction airflow through the surface introduced a new factor into considerations of strength, stiffness and environmental resistance.

Aerodynamic requirements: An extremely smooth wing surface, uniformly porous or slotted with varying suction velocities and close control of pressure drop tolerances, Suction System requirements: Maintenance of the required airflow in the main trunk ducts with minimum losses, avoiding right angle turns and constrictions, Metering and control of suction airflow from the surface to the suction pump.

In addition to the integrated discipline requirements, there were four primary structural design questions to resolve: I) Porous or slotted surface?

2) Integral or gloved LFC surface?

3) Composite or metallic structure?

4) Spanwise or chordwise surface air collection?

2o5 z- _.

9.1.2 Initial Concepts Considered Initially, twelve concepts were considered and each was assigned a code letter, see Table 9-I. Three wing box materials(composite, aluminum, titanium), porous and slotted surfaces, and a removable glove panel versus an integral LFC surface were the alternatives considered.

Six variations in the suction surface structural design were also considered, as illustrated in Figure 9-I and described below: A. CORRUGATED, with air collection spanwise in the flutes.

B. CONVENTIONAL (X-21) and EXTERNAL BLADE (if gloved), with spanwise air collection between the stiffeners.

C. ARCH-WEB, rib support is distributed into a continuous web which separates the fuel tank from the chordwise air collection. The structural cover can be an open grid as shown, a honeycomb sandwich, or a monocoque shell depending on the spacing of the chordwise formers.

D. ISOGRID-STRINGER, an open grid, forms the primary structure through which air is collected spanwise between the stringers. The lower sheet is a fuel pressure barrier and seal.

E. HONEYCOMB, with slotted metallic or "peel ply" porous outer surface.

Air is drawn spanwise for short distances within the core and then ducted through the fuel tank to larger ducts in the spar areas.

F. INTERNAL BLADE, (or conventional stringer-stiffened), in which air is collected through a surface panel to a ducting layer entirely outside the primary structure.

with a 1990-1995 technology frame of reference, there should be sufficient time for the development of new and original types of structures, if selected.

®

OF POOR QUALITY / TABLE 9-I. INITIAL EVALUATION MATRIX PANEL INTEGRAL (i) GLOVE (g) STRUCTURE POROUS (p) SLOTS (s) SLOTS (s) POROUS (p)

(c) A A

Compos i te A Acis cip cgp cgs A1 umi num

(a) A A A

A ais aip agp ags Titanium

(t)

Atis Atip Atgp Atgs Surface mechanically Surface mechanically fastened from inside fastened from outside.

or permanently fastened.

Fab/Assy Scheme II Fab/Assy Scheme I The question of a non-removable LFC surface versus removable LFC panels is especially important from an inspection and maintenance standpoint. Integral ducting and integral surfaces could make structural inspection and repair difficult. An early decision, influenced by uncertainty of the durability of the suction surface \c,._ at that time, was to favor removable LFC glove panels fastened from the outside, to minimize service maintenance costs. It was recognized this would entail fastener smoothness problems and some weight penalty which would be offset by the aircraft performance improvement due to laminar flow and the advantage of having a more easily maintained LFC structure.

Table 9-2 shows some advantages and disadvanta, ges of glove (porous) and integral (slotted) systems. Two separate trade studies, glove versus integral and slots versus porous, were linked in this evaluation by the following rationale: Precision slots must be cut in a metallic material which by virtue of its stiffness attracts load. Slotted panels should therefore be integral with primary structure and fastened accordingly. By contrast, porous materials tend to be of lower stiffness and be less highly loaded. Comparing porous and slotted arrangements, it was concluded that the disadvantages of the slotted systems were more difficult to overcome than those of a porous glove. The listed disadvantages for a porous glove were therefore regarded as areas for design improvement, especially load acceptance. For trade purposes, the porous glove was considered non-structural but attached so as to strain with underlying structure.

TABLE9-2

GLOVE VERSUS INTEGRAL LFC SURFACE

, Glove- Porous Panel Disadvantages Advantages o Must strain with structure o Removable for inspection and repair o Glove structure less efficient o Low sensitivity to variations in po ro s i ty o Possible clogging and cleaning o Fewer joints and discontinuities to fair problem o Fatigue resistance uncertain at o Fuel leak seals separable this stage.

o Potential lower fabrication cost o External fasteners must present a smooth surface.

° Integral - Slotted Surface o Efficient structure o High cost, precision construction o Fine slots difficult to machine in corrosion resistant materials o Sensitive to suction variations o Difficult to inspect for cracks and corrosion in substructure and integral ducts.

o Difficult to maintain and repair o Slots require stiff metallic materials highly stressed integral design required.

o Difficult to control slot width under structural loading.

9.1.3 Preliminary Design for Strength A preliminary wing geometry with an aspect ratio of 12 and a 381 m2 (4100 ft 2) area, as shown previously in Figure 5-2a, was used to generate loads for sizing the structural sections of the candidate designs. The baseline structural concept selected initially for modeling to determine loads, was a corrugated graphite epoxy wing skin covered with a porous glove panel, concept A, as shown in Figure 9-2. A later version is shown in Figures 9-3a and b. The latter provided spanwise flow channels closely matching desirable structural sizes. The resulting wing shear and moment diagrams are shown in Figure 9-4.

O'"'._:.,'.i i ::-: i:__ OF POU,-_ QUALi"_ AIR INTEGRAL GLOVE COLLECTION SLOTS-_ _ _ r_ SPANWISE A _-- , .... T __ ' SPANWISE (x-2z) /-SLOTS OR POROUS CHORDWISE _SLOTS OR POROUS SPANWISE I PEEL _":<_ :_ ...................

CHORDWISE POROUS- 7 CHORDWISE ,.(----_ • _.. ..... _ ......... _....

• << _/ .....

7-GEN.22075"| FIGURE 9-1. LFC STRUCTURAL DESIGN CONCEPTS STRUCTURAL SCHEMATIC -- TYPICAL UPPER/LOWER SURFACE II1_1EN-2251 | FIGURE 9-2. INITIAL BASELINE STRUCTURAL CONCEPT ORIGt_'i_:_]- ,;r,,_,-., .-,_,.. IS OF POOR QUALITY • GLOVE CONCEPT • CORRUGATED PRIMARY STRUCTURE • INTEGRATED SPANWlSE AIR COLLECTION • CONSTANT NUMBER OF SPANWlSE FLOW CHANNELS • SEE FIGURE 9-3b FOR SECTIONS INDICATED FS C E CORRUGATION _ SUCTION ENGINE RS SUCTION ENGINE FIGURE9-3a. BASELINE WING STRUCTURE SECTION A-A STATION 200 I--POROUS (OR SLOTTED) PANEL , r'-METERING SHEET _ F--OPEN STRUCTURAL GRID

- i i t

CAP MATERIAL,--_._._ i _- -4 _- ..... --_ _+ 45, 90-DEG O-DEG FIBERS _'___ ---_-- -_J SECTION B-B _----- - ------J FIBERS STATION 600 ! _Y_:c__.-_,iECTION C-C\_---_-_--_--JJ ' 7-GEN-22755.| STATION 887

C_

FIGURE 9-3b. CORRUGATED STRUCTURE (COMPOSITE) • .

Two critical wing load conditions were identified, both gust conditions. The low wing loading of 4.69 kPa (98 psf) made the large wing gust critical. Two fuel conditions were included in the analysis. With fuel outboard maximum bending relief was obtained, resulting in 20-30 percent less bending moment in the inboard wing region. The 3.24g gust response with minimum fuel in the wing produced 40 percent higher shears in the outboard wing and this sized the spar webs.

The bending material effective spread thickness (_) requirements were generated initially on an Mc/l basis only. This factor (_) was the significant parameter in the wing bending material requirements used for concept comparison. The bending load intensity peaked at approximately 40 percent span, Figure 9-5, because of the aerodynamic break in planform and thickness controlled by the thickness-to- chord ratio; see Figure 9-6.

The loads and sizing procedures were also run for an aspect ratio 14 geometry for comparison with the initial baseline AR of 12. This produced an even higher peak in the bending load intensity (Figure 9-5). The higher load intensities for f" • r, the LFC wing were due to aerodynamic requirements for reduced thickness/chord ratios, Figure 9-6.

9.1._ Flutter Penalty A preliminary flutter analysis was run using the initial baseline wing bending (El) and torsional (GJ) stiffnesses for both aspect ratios. The modal representation used in the flutter analysis consisted of three rigid body modes (symmetric and antisymmetric), five uncoupled wing bending modes, five uncoupled wing torsion modes, and one quasi wing-roll mode. This allowed independent variation of bending and torsion rigidities, which is feasible with composite material construction due to directional characteristics of the weave. The strength-only design rigidities were used initially and the analysis of flutter speeds versus percent fuel showed that the most critical loading was with full fuel tanks although even zero fuel produced a flutter speed that was less than the 1.2V D requirement. Substantial increases in wing stiffness were indicated to meet the flutter requirement, see Figure 9-7.

_i_ ¸, OF POOR QU_L;IY N-m (LB-IN.)

(KIPS)

(200)

(400)

kN

2O COMPOSITE OF TWO GUST CONDITIONS Z

1500 -

o_ ,-I (150) .J o: (300)- m MJ =E Z I (/3 ..J < 1000- Z (11:10) uJ o 1200) p- _----- VERTICALMOMENT C¢ IJJ ,.J 500 (100) (50) _O P- MJ 0 0 O- 0 (0 (500) (1000) (1300) I I I 0 10 20 30 SEMISPAN STATION- meters (IN,) FIGURE 9-4. WING MOMENTS AND SHEARS t_ -'L ,..

) (KIPS/IN.

kN/mm 6 CONDmON: 2.74-GGUST WITH OUTBOARDFUEL

(4°) I co=Posn coR.u= o s

AR14 (30)1 LOAD INTENSITY 0 O_ o (,5o0) (looo) (t3oo) IN.

I I I I 0 10 20 30 m SEMISPAN STATION =="'=' FIGURE 9-5. WING PANEL BENDING LOAD INTENSITY - SPANWISE VARIATION AC/4 "-- 30 DEGREES MCRUISE -- 03 OF PO0_' QUALITY l- Z 14J r,* % r, 14 I LAMINAR FLOW (MODEL 3128) / _o CtL DESIIGN = 0.50 % I ¢Y ",, TURBULENT (MODEL 3127) e_ Og CL I O t _)ESlGN _0.64 ul Z X m Z I--

.J 10 I

t/c BASEDON TOTALCHORD INCLUDING "YEHUDI" < O .J 0 10 20 30 40 50 60 70 80 9O 100 7 -GEN*22647- | PERCENT SEMISPAN FIGURE 9-6. PRELIMINARY WING THICKNESS DEFINITION (KEAS (550) m/s ..- _AR14. 250% STRENGTH DESIGN 6.0 Hz 6.0 Hz (450) - (FLUI"rER SPEED REQUIRED) N_ AR12, 220% STRENGTH DESK; FLUTTER SPEED 200 I (350) AR12, STRENGTH DESIGN 4.1 Hz 4.6 Hz (250) I I I !

0 20 40 60 80 PERCENT FUEL 7-GEN-227711"1 FLUTTER SPEED VERSUS PERCENT FUEL - INITIAL BASELINE FIGURE 9-7.

STRENGTH DESIGNED WING Increasing the bending stiffness (El) over the strength designed values was found to have no effect on the flutter speed.

The required flutter speed was therefore attained by increasing the wing torsional rigidity (GJ), assuming constant percentage increases over the entire span. A factor of 2.2 times the strength-designed "GJ" values was required for the aspect ratio 12 baseline wing, and 2.4 times for an AR = 14 wing. Future stiffness tailoring along the span could significantly reduce the material required to meet flutter speed, however this analysis highlighted the effect of high aspect ratio on flutter for a 30 ° swept wing. Metallic wing designs, with a fixed E/G ratio, would incur larger penalties in meeting flutter stiffness requirements than graphite epoxy designs in which ply direction tailoring can be used to augment either El (bending stiffness - 0 ° fibers) or GJ (twisting stiffness - 45 ° fibers) to give minimum weight penalty, see Figure 9-8. The resulting initial wing stiffnesses adjusted for flutter, but before aeroelastic analysis, are presented in Table 9-3.

9.1.5 Evaluation of Initial Concepts Design sketches were generated for the structural concepts considered. Three of the concepts (arch-web, isogrid and peel-ply) were discarded early for various reasons as discussed in the following paragraphs: 0 Concept C, Arch-Web (Figure 9-9) was initially attractive since it promised a very low LFC weight penalty and is geometrically shaped to favor easy integration of chordwise airflow collection and suction air/fuel separation with an efficient structure. The outer structural shell could be slotted honeycomb sandwich or monocoque. For a porous outer suction panel, the structural shell could be of isogrid or sandwich construction. Drawbacks of this design are (I) possible buckling instability for a continuous rib-' supported wing cover that would be difficult to analyze, (2) lack of fuel slosh baffles (weight penalty) and (3) difficult fab/assembly sequence.

This design could still be attractive for a fuel-free wing or stabilizer.

A Y

OF POOR QUALITY

TABLE 9-3 INITIAL BASELINE WING STIFFNESS DISTRIBUTION (Strength Designed Wing Stiffened For Flutter Only) AR 12 WING - 381 m2 (4,100 ft 2) GRAPHITE EPOXY STRUCTURE El DESIGNED FOR STRENGTH GJ DESIGNED FOR FLUTTER* MNm2 Billion Lb-ln 2 MNm 2 Billion Lb-ln 2 Sta 200 4178 1456 3814 1329 2936 1023 2594 904 400 1940 676 1685 587 1188 414 IIII 387 542 189 600 720 251 700 465 162 336 117 281 98 227 79 1 000 75 26 52 18 AR 14 WING - 381 m2 (4,100 ft 2) GRAPHITE EPOXY STRUCTURE 200 4729 1648 4698 1637 296 3472 1210 3340 1164 4O0 2376 828 2175 758 1 570 547 5O0 1 550 540 309 71 7 250 600 887 7O0 591 206 453 1 58 800 385 1 34 304 106 1 000 1 32 46 98 1 2OO 2 3 8 17 6 *2.2 GJ factor over strength design for AR 12 2.4 GJ factor over strength design for AR 14 ORIGINAL PAOI_ i_ OF POOR QUALITY m/s (KEAS) ( (550) GJ REQ'D FLUTrER SPEEDREQUIRED 1.2 VD / FLUTTER (450) SPEED J (350)

I

WING-- 381m2 (4100 F'r'2) CORRUGATED GRAPHITE-EPOXY STRUCTURE (250) 0 0 40 60 80 100 120 140 0 2O PERCENT GJ INCREASE OVER STRENGTH DESIGNED VALUES (CONSTANT El) 7-GEN-22720* 1 FIGURE 9-8. WING TORSIONAL STIFFNESS INCREASE REQUIRED FOR FLUTTER OR SOLID MONOCOQUE A _ R.S.

F.S.

A L AIR TANK WALL AND DISTRIBUTED RIB WEB WING BOX BULKHEADSAT

' 'f' I '

3.81-m (150-1N.) PITCH

iiiiii iliiii i

SECTIONA-A ,._E,.,,.,,_ SLOT (TYPICAL) ARCHWEB-STIFFENED CONCEPT - CHORDWISE COLLECTION -- SLOTTED FIGURE 9-9.

LFC SURFACE Concept D, Isogrid-Stringer, Figure 9-10. The isogrid concept studied was of aluminum construction although it would be feasible in graphite epoxy.

The isogrid was sized to take an estimated 945.7 kN/m (5400 Ib/in) shear load only, since the spanwise grid bars would otherwise be required to take an estimated 4.378 MN/m (25,000 Ib/in) bending load intensity and would be very large, penalizing the chordwise grid direction. Stringers were therefore added to take the major portion of the bending load. The spaces between the added stringer were used for air collection and a 1.27 mm (.050 in) diaphram acted as fuel and pressure barrier. The section weight of 51 .3 kg/m 2 (10.5 psf) compared unfavorably with the 31.7 - 34.2 kg/m 2 (6.5-7 psf) weight then being obtained for the graphite and/or titanium corrugations and honeycomb sandwich concepts.

Peel Ply Concept, Figure 9-11. This was a derivative of the corrugated concept "A", except for incorporating an unusual semi-flexible, three- dimensionally porous blanket over air collection holes in the structural skin.

This promised to be one of the lightest designs, but cursory materials work with controlled flow resin-impregnated Scott foam under Dynapore surfaces failed to disclose a feasible materials combination for the porous blanket.

The objective of the "_eel-ply" concept was to eliminate mechanical fasteners in the surface yet provide a readily removable surface for cleaning or replacement.

Weight numbers were obtained for other concepts judged to be feasible (Concepts A, B, E, F in Table 9-4). A multi-station analysis using the loads and stiffnesses described previously, provided bending and shear material requirements for the seven remaining concepts. Table 9-4 compares weight efficiency results at the 45-percent semi-span station. This station coincided with peak loading intensity, however an analysis using a larger matrix of stations indicated the same general trends.

_c i

)

HOLES IN FLOW REMOVABLE EB PEF 1 METERING SHEET POROUS SURFACE "/FUEL BARRIER DUCT POROUS-- PANEL FASTENER FIGURE 9-10. ISOGRID STRINGER CONCEPT -'_ .. J rCOMPARTMENTATION _ r(AS SHOWN) ;:.-) >< _:.- _ f_ll_/ _.;q_ J ,. /_"f- _..,ee'_B_ 'r_ \ _----POROUS FOAM CORE _,\ .j _ v 11" _,,_r.._J_tB_ \ \ TUBULAR SHEET ADHESIVE , ='SUCTIO -J _ _ "_\ / / _ - N DUCT FLOW 7-GIrN-22719-| FIGURE 9-11. PEEL-PLY CONCEPT OF,,iCii_,._j , .......

. _ OF POOR QUALiTy TABLE 9-4 / .

PRELIMINARY STRUCTURAL DESIGN ASSESSMENT WING BOX COVER INCLUDING LFC SURFACE AT 45 PERCENT SEMISPAN FOR STRENGTH AND FLUTTER STIFFNESS WEIGHT OF UPPER COVER BOX STIFFNESS RELATIVE GJ-MNm 2 EI'MNm2 kg/m 2 (LB/FT 2) WEIGHTS (LB.IN. 2 (LB.IN. 2 PER UNIT STRUCTURAL CONCEPT BILLIONS) BILLIONS) STRUCT GLOVE TOTAL AREA A, CORRUGATION 549.9 694.5 46.29 8.93 65.22 A©ip, COMPOSITE - POROUS (242.0) (9.48) (1.83) ( 11.31 ) <_1.52 (191.6) 549.9 Acts, COMPOSITE - SLOTTED 694.5 46.29 0 46.29 1.28 (191.6) (242.0) (9.48( (9.48) 483.3 744.7 72.11 8.93 81.05 2.24 Atlp, TITANIUM - POROUS (259.5) (14.77) (1.83) (16.60) (168.4) EXTERNAL BLADE II II 11 553.0 639.1 44.09 10.35 54.44 _ 1.5 Bcip, COMPOSITE - POROUS (192.7) (222.7) (9.03) (2.12) (11.15) E. HONEYCOMB SANDWICH COMPOSITE - Ti SLOTS _j_ [_LU_I_L 688.3 721.8 Eels, 36.23 0 36.23 1.0 (206.0) (251.5) (7.42) (7.42) F. INTERNAL BLADE" 543.0 I I I I 920.1 95.01 11.42 106.43 2.94** FDIm, ALUMINUM - POROUS (189.2) (320.8) (19.46) (2.34) (21.80} 541.2 646.3 42.04 11.42 53.46 F©ip, COMPOSITE - POROUS (225.2) (8.61) (2.34) (10.95) _1"47"° (188.6) GLOVE TRADE STUDY COMBINATIONS SELECTED FOR FURTHER STUDY "'RELATIVE OVERALL WEIGHT WOULD BE LESS DUE TO SMALLER CHORD MAIN BOX WITH CHORDWISE AIR COLLECTION 7-GEN-22757-| Each design was brought to the same level of required stiffness for flutter. The metallic designs incurred greater stiffening weight penalties than the composites since, in meeting GL] requirements, they acquired excess El material due to the fixed E/G ratio of metallic isotropic materials. The titanium corrugated primary structure was abandoned because it was one of the heaviest designs considered.

The aluminum internal blade design that survived earlier screenings was the heaviest, due to the added stiffness penalty. This configuration was closest to conventional wing design.

As expected, the designs with structural material closer to the outer surface were lighter. The honeycomb design illustrated in Figure 9-12, although shown to be the lightest, would be heavier than indicated when the weight of air collection through the fuel tank and joints and attachments are considered; it was not considered further however because Douglas design philosophy TOr honeycomb structures* is to avoid its use for primary structure. The three glove panel designs with graphite epoxy primary structure are approximately equal in weight when glove panel weights are included. Of these three (corrugated, external blade and internal blade), either of the blade-stiffened designs would be preferred on a basis of probable cost. System weights on a whole-wing basis, including the weight of additional ducting, and relative efficiencies of the spanwise versus chordwise air collection schemes were not considered in this preliminary analysis.

*Honeycomb Design Philosophy. Douglas design philosophy allows the use of honeycomb material for non-structural components, lower life fairings, and replaceable control surfaces if overall evaluation shows it to be advantageous. The reasons for avoiding honeycomb construction for primary structure are: It is extremely vulnerable to moisture absorption, corrosion (if aluminum) and delamination. This would apply particularly to an integral air collection LFC honeycomb design where atmospheric moisture is drawn directly into or through the core.

0 It is vulnerable to fuel absorption particularly if the air collection pipes enter the fuel bays through the inner surfaces of the honeycomb sandwich.

o Attachments and joints are complicated, leading to higher cost - much of the apparent weight advantage may be lost due to heavier joints.

o Repair is difficult - especially for highly-loaded composite/honeycomb panels.

o Airline resistance to honeycomb construction is already very strong - due to experience of premature failures in service and costly maintenance procedures.

9.1 .6 Aeroelastic Penalty The stiffness distributions derived for the initial baseline configuration AR 12 and 14 wings, considering strength and flutter requirements only (Table I) were subjected to an aeroelastic analysis for roll control system effectiveness. Both wings were found to be deficient because of excessive bending and twisting deflection leading to loss of aileron effectiveness at required speeds. Initially, an aeroelastic stiffening factor of 1.5 was applied uniformly spanwise to both E1 and GJ of the i graphite epoxy structures of Table 9-4. The subsequent multi-station analysis indicated additional stiffness increase would be required to meet the roll stiffness criteria. High modulus GY-70 fiber was therefore introduced in place of T-300 (high strength) fiber material in the _+45° plies. This resulted in reduced shear material weight with increased shear stiffness. The T-300 material was retained for the 0 ° spanwise fibers because of uncertainty about the ultimate strain allowable for GY-70 fibers at this time. The resulting weights for AR 12 are listed in Table 9-5, with an all T-300 weight calculation for comparison. The results clearly indicate the advantage of using GY-70 fibers. The corrugated main box design was the heaviest and the two blade stiffened design weights were not significantly different.

9.1.7 Concept Selection.

The weight of the external blade design was slightly higher than for the internal blade (Table 9-5) but provided the advantage of integral spanwise ducting of increased area compared with the corrugated design. The corrugated design would FkT '_i_i _ _ .i_i ,...... ' QJ r'_ _-- _ 4-) "_ _- 0 _ I_ I._ _- •r- 0 0 E e,J _ _.0 0 I _ r-- I_ '_ I_ ._ kid I-_ LO 0 4-) I \ ÷ I'-- _-- O0 L_ UJ h- n- _I I.I- °r- _ "1- O •_- _J O _ Oh oO LLJ I-- ¢._ C) tO O_ O0 ¢0 --J 0 0 0 _oO E I-- 0 _J 0") n,- Q; LI..I -I" O t_ I U3 • r- _ I. %- I--" °_- E U.J I.L I J- e- QJ e" cO E 0 0 QJ u_ QJ rs 14- _0 ,-- 4-J °l-- e- 4-J _- -I-J _" O_ .,-. --_ e'_ _P O CJ tJ _- 0 "_ __ _..

E _- O O 4-_ C.J _.- _ QJ C.JO0 O X _- f_ also incur additional weight penalties because of its inefficiency in carrying _J chordwise concentrated loads. Another factor against the corrugated concept was that aft of about 65 percent chord the corrugations were not big enough for the increased suction airflow collection, by a factor of two.

An early external blade design sized only for strength and flutter stiffness is shown in Figure 9-13. The exterior paneling was considered at this stage to be fastened mechanically to a glove panel support grid. The grid was then fastened to the split blade stiffened cover below. Flanges were not used to stiffen the blades in order to provide the maximum surface area for porosity; the grid effectively provided both blade stabilization and an outer structural cap. Elements of the trailing edge control system could be readily attached with this design as shown in Figure 9-14, due to the continuous load path provided by the inner structural skin. Rib attachment to the main panel is simplified by the absence of internal stiffeners.

The internal blade structure is shown in Figure 9-15, which also illustrates an option of chordwise air collection. The blades are parallel to each other and to the rear spar, and the wing ribs are attached using separate shear clips as in a conventional wing. Suction airflow from the surface passes through a grid, supported outside the primary wing skin, on chordwise standoff strips. The standoff strips act as chordwise "cross stiffening" and as air flow dividers.

The porous outer panel is attached to the grid and can be removed separately.

With chordwise collection, the spar location must be moved aft (from,15 chord to .19 chord) to provide space for ducting in the leading edge.

The Internal Blade design is similar to a conventional wing structure with respect to trailing edge control system attachment, Figure 9-16. Air was not ducted spanwise aft of the rear spar with the chordwise collection system because of interference with control system bracketry, Thus both upper and lower surface air is collected forward into the leading edge region where it is ducted spanwise before being ducted into the dry bay at the suction engine. It should be noted that at this stage of conceptual design, LFC suction was being used on both upper and lower surfaces to .7 chord, PHASE-OUT IN THE TRAPEZOIDAL SHAPED AREA CORE 12.8 kgJm 3 1.52 mm (0.06 IN.) DIA HOLES WEDGE OF (8 LB/FT 3) 6.35 mm (0.25 IN.) PITCH PSU EDO-ISOTROPIC G/E 100%- 0 ° -O.O03O TO 0.0070 50,8 mm t I= ± 0.0005 (2.0 IN.) 1 SLOT CUT AFTER -9 mm 120 Ti OUTER SKIN CONST ' ASSEMBLY (0.36 IN.)

-10 mm I (0.39 IN.)

AIRFLOW mm (0.25 IN.)

INNER SKIN IS PSEUDO-ISOTROPIC G/E CROSS FLOW DUCTS 889 mm (35,0 IN.) APART 12.7 mm (0.50 IN.} ELLIPTICAL HOLE

MOLDED THROUGH t

COMPOSITE FIGURE 9-12. HONEYCOMB CONCEPT LOCAL POTTING_TION POROUSPANEL FOPEN STRUCTURAL GRID

METERING SHEET _ II I

AND DUCT WALL _ U t, STATION 600 STATION 887 7-GEN-2283SA FIGURE 9-13. EXTERNAL BLADE STIFFENER CONCEPT AILERON ATTACHMENT SECTION : +_ +_ SPOILER SUPPORT FLAP SUPPORT STATION 7-G_'N-22930 FIGURE 9-14. CONTROL SYSTEM ATTACHMENT -- EXTERNAL STIFFENER BLADE CONCEPT PLATE/GRID POROUS G LOVE-_ D I COLLECTORS

U L)

A-A PLAN VIEW FIGURE 9-15. INTERNAL BLADE STIFFENER CONCEPT - CHORDWISE AIR COLLECTION The preliminary wing weight comparison of Table 9-6 clearly shows the corrugated structure to be heavier. The trade study was therefore reduced to internal versus external blade configurations and chordwise versus spanwise air collection. A comparison of the blade stiffened designs with only one suction pump/engine installation per side is shown in Table 9-7. In spite of the increased duct size required in the leading edge and the further aft location of the front spar, the internal blade stiffened wing was lighter. With the weight difference being small, the decision was made to proceed with the External Blade stiffening at this stage because of its inherent spanwise air collection capability, but the internal blade stiffened structure would be more efficient if a chordwise suction airflow collection system were finally selected.

9.1.8 Aspect Ratio Study In order to select the LFC wing planform and thickness ratio, parametric data concerning the effects of aspect ratio and thickness on wing structural weights were produced for use in the aspect ratio trade study discussed in Section 7.3.

The general LFC wing concept selected for study was of unconventional material (graphite/epoxy composite) and of unconventional structural configuration (gloved- external stringer); there was therefore no reliable data base for estimating wing weights. A special design/analysis was required making use of conventional analytic tools for strength and stiffness analysis appropriately adjusted for aeometry and materials properties. Three cases were carried in parallel for wing aspect ratios I0, 12, and 14. The starting planform family was derived from the initial baseline LFC configuration with suction on both surfaces, however, the wing area was updated to 334 m2 (3600 ft 2) from the original 381 m2 (4100 ft2).

The general procedure used to establish material thickness and volume requirements for weight estimation of each of the three wing geometries was as follows: a. Define design load conditions. Again, the load factor Nz for gust exceeded the 2.5g maneuver factor by a small amount.

b. Define external design loads and l.Og twist requirement.

t TABLE 9-6 WING WEIGHT COMPARISON CORRUGATED VS EXTERNAL BLADE STIFFENING External Blades Corrugated Panels kg (Lb) kg (Lb) 12,796 (28,210) 13,288 (29,163) Bending Material LFC Box Panels 4,480 (9,876) 4,480 (9,876) L.E. With Ducts 877 (I ,934) 877 (I ,934) 1,104 (2,434) 1,104 (2,434) Engine/Pump Assemblies* 231 (510) 231 (510) Engine/Pump Nacelles

413 (910) 413 (91o)

Ducting to Pumps 10,499 (23,147) 10,282 (22,668) Additional Weight (Ribs, Webs, Flaps, Etc.)

Total 30,832 (67,974) 30,183 (66,542) *Two Suction Engines per Side TABLE 9-7 WING WEIGHT COMPARISON BLADE STIFFENED STRUCTURES Internal Blades External Blades kg (Lb) kg (Lb) 12,796 (28,210) 12,325 (27,172) Bending Material LFC Box Panels 4,1 54 (9,1 58) 4,480 (9,876) L.E. With Ducts 877 (I ,934) l ,llO (2,448) 683 (l ,506) 683 (I ,506) Engine/Pump Assemblies* Il I (244) III (244) Engine/Pump Nacelles 409 (902) 409 (902) Ducting to Pumps I0,503 (23,155) 10,282 (22,668) Addi:tional Wing Weight (Ribs, Webs, Flaps, Etc) Total 29,638 (65,340) 29,295 (64,585) *One Suction Engine per Side Establish a structural model geometry and materials allowables for C • each aspect ratio study wing.

Because of earlier experience on the preliminary baseline (see Section 9.1 .6), GY-70 graphite fiber was used at the outset for 345 ° plies, but the spanwise fibers remained T-300. The ultimate strain allowable was set at 0.004 for spanwise strains.

d. Establish GJ and El values distributed along each wing span to meet external loads strength requirements and a l.Og aerodynamic twist requirement.

At this stage it was noted that for the aspect ratio 14 wing, the high shear stiffness provided by the GY-70 fibers was necessary in the outboard wing merely to meet the Ig aeroelastic twist requirement for obtaining the desired spanwise lift distribution.

e, Establish the weight distribution of each wing represented, as lumped masses along the elastic axis, for the flutter analyses.

f, Calculate the spanwise lift distributions and the aerodynamic centers of pressure.

Conduct flutter analyses for the strength designed wing and establish , flutter margins and the effect of stiffness variations• (The wings were flutter free without additional material at this point.)

Conduct an aeroelastic analysis for each of the three wings using h ° stiffness distributions established by the strength and flutter analyses.

Revise GJ and El distributions per aeroelastic analysis and calculate the j • distribution (bending material equivalent thickness) on which to base wing weights• An aeroelastic stiffening for each aspect ratio was found necessary, see Figures 9-17 and 9-18.

P

-_- ACCESS---------

SECTIONAT AILERONACTUATOR

OF POOR QuALiTy

TYPICAL SPOILER ACTUATOR TYPICAL FLAP SUPPORT STATION 7-GEN-22929 FIGURE9-16. CONTROL SYSTEM ATTACHMENT - INTERNAL BLADE STIFFENER CONCEPT 3.O 2.5 RELATIVE WING 2.0 STIFFNESS AND WEIGHT 1.5 WING WTsTI FFNES S WING WTSTR ENGTH _..,.._ 1.0 11 12 13 14 9 10 ASPECT RATIO Ik3Ef¢_ FIGURE 9-17. INCREASED STIFFNESS REQUIRED VERSUS ASPECT RATIO k. Calculate wing weights for AR = I0, 12, 14.

l ° Vary t/c of airfoil and calculate corresponding variations of to meet final strength and stiffness levels. This is subsequently used to define wing weight variations as a function of t/c, see Figure 9-19.

m. Provide weights data for the aerodynamic performance analysis with suction on both wing surfaces.

Establish t for the "upper surface suction only" configuration to meet n • the stiffness rqu-irements, as above.

o. Calculate t variations for airfoil t/c variations.

Calculate weights and provide input to aero-performance analysis for p.

comparison in "suction on both surfaces" versus "upper surface only" trade study_ 9.1.9 Revised Winq Design Details After the sizing and aspect ratio study, wing aspect ratio lO and 311 m2 (3350 ft 2) area were selected for the baseline LFC aircraft, with suction on both surfaces to 70 percent chord. Loads and stiffness requirements were then extrapolated for the new wing. Scope drawings of the wing box structure were prepared. Figure 9-20 shows the planform and a typical rib and Figure 9-21 shows structural details of the wing upper surface at the suction engine dry bay station• Elliptical holes were used to reduce spanwise load stress concentration in the wing skin and allow smooth air collection from each stringer bay. The suction engine support is a box beam cantilever extending from the rear spar. Figure 9-22 shows the lower wing cover at the suction engine. The large hole aft is for routing leading edge and upper surface suction air to the suction pump below the wing. The two holes forward are access door locations and the smaller holes aft are to route air from the lower triangular duct aft of the rear spar. The truss web construction was used to provide the additional ducting area required for the increased suction airflow aft of approximately 63 percent chord.

O _")Tt.'_. ln_,..

OF POOR Qb',qLiTY 2.0 ,_R = 10 = [ - -1.35, _.,,a=e- STIFFNESS STIFFNESS FACTOR, FACTOR, I N N I I 0.33 0.90 0.90 I t I I I I I I J J i 0 0 0.2 00 02 0.4 0.6 0.8 1.0 0.8 1.0 4.0 AR = 14=

I

STIFFNESS - 2.43- FACTOR, 2.0 _I

i\

I I -'r-" i I 0.90 0.33 I 0 L 0 0.2. OA 0.6 0.8 1.0 B-GEN.Z2(II9 FIGURE 9-18. SPANWlSE VARIATION OF INCREASED STIFFNESS 12 14 ASPECT RATIO T/C 0.09 I 0.11 [ 0.14 0.0910.11 [ 0.14 VALUES FOR ??0.32 (770.75 SIMILAR PROPORTIONS) m t 2- (IN) ---STRENGTH-- --STI FFN ESS -_'_". _ _ B-GEN-22597 FIGURE 9-19. EFFECT OF THICKNESS/CHORD RATIO ON INCREASED STIFFNESS REQUIRED OF POOR QUALITY ,_F-'- POR0 US PANEL . ! ..... I ....

N AA DRAINHOLES _--" w - " "" ' t :-_ 2

F_ON_ S,ARPLANE_ \___'_--_ / I

• CONVERGENT STRINGERS LOWER--'-"'t lJ_-:-S-'_-'J_ I I • SINGLE SUCTION ENGINE wl _ I ,_ A ,,,_//. j:;_:_-" "_ / • ACCESS I <" _./t . :. :,'_;!:i_' _- _ .... ,I • SUCTION TO 0.70C | \/,.<_.. .. .::/:..-:_;rL_-. - ________. ----_ T

• EXTERNAL ST, F_ENED I/////////_:_-_"% " "_F----_ '

87 RO2 SPAR PLANE X w = 464.1 8C, EN-_.,L Xw = 1097.524 FIGURE 9-20. STRUCTURAL ARRANGEMENT - SUCTION THROUGH BOTH SURFACES _! // VIEW LOOKING DOWN AT UPPER SURFACE (GLOVE PANEL NOT SHOWN) FIGURE 9-21.

STRUCTURAL ARRANGEMENT -- UPPER WING BOX PANEL AT _. OF SUCTION PUMP SYSTEM A typical rib cross section with upper surface suction to 85 percent chord % is shown in Figure 9-23. In this configuration, the spoiler and flap control surfaces were located aft of 85 percent chord and the truss web box extended further aft to provide ducting for the larger volume of upper surface suction air and still provide support for the control surface bracketry and actuators.

The lower wing cover was of a more conventional internal blade stiffened skin graphite/epoxy construction. The weight trade study was favorable for this upper surface suction only structure, including the porous panels, compared to the previous baseline configuration with suction on both surfaces. This was primarily due to increased effective structural depth of the wing section without lower surface air collection.

9.1.10 Wing Box Structure Conclusions 0 As expected, the wing span (aspect ratio) and wing thickness are dominant parameters in establishing structural bending material require- ments, but in all cases wing weight was influenced primarily by aero- elastic stiffness requirBnents to assure roll control effectiveness.

o Increasing bending stiffness was more effective than increasing torsional stiffness in meeting the aeroelastic requirements with minimum wing weight.

o Flutter was not critical for any of the wings considered.

GJ flutter margins indicate that further weight reduction could be achieved by optimum balancing of material characteristics to meet GJ, aeroelastic El, and Ig twist requirements. The optimum would have increased bending stiffness at the expense of torsional stiffness.

0 Duct space could also be reduced to some degree in order to increase the effective wing depth, since in this study the stiffener space allocation was never fully utilized for airflow (except at_ : .z}, AR 14). A reduction of stringer height must, of course, consider stringer/panel compression stability, but this did not appear critical for the proposed mixed modulus graphite composite structure.

LOOKING DOWN AT LOWER SURFACE DRY BAY BULKHEAD STRUCTURAL ARRANGEMENT SHOWING LOWER SURFACE AND FUEL FIGURE 9-22.

BULKHEAD AT __ OF SUCTION PUMP SYSTEM f POROUS PANEL 0.8 i C .......... _ _ .... "" .-, I I _ REAR SPAR PLANE (0.67 C) TYPICAL WING RIB - SUCTION THROUGH UPPER SURFACE ONLY TO FIGURE 9-23.

0.85 C An aluminum LFC wing of high aspect ratio with reduced effective . r, structural thickness would carry a large weight penalty compared with a graphite epoxy structure.

9.2 POROUS SURFACE SELECTION ArID LFC PANEL DESIGN 9.2.1 Initial Survey of Possible Materials & Construction The selection and development of the porous LFC panel system was dependent on suction air collection requirements, Section 7.1.3, and LFC performance as measured in the Douglas wind tunnel, Section 8.1. Only those porous materials that could meet these basic requirements were considered structurally. Initial experience with porous materials indicated that most were of low modulus and low yield strength; they were therefore not considered to be part of the primary str_cture.

in addition to strength anJ stiffness the following factors were considered: o Replaceability in case of damage.

o Porosity - overall, in strips or in geometri _- Datches.

o Smootnness of the surface and a; joints o initial imperfections and deflection under ;oad to meet wa_,iness criteria.

o Strain characteristics to be compatible with s;rain levels that could occur in the primary structure.

o Durability in operational environments o Clogging resistance and cleanability.

Some prior experience with porous panels was obtained under NASA Contract NASI-I_08 "Development of Technology for the Fabrication of Reliable Laminar Flow Control Panels on Subsonic Transports", see Reference 9.7-I. This program suggested suitable porous materials and areas needing further investigation. These included yield, fatigue strength and bondability of 316L stainless steel Dynapore porous sheet; strength of Kevlar 29/Epoxy porous Doweave laminate; resin selection, methods of controlling the porosity of a porous laminate, further investigation of methods of stiffening porous panels for LFC applications a_d clogging and cleaning of porous surfaces.

The following porous surface materials were considered initially:

o 316L Stainless Dynapore - "flicroperforated plate", 50 x 250 and 24 x llO

weaves, used for LFC surfaces.

Kevlar 29 (Kv 29) Doweavefabric (woven in "basic" weave pattern from

o

200 denier yarns) and impregnated with Cor_ar 5134 epoxy resin (250°F

_uring controlled flow resin, E.I. DuPont) used for porous layers in

the body of the panel.

o 80 x 700 Dynapore; other resins and other weaves of Kevlar and glass were

variants introduced.

LFC panel constructions considered iniLiai!'y included the following: Lockcore - a tr'Jss core construction using __ewing techniqL;es to _n: truss nodes to ,_acin_s e Figures ....

mechanically' ioc!_ _ _ " _ , se Q-2_ a,qd _-_5 Various composite non-metal]ic materials can me used, Fieure 9-26.

o Composi ÷__ iso_ri4, , an interlocW';ng., g_:_,_ tri-_×i_i qrid_ m_de., from continuous fleer '/arns and epoxy, -ig_res _-27, 9-23 and 9-29.

Honeycomb sandwicr, with porous facings, see Figure 9-30, which contrasts o isogrid and honeycomb surface suDports as revealed by reticulating adhesive bleed-through at the porous surface in pressure drop test specimens.

The isogrid and honeycomb panel concepts considered require the airflow to be drawn completely through the outer stiffening sandwich before being ducted laterally; this resulted in the need for a double sandwich glove panel or a separate collection structure under the outer sandwich. The Lockcore, on the other hand, already provides a tubular air collection structure once the air is drawn through the outer surface. This feature of LOckcore is an additional advantage.

017' -:_ " OF PC;},.;, x I _ 0.254+_0.05 mm RAD (0 375 IN ) + 9"54 mm' / \ (0.01 _ 0.002 IN.)

11 mm CORE MANDRELS _ALL NODES SEWN AND COCURED X' /I_ RESIN FILLET (TYPICAL) _// \v/'-_ \x.// Xv/; - "_""--_0.027 IN.,)(O/90) 3 WEB SECTION THROUGH CORE FIGURE 9.24 LOCK CORE BASIC PANEL DEFINITION / ,'_'; YERS , _,-i I ÷I ,-f, !

BOND

__L

/T T /)

2 _/ I/ /_,.,. / /_' / _ /

\I// ., ./ ././ /

- ' , , l//f/ t__ SEWN N E JOI_TL_ FABRIC LAMINATE _ODcocuRENSAND BOND ,G,N _,_B

(

FIGURE 9-25. LOCKCOREPANEL _._'. i_," OF POD;:,: ".:_,-,.-'_ -- 200 DENIER KV/E DOWEAVE, 10190 I' DOWEAVE NODES SEWN WITH 1/4-1N. STITCHES , 50 x 250 MESH DYNAPORE (OUTER SURFACE) STEEL 316L 0._ I ' - ............................ !FILLET (TYP) 316L STAINLESS STEEL BOLTING CLOTH NO 16 I I r _ ALL COMPONENTS ARE BONDED SIMULTANEOUSLY FIGURE 9-26. EARLY MODIFIED LOCKCORE PANEL CONCEPT VACUUM METALLIZ.ED SURFACE

T/

!

_--- KEVLAR OR GLASS-CONTINUOUS FIBER GRID L POROUS FABRIC LAMINATE ALTERNATE: 2 DYNAPORE FACES FIGURE 9-27. POROUS !SOGRID PANEL FIGURE 9-28. KEVLAR/EPOXY ISOGRIDCORE ;/ FIGURE 9-29. ISOGRID CORE SPLICE The exploratory testing included the following: o Fabrication Development o Short Beam Shear Tests o Celanese Compression Tests o Block Compression Tests o I.I.T.R.I. Tension Tests o Sandwich Beam Bending Tests o Tension Fatigue Tests o Coefficient of Thermal Expansion Test o Airflow Testing (including salt solution clogging).

The conclusions from this initial testing are summarized as follows: Of the various constructions studied, the <luted "Lockcore" porous sandwich construction was the moss promising for meeting design requirements, so a decision was made to limit further development to this type of construction.

It was not necessary to pull suction air entirely through an outer sandwich and initial collection could occur within the integral ducts of the self-stiffened Lockcore sandwich.

Although the softness for impact, low yield strength and difficult bondability of the 316L stainless steel "Dynapore" porous surface presented challenges for development into a reliable LFC panel structure, it was the leading material from an aerodynamic standpoint as indicated by the flat plate wind tunnel tests, see paragraphs 8.1 and 9.2.2.

The basic Doweave fabric woven from Kevlar fibers and produced as a porous laminate was too porous and lacked sufficient strength; and stiffness for suitability as an LFC glove panel support structure. Con- structions employing tighter weaves of glass or graphite (rather than Kevlar) fiber were recommended for overcoming these deficiencies. Kevlar was not as good as glass under impact loading with a Dynapore surface.

OF POOR QUALI'FY Clogging tests with salt (N Cl) solutions indicated that the Doweave a laminate faced with micro-perforated plate clogs relatively little and is easily cleared to its original porosity. Dynapore-faced laminates are susceptible to greater clogging and may clean to lower percentages of their original porosity depending on the original porosity. This tendency needs further investgiation with more representative contaminants.

Resin content and formulation experience led to selection of E 719 (U.S. Polymeric Co.) resin for continued porous laminate and co-cure bonding work.

Attempts to "meter down" Doweave with #120 fabric interleaved plies were unsuccessful in controlling the airflow to meet LFC requirements.

Much more work with pressure drop technique was indicated.

The conclusions from an additional test program consisting of bending tests of environmentally conditioned Dynapore-faced Lockcore panels and impact tests of various constructions, were: I) All Dynapore surfaces yield under low impact levels unless supported by subl ayers.

2) A titanium surface was far more resistant to surface fracture under impact than Dynapore.

3)

An improvement in bond strength bewteen 316L Dynapore and the body of the panel as well as an improvement in cohesive bond strength between underlying layers of the porous non-metallics was necessary.

4) The bending specimens that had been subjected to freezing water and temperature and humidity cycling showed no degradation of strength.

F 9.2.2 Aerodynamic Smoothness Test Panels Stiffened porous panels 280 mm chord x 1190 mm span (I ft x 4 ft), were designed and constructed for wind tunnel testing (see Section 8.0). The panels tested are listed in Table 9-8. They were primarily aerodynamic models not strictly representative of glove panel construction. A Doweave laminate was used for porosity metering under the 50 x 25C Dynapore and Fibermetal surfaces but the panels were found to be consistently too porous compared with predictions based on a series resistance analogy. A bonded honeycomb core support was used to hold the thin Dynapore surfaces flat and rigid under suction pressure, although it would not be adequate to develop necessary panel impact resistances, if used alone under the low-yield surface material. The Dynapore backed with _erforated sheet material with holes as large as 6.35 mm (0.25 in) performed satisfactorily but smaller holes would be used to provide acceptable impact resistance. The 80 x 700 Dynapore with a low porosity r25 SCFM/14 PSF) performed the best, so a combination of 80 x 700/80 x 80 surface and perforated fiberglass sublayer (40 percent open perforation pattern) was bonded to honeycomb for additional wind tunnel tests. The resulting porosity of 0.508 m3/s/m 2 @ 670 Pa (IOD SCFM/ft 2 @ 14 PSF) gave poor results due to an inflow/outflow condition caused by a small pressure gradient over the panel. The panel porosity was therefore reduced to approximately 0.178 m3/s/m 2 @ 670 Pa (35 SCFM/ft 2 @ 14 PSF) by the addition of porous glass plies on the backside. This panel worked well and a subsequent test panel of the same average porosity, achieved by using a perforated sublayer containing 3.18, 4.76 and 6.35 mm (I/8, 3/16 and I/4 inch) diameter holes beneath the 80 x 700/80 x 80 Dynapore surface, worked equally well in the tunnel.

.

TABLE9-8

AERODYNAMIC SMOOTHNESS TEST PANELS

(Suction Surfaces)

o Baseline Flat Plate - 3.18 mm(I/8 inch) thick Aluminum sheet

o 50 x 250 Dynapore - Honeycomb Sandwich

o Sintered Fibermetal on Doweavelaminate

o 50 x 250 Dynapore - with strip porosity obtained by controlled adhesive

blockage.

o 50 x 250 Dynapore on Isogrid stiffening.

o Metallized Doweavelaminate.

o Metallized #120 molded fabric surface on Doweavelaminate.

o Microperforated Plate (Dynapore) #19 on DoweaveLockcore with flutes

running spanwise.

o Slotted Aluminum panel.

o 80 x 700 Dynapore on Honeycomb.

o 80 x 700/80 x 80 Dynapore on perforated fiberglass sublayer - Honeycomb

stiffened.

o Electron BeamPerforated Titanium 0.102 mm(0.004 inch holes)

OF ?_b:_:_ {)'..;_Llll t The slotted panel was fabricated by precision machining aluminum sheet and fitting it together with shim spacers during the bonding operation to produce 0.127 mm (.005 inch) slots. The first slotted panel was not smooth enough but a satisfactory test panel was produced after modification of the assembly tool.

The metalized Doweave and metalized #120 fabric surfaces were found to be not smooth enough for laminar flow in the wind tunnel, even though molded against a Mylar sheet. The electron beam perforated titanium panel was procured and tested using a hole pattern suggested by NASA. It is not known whether Dynapore coarser than 50 x 250 would have been too rough, however, the heavier Dynapore was difficult to roll precisely to low porosities during the calendering process.

Figure 9-31 shows the effect of calendering on porosity.

9.2.3 Pressure Drop Through Combined Porous Layers In support of the wind tunnel flat plate tests, pressure drop testing was conducted on various numbers of plies of Doweave that were stacked in alternate weave directions [0/90] n to avoid uneven porosity mistribution in the ply stack.

A goal was to discover a method of predicting the porosity of such porous layups and to produce porosity design charts. Figure 9-32 represents an attempt to organize the test data, and indicates the effect of ply stacking of basic Doweave on pressure drop at a single .012 m/sec (.Oa ft/sec) flow velocity through the surface. This was the lower limit of velocity through_ the suction region. The target pressure drop for one facing of a two-faced porous sandwich, assuming no contribution of the core, was 335 Pa (7 psf) as indicated in the right of the chart. The total sandwich &P was 14 psf at that flow, with the relationship for separated porous facings assumed to be suction velocity _PI '_P2 _PTotal where Vs = + pressure drop V s V s V s and _Pi = ORIGINAL PAGE i$ OF POOR QUALITY'

:' _}i_/!i_!i_!/,_ _

FIGURE 9-30. ISOGRID AND HONEYCOMB CORES BONDED TO DYNAPORE FACINGS (IN.) mm AS WOVEN _ 150x 2SOl 0.012 F 031 , , , -_, -]-zOO

!

0_)11 [ 0.008 _- ^.H--AREA OF --H ./ /I I--I /

u_ [_--INTEREST-- I I 1t" / I 1t7o /

x" /

h ..-t I! +=

J o"'""" ,.....,.,_.._ _"-50 x 250 14-5o / o_)oe'-o.151' _e..-® '_ l i| Jm, 0.01 O.10 1.0 10.0 PERCENT m3/SEC/m2 AT 670 Pja ASWOVEN I I 1(_1010001 THICKNESS 1.0 10 (SCFM/FT2 AT i4 LB/SQ FT) , ....... ,_ FIGURE 9-31. EFFECT OF CALENDERING ON POROSITY OF DYNAPORE MATERIALS It was apparent that increasing the number of Doweave plies alone could not produce the required pressure drop. Inserting of (l, 2 or 3) more closely woven #350 fabric plies regularly separated by (0/90) n Doweave plies, seemed to have a metering effect on the stack. Extrapolation to 5 inserted metering plies indicated that this would produce the required 670.3 Pa (14 psf), however, the panel made for the wind tunnel model in this way was found to have a pressure drop more than one magnitude less than expected.

No simple relationship between the number of plies and the pressure drop could be determined so porosity control using additional porous layers continued to be a "trial and error" process.

9.2.4 Porous Surface Selection Following the initial survey and testing of a variety of porous surfaces, two were selected for further development. These were: (1) stainless steel Dynapore mesh, and (2) electron beam perforated titanium sheet. These two surface materials were then tested more extensively to obtain the physical and mechanical properties needed for comparison and for LFC panel design.

The evaluation of Dyanpore surfaces was limited to an external 80 x 700 mesh weave diffusion bonded to an 80 x 80 mesh inner weave calendared to a porosity producing 0.813 m3/s/m 2 (160 SCFM/ft 2) airflow at 670.3 Pa (14 PSF) pressure, see Figure 9-33.

Both 316L and Nitronic 50 stainless steel wire materials were evaluated. In addition, work was conducted on the configuration in which the Dynapore layers were also diffusion bonded to an internal supporting electron beam perforated 316L stainless steel sheet.

The evaluation of electron beam (EB) perforated titanium sheet surfaces was initially with material supplied by the Farrell Company, Connecticut. Testing confirmed the possibility of using EB perforated titanium an an LFC surface and its relative toughness was highly desirable, particularly for leading edge surfaces where high impact resistance is needed.

The smallest holes obtainable from Farrell in 0.685 (.025 inch) thick titanium

were 0.I02 mm(.004 inch) diameter. There was someconcern that suction through

holes of this size could trap particles on the surface large enough to cause

transition. The use of elongated holes was therefore investigated to reduce

this possibility.

The Farrell Company subsequently decided to discontinue EB perforation and

additional material was obtained first from Steigerwald in Germanyand then

from Pratt and Whitney in Connecticut. Pratt and Whitney were able to produce

EB perforated holes as small as 0.064 mm(.0025 inch) diameter in 0.64 mm

(.025 inch) thick titanium sheet material. This reduced the hazard from possible

trapped particles on the surface so development was continued using circular

holes only. The maximum ratio of lO:l in sheet thickness to hole size was

confirmed as a limit by Steigerwald, the manufacturers of the electron beam

perforating equipment, although Dr. Steigerwald has since stated that his

equipment could be modified to produce even smaller holes.

A statistical analysis was madeof EB perforated hole sizes. The sampling

technique used was to photograph a set of holes using a metalograph (inverted

microscope). These photographs were then used to determine hole size variations

and obtain a statistical distribution. The results showedvery little variation

in hole size indicating that a reliable uniform porosity was obtained.

Figure 9-3_ shows an example of the uniformity of EBhole sizes and distribution

obtainable and Figure 9-35 shows an electron microscope photograph of an EB

perforated hole highly magnified.

The Pratt and Whitney material was of high quality. Sheets of 0.635 mm(.025 inch)

thick 6AL4Vmaterial perforated with 0.064 mm(.0025 inch) diameter holes spaced

0.813 mm(.032 inch) on centers in a square pattern were obtained and used for

wind tunnel testing. Porosity tests indicated a flow rate of 0.014 m3/s at 670 Pa

(30 SCFM at 14 PSF) pressure differential for this material, which was acceptable

for the wind tunnel models.

I 0,001 (0.048 Pa ) FIGURE 9-32. PRESSURE DROP PANEL DESIGN - POROUS COMPOSITES

56X

140X .......... _ _ __-_-'- H t+l+> , FIGURE 9-33.

80 BY 700 DYNAPORE SURFACE PLUS DIFFUSION-BONDED 80 BY 80 SUBLAYER ORIGINAL PAGE |9 OF POOR QUALITY FIGURE 9-34. ELECTRON-BEAM-PERFORATED HOLES :i FIGURE 9-35. ELECTRON-BEAM-PERFORATED HOLE During fabrication development with EB perforated Ti material, it was determined that normal etching and cleaning procedures before bonding would enlarge the hole sizes by about 0.025 mm (.OOl inch) diameter. A new preparation procedure was therefore developed using a protective coating applied to the outer surface of perforated titanium. The coating had been developed previously for preferential chemical milling procedures. The coating inhibited enlargement of the perforated holes at the surface and permitted normal etching and cleaning of the interior surface prior to bonding.

Structural testing of Dynapore and EB perforated titanium is described in Section 9.3. The EB perforated titanium LFC surface was finally selected.

9.2.5 Thermal Analysis of Dissimilar Materials The effect of using dissimilar materials for porous panel surfaces, panel stiffening and for the basic wing structure was considered. The effect on LFC panels, which being securely attached to wing structure, strain to the same extent as the primary structure, is illustrated by Figure 9-36. In the example indicated on the chart, a panel material with a 68.9 GPa (lO million psi) modulus would reach a stress level of 186.2 MPa (27,000 psi) when strained with a graphite/ epoxy that is allowed to reach 68.6_m (2700 micro-inches) at limit load.

A titanium wing structure would be allowed to reach higher strains and would therefore strain the glove material to a higher stress level. A graphite/epoxy basic wing structure results in the lowest glove stress. Thermal strains that may be present, due to the difference between fabrication temperature and service temperature in bi-material systems, must be added algebraically to the load strains to obtain the resulting strain level of the glove material.

The glove panel should be designed to not yield or buckle below wing limit load taking into account any thermal strains present. Thermal strains must therefore be considered when determining the feasibility of structure and panel material combinations. Because of a lack of proportionality indicated by a pronounced "knee" in the stress-strain curve for Dynapore, the secant modulus (Es) rather than the modulus to proportional limit (Ep) was used in the thermal analysis of Dyna pore panels.

The thermal analysis was programmed to accept orthotropic properties of two different layers intimately connected and subject to a temperature change. The program directly solved for single-materials (self-stiffened) Kevlar, glass or steel mesh panels on each of the four primary structural materials: aluminum, titanium, boron epoxy or graphite/epoxy. Multi-material panels were solved in two stages, considering first the bonding temperature of the panel to determine the balanced stresses within the panel layers when they are being assembled to the primary structure, then the effect of temperatures down to -65°F on the combined assembly. The maximum strains in the panel layers occurred at -65°F or room temperature depending on the material combinations and thicknesses.

Generally, the thermal strains in the primary structure are very low since it controls the panel strains by virtue of greater cross sectional area, Figure 9-37 summarizes rationale and materials combinations.

In the examples given in Table 9-9, the first column shows the calculated maximum thermal strains induced at -65°F or 70°F for several multi-material panel/structure combinations. It shows the strain available for limit tensile wing loading after subtracting thermal strain from the allowable layer strain. If a layer available strain is less than the structure strain, then that combination is infeasible.

The very thin microperforated plate (MPP #21) appears to be infeasible on any structure. 24 x II0 Dynapore on graphite and 50 x 250 Dynapore on titanium wings were indicated as problem combinations although 50 x 250 Dynapore on graphite is feasible. 50 x 250 Dynapore is feasible even with an aluminum wing, at least 4"" ',J I" . =.... ; " (KSI) 50 100 150 (GPa) (100) I f MPa (90) (THERMAL STRAINS ARE ADDITIVE) (80) PRIMARY STRUCTURE MATERIALS ._\_ 500 - (70) REQUIRED GLOVE YIELD 400 - (60) STRENGTH (50) AT LIMIT LOAD 300 - (40) 200 - ( 3O ) 7-i.- ....... . CO PATIBLE (20) 100 -(10) i (20) (PSI x 10 -6 ) (5) (10) (15) GLOVE PANEL MATERIAL STIFFNESS (E) FIGURE 9-36. STRENGTH AND STRAIN REQUIREMENTS FOR LFC GLOVE PANELS PANEL - BONDED AT 350°F 350 ° - 70OF PANEL-INSTALLED ON STRI.1CTURE AT 70°F COMBINATION OPERATED AT 70 ° _ -65°F MATERIALS COMBINATIONS STRUCTURE SURFACE PANEL • ALL KEVLAR/EPOXY • ALL GLASS/EPOXY • TITANIUM • ALL DYNAPORE • BORON/EPOXY • ALUMINUM • DYNAPORE ON KEVLAR/EPOXY LOCKCORE • GRAPHITE/EPOXY • DYNAPORE ON GLASS/EPOXY LOCKCORE • DYNAPORE ON KEVLAR/EPOXY ISOGRID 7-GEN 2275/0 • DYNAPORE ON GLASS/EPOXY ISOGRID FIGURE 9-37. THERMAL ANALYSIS - MODEL AND MATERIALS ORIGIDIAL Pf'_GE r_ OF POOR QUALITY _.J O O Z >- C_ ¢"3 ,,"_ _,q "_-.,__ _.

_,Z ."-- __ .-.., <ZZ v 00_ O00 r--. 0 _') __, q_) .,--_ b.-._ .._._ %.. C'_ ,,_- O,J C_, r'-- _ i,..-- ......j ,,_..

r,- I-- L,LJ >.- O_ .,..J C v v [,., 00 0 0O O_C_ -'r- _-- L...) _ O.JO c'o O" CO Zl.-- E • r-- 0 c • .._J ._r:, _- $..

_ Z 0 "_ '-0 t-- 4--) _.,-- r'_ 0 $-- "_- c- 4-.) ._ L.0 e--- _. -r-- I.I_ O"J o _ 0 _J U 0 f.. v 4.._ I ",=" X -.J • "0 0 C,_ _.-- 0:_ f o.

.-J L_ Z O,,i "_" t,_ I'-- r_ Z CF TL,_I " (_i;::.'.JEY' in tension. There was uncertainty about the appropriate stiffness input for a Kevlar isogrid panel stiffening layer and when theoretical KV49 values were used the Kevlar isogrid combinations were infeasible. When glass or lower stiffness Kevlar 29 theoretical values were input the combinations were feasible.

9.2.6 LFC Panels Arranqement A design study of the porous glove panel structure was undertaken and details of a glove panel design and its assembly to the basic wing structure were drawn.

This design provided the basis for glove panel structural testing and development.

Figures 9-38 and 9-39 show the arrangement of the panels on the wing.

The LFC panels were trapezoidal rather than rectangular in shape so that all streamwise flow paths would have a reduced non-sucked distance to travel across structural joints, and no joints were made parallel with the stramlines. Sub- surface blockages, such as at node lines, were also angled parallel with the panel end joints. The angled fluting made the panel difficult to analyze structurally, however panel tests showed this arrangement to be stronger than with the flutes at 90 ° to the stringers. Details of the glove panel at this stage of development are shown in Figure 9-40. The double plate butt splice design shown in Figure 9-41 was used for the chordwise joints. Acceptable smoothness of these panel joints was not established in this program but structural feasibility was demonstrated.

Because of panel length, interior fasteners, or "field fasteners" were used for added support and stabilization, see Figure 9-42. The overall concept showing how the suction flow was controlled and transferred to the integral ducting of the basic wing structure is shown in Figure 9-43.

9.2.7 Superplastic Formed/Diffusion Bonded LFC Panel To examine the feasibility of using a Superplastic Formed/Diffusion Bonded (SPF/DB) structure for LFC glove panels, a small panel of this type was fabricated. The formed panel is shown in Figure 9-44. The dimples formed in the shallow channels ,i "" were subsequently machined to create rows of holes and allow suction air to flow from the surface to the integral channels. The resulting panel is shown in

OF POC2

QUALITY PLAN VIEW---AR=-IO WING SUCTION BOTH SURFACES 8_,EN.22_774 FIGURE9-38. GLOVE PANEL INSTALLATION AIR CHANNELS FRONT SPAR STREAMWISE FLOW SPAR LFC GLOVE PANEL JOINTS 8-DP-80053A / FIGURE 9-39.

TYPICAL WING BOX GLOVE PANEL ARRANGEMENT ,OF P_:_:".. _,,": _',:'. , _: STREAMWISE

___PERFORATED S-G,ASS,AM,NAT'E-_

<,MPACT. 'AYER i

80 x 700/80 x 80 DYNAPORE SURFACE /-- / ///--GLASS" LAMINATE ' i "-- ,_' I I "1 ':'_ PANEL 8.DP-80062 I I FIGURE 9-40. LFC GLOVE PANEL WITH PERFORATED FIBERGLASSSUBLAYER "E" GLASS/EPOXY SPLICE WITH 80 BY 700/80 BY 80 DYNAPORE BONDED TO BOTTOM SURFACE "_ ! / Ii¢ , 4 IN. O.C. EXCEPT TYPICAL SPANWlSE SPLICE b _1 _ COMPRESSION SHIM 8-GEN-22773 TYPICAL CHORDWISE SPLICE _ FIGURE 9-41. GLOVE PANEL SPLICES

OF PO011 (_U,-",LI"£Y

._(,,j_ R EAMWISE FLOW ILi f i !

FLUTE AIRFLOW SEPARATOR FIGURE 9-42. TYPICAL FIELD FASTENER - LFC GLOVE /'_LEADING EDGE DYNAPORE OUTER SURFACE LAYER PERFORATED FIBERGLASS LAYER "--, STREAMWlSE FLOW DIRECTION ANEL FC GLOVE_ SUCTION FLOW \ FRONT SPAR COLLECTOR CHANNEL METERING NOZZLE MAIN BOX SKIN 8-GEN-_777 FIGURE 9-43. LFC GLOVED WING STRUCTURAL CONCEPT Figure 9-45 and an enlarged view of the holes is shown in Figure 9-46.

The LFC glove panel could be completed by bonding either a layer of Dynapore

or EB perforated titanium to the outer surface. The SPF/DBconstruction was

judged to be acceptable for LFC panels and further development of this concept

is recommended.

9.3 STRUCTURAL TESTING

9.3.1 Porous Materials Strength Properties

To obtain tensile and compressive stress-strain data on porous surface materials, Douglas standard honeycomb bending test beams were used in a four point loading rig. These beams are those normally used for thin composite materials being tested under compression loading. Figure 9-47 shews a typical test beam configuration.

Six beams were fabricated with a 316L Dynapore surface: 3 with an 80 x 700 surface diffusion bonded to an 80 x 80 sublayer, and 3 with an 80 x 700 surface alone. Four similar beams were fabricated with EB holes .0045 x .0073 long in 6AL4V titanium (2 of each with holes in D° or 90 ° directions). Strain gages were installed on the beams for testing for strain/yield characteristics. Each beam was tested to a strain of .OOa using 2 tests at each of 7 increasing load steps followed by 2 tests at each of 5 decreasing load steps. All initial tests at increasing strains in Dynapore showed that for strain levels up to .002 some small residual strain existed upon unloading which disappears upon reloading up to a strain of .002. For strain levels of .003 and above, the residual strain was reduced after the second loading but did not "disappear" as indicated for lower strain levels.

Prestraining to ,003 resulted in no residual strains remaining after loading at lower levels. Both fill and _45 ° warp directions in the Dynapore weave were tested. Figure 9-48 shows the stress-strain relationship in tension and compression for 80 x 700/80 x 80 316L Dynapore bonded to a fiberglass sublayer

[

ORIGINAL P#,,3:_ _,_ OF POOR QUALLTY FIGURE 944. SUPERPLASTIC FORMED/DIFFUSION-BONDED TITANIUM SANDWICH PANEL - BEFORE MACHINING FIGURE 9-45. SUPERPLASTIC FORMED/DIFFUSION-BONDED TITANIUM SANDWICH PANEL ORIGINAL PAGE _S OF POOR QUALITY

__o • co •

,, .... . - '--Z_." -- -- "

F'oo 0 • *o*o • • • • •

.o O.OJ__o _e'_ _ • o o O O"

FIGURE 9-46. SUPER PLASTIC FORMED/DIFFUSION BONDED TITANIUM SANDWICH PANEL SUCTION HOLES

1]

STRAINGAGE---- S

I

I (LINEAR) _ '----- HONEYCOMB CORE

/

__1 11Z mm 5,1,11,1, \ TEST FACING (2: _.0) 111111 (4.0) (2.01 • 55¢.

/

h 76.2 mm mm (3.0) '_1 31.75 U I AREA

_t__ k ---I, o01---I

TEST !

(11.0) 279 mm

t

{9,5/ I 241.3 mm (1.50) A A

t_ __J

38"ir / 1.

/ F..3 AONES,VE \

OR EOO,VALENT \

0.1_3 7075-T6 A-A.

FIGURE 9-47. TEST BEAM and Figure 9-49 is the stress-strain curve for 80 x 700 Dynapore made from Nitronic 50 wire. The offset of the curve at the axes intercept is the result of thermal stress introduced during the Dynapore to fiberglass bonding process.

Calendered woven metallic materials exhibit a combination of elastic and mechanical strain characteristics depending on load direction relative to the weave. The only requirement for the LFC surface is that the porous material must withstand a limit load strain of ±.0027 without surface deformation after the strain is reduced to the Ig flight condition. The .0027 strain corresponds to a .004 ultimate strain for advanced graphite/epoxy composite materials (current design practice limits ultimate strain in graphite/epoxy materials to 2.003). The test results indicate that some damage may be occurring with the 316L material but Nitronic 50 appears to be still behaving elastically within its limit of proportionality at the required strain level.

All of the 6AL4V titanium faced beams tested with .114 mm x .185 mm (.0045 inch x i- .0073 inch) long perforated holes exhibited the same stress-strain characteristics as the original sheet materials. The percent porosity was too low to significantly affect the stress-strain relationship up to the .006 maximum strain limit of the tests.

9.3.2 Porous Material - Fatigue The Dynapore fatigue test beams were as shown in Figure 9-50. Specimens include 316L 80 x 700 warp and +45 ° plus an equivalent set using Nitronic 50 wire, as shown in Tables 9-10 and 9-11. Tests were run at "average" reversing strain conditions (R=-2.5) and a design objective of 120,000 cycles at .0016 maximum compressive strain.

Figure 9-51 shows a set of fatigue test sandwich beams with electron beam (EB) perforated test facings. Note that the fatigue beams are wider than the normal one inch wide beam in the foreground which is used for materials properties evaluations. The wider beams are used in the fatigue test rig to prevent loading in the plane of the test facing from the test fixture.

_,_::? i _ ;. _i:_i ' t-*: MPa (KSI) (20) Ioo 0.001 0.002 0.003 0.004 -0.003 -0,002 -0.001/1 STRAIN

/]

--50 -- (-lo) 80 BY 700/80 BY 80 316L DYNAPORE WITH FIBERGLASS SUBLAYER - 1__- 11(1(1 '(-20) 5O FIGURE 9-48. STRESS-STRAIN CURVE-DYNAPORE ON FIBERGLASS MPa (KSI) 40- 30- 20- 10- STRESS 0- -10 - -IO0 -20 - -200 -30 - -40 - m -300 I I I I -0.003 -0.001 0 0.001 0.003 0.002 0.004 -0.002 B3 GEN 23297 STRAIN FIGURE 9-49. STRESS-STRAIN CURVE - NITRONIC 50 DYNAPORE ORIGIF'AL P_(_;_]" i_ OF POOR QUALITY TABLE 9-I 0 FATIGUE LOADING REQUIREMENTS Strain x 106 Test Beams No. Cycles Tension 640 1 each 120,000 R = -2.5 dash no.

Compr. -I 600 or Failure 1 each Compr. -2400 Record Fatigue R = -2.5 Tension 960 dash no.

Cycles to Failure 1 each Compr. -3200 Record Fatigue R = -2,5 Tension 1280 dash no.

Cycles to Failure or Higher as Indicated TABLE 9-I 1

Y

FATIGUE SPECIMENS AND IDENTIFICATION Dash No.

No. Beams Facing Material -I 80 x 700 only, 315L, warp lengthwise.

-2 Same as -l except _45 ° -3 80 x 700 Nitronic 50, Warp lengthwise.

-4 80 x 700 Nitronic 50, _45 ° 12 TOTAL ORiG;NAL ......

OF" POOR QUALITY S HONEYCOMB CORE DYNAPORE TEST FACING -- BOND [)-DEC WITH NUMBER 120/7251 \ _ 0.0B0 NUMBER 120 GLASS/EPOXY 90-DEG-_ \ WARP IN O-DEG DIRECTION : \ \ (3.0) "l LINEAR I

\ \ ;,6.2 ,,,, /

STRAIN

I I- \ \ (2.0)--.- I----(o.s)

(0.25) 50.8 mm 12.7 mm (22.0) 6.35mm--_ _-_ I 1; 559m_ \ "J '_'--FILM ADHESIVE (11.o) (2.0) 279 mm EDGE HOLDDOWN 50.8 mm I

II

FM73 ADHESIVE OR EQUIVALENT i (2.75) _- _ (0.12) 69.8 mm 3. 05 mm 0.813 mm 7075-T6 K (0.032) SECTION _""_ FIGURE 9-50. FATIGUE TEST BEAM FIGURE 9-51. ELECTRON-BEAM-PERFORATED TITANIUM SHEET FATIGUE SPECIMENS Fatigue testing of 80 x 700/80 x 80 Dynapore in reverse loading conditions was initially at R=-2.5, 650_ strain in tension and -1600y strain in compression.

This test simulates normal maximum operating stress conditions for a ground to air to ground cycle. The tests showed the material fatigue life capability at this strain to be in excess of 2 million cycles (7 30 lifetimes) for both warp and _+45° orientations.

Due to the high number of cycles to failure on the Nitronic 50 specimens, a cutoff point of I0 million cycles per specimen was established to limit test times. The fatigue tests results on 316L Dynapore, Nitronic 50 Dynapore, and Electron Beam perforated 6AL-4V titanium porous materials are shown in Figure 9-52 together with the target design requirement for the reversed loading conditions.

The Dynapore testing was conducted in 2 directions: parallel to the warp fiber (spanwise on the wing), and _45 ° to the warp fiber direction. The i- perforations in the EB Ti material were approximately 0.114 x 0.185 mm (.0045 x .0073 inches) and the specimens were tested in a direction parallel to the slotted direction, (spanwise orientation on the wing) and perpendicular to the slots (indicated by II and 1 respectively in Figure 9-52).

The test results indicate that LFC panel designs using either 316L Dynapore or EB Ti porous surfaces would be adequate for fatigue resistance. The added expense and difficulty in fabricating Nitronic 50 Dynapore is not required to provide fatigue resistance. The EB perforated titanium also exceeds design requirements with holes elongated in either direction.

j ' 9.3.3 Impact Testing The test specimens for the initial impact evaluation of perforated stainless steel backed Dynapore surfaces and electron beam (EB) perforated titanium sheet are described by Figure 9-53 and Table 9-12. The EB perforated Ti sheet materials were purchased from the Farrel Company. The test results are shown in Figure 9-54. It should be noted that the results shown are for maximum depressions which actually occurred at the center lines of the flutes in the substructure.

Final impact testing was conducted using standard Gardner tests on all of the materials listed previously plus a Dynapore surface configuration consisting of Dynapore diffusion bonded to a sublayer of 0.635 mm (.025 inch) thick EB perforated 316L stainless steel sheet with .254 mm (.010 inch) diameter holes spaced 2.54 mm (0.I0 inch) on center in a square pattern. The test results are shown in Figure 9-55 together with results for non-perforated materials normally found on leading edges of commercial aircraft. The 0.635 mm (0.025 in) thick electron beam perforated titanium was more resistant to damage than the other materials tested, including conventional leading edge materials of 1.27 mm (0.05 in) thick 7075-T6 aluminum.

9.3.4 Rain Erosion Testing A total of 44 Dynapore and 20 EB titanium porous surface rain erosion specimens were tested to evaluate their resistance for leading edge surfaces.

Initial rain erosion test specimens were defined as shown in Figure 9-56 and Table 9-13. The testing was arranged through Mr. George Schmitt, Jr., AFML, and was conducted at Wright Patterson AFB. The test facility consists of a powered 2.4 m (8 ft) double arm propeller blade mounted horizontally and the specimens were mounted in the blade tips. A simulated rainfall of one inch per hour was produced with controlled water droplets, which impinged on a pair of specimens as they are rotated. The specimens were tested initially at 179 m/s (400 mph).

L 0.005 \ MAXIMUM COMPRESSION STRAIN R = = -2.5 MAXIMUM TENSION STRAIN 0.004 RONIC 50 ( +_45 °) _TRONIC 50 (WARP) 0.003 MAXIMUM _'_*_ \_\ PORE (WARP) __ _- COMPRESSION C] _6L DY'_APORE ( 45 o) STRAIN 0.002 EB Ti (11) __ _ ""0-.-_ EB Ti (1) 0.001 DESIGN REQUIREMENT I I I I I I 0 ..... ,,,I , , , ,,,,,I , , , ,,,,,I , , 10,000 100,000 1 MILLION 10 MILLION NUMBER OF CYCLES FIGURE 9-52. POROUS SURFACE FATIGUE 190.5 mm / (7.5 IN.)

12.7 mm P I 41.7 mm 19.6 mm / (1.64 IN.) TYP"_="

-'t (0.77,.,r-- I (0.5 IN.)

TYP 3 PLACES

r-

I I I I LAYER 1 LAYER 2 _/i///////_////_ E///IY_///A V//_///////A _1//////_ 7//a v///J.//////////_ /////./.dl V.////X, LAYER L_ _ 21.3 mm LAYER (0.20 IN.)" In ':,n J_l _ L m _"/ 5.1 mm (0.37 IN.)

9.4 mm .

FIGURE 9-53. IMPACT TEST SPECIMEN (SEE TABLE 9-12 FOR DEFINITION OF LAYERS) OF pOOR QUALITY TABLE 9-I 2 IMPACT TEST PANELS Panel Layer Configuration 1 2 3 4 5 80 x 700/ 0.508 mm 2.285 mm 1.016 mm 0.762 mm Dyna pore D-I 80 x 80 (.020 in) (.090 in) (.040 in) (.030 in) Dyna pore Perf SS Glass/Epoxy Glass/Epoxy Glass/Epoxy r23%_ open) Lain (4) Lam (4) Lam (4) Dyna pore 80 x 700/ .508 mm 1.854 mm 0.991 mm 0.991 mm D-2 80 x 80 (.020 in) (.073 in) (.039 in) (.039 in) Dynapore Perf SS (5) (5) (5) (23% open) EB Perf Ti EB Perf Omit 2.286 mm l.Ol6 mm 0.762 mm T-I Ti (.090 in) (.040 in) (.030 in) (See AVl NS-20) (4) (4) (4) EB Perf Ti EB Perf Omit 1.981 mm 0.991 mm 0.991 mm T-2 Ti (.078 in) (.039 in) (.039 in) (See AVI NS--20) (5) (5) (5) Note : ( 1 ) For configurations Dl and D2, resistance spot weld layer l to layer 2 approximately I/40.C to simulate a diffusion bonded panel from the vendor.

(2)

Sew and cure layers 3, 4, and 5 prior to attach of layers l and 2.

(3) Bond layer 1 (or layer 2 if indicated) to cured substructure with

film adhesive using strips on layer 3 surface contact area only.

(4)

120/7251 Glass-Epoxy cloth laminate.

(5)

T300/5203 Graphite-Epoxy cloth laminate.

f OF POOF,t Q_ _,_.,;y mm (INCH) I 80 BY 700/80 DYNAPORE i RESIST. WELDED TO 0.635 mm (0.025 IN.) PERFORATED (23% OPEN) • 2"0I(0"01 t STAINLESS STEEL ON GL/EP _" I SUBST RUCTLJ RE -X 1 t 1.5 r ( O.Oq /

IMPACT /

DEPRESSION | , m

,.o[(o.o

i j t EB PERFORATED 0.635 mm (0.025 IN.) 6AI-4VTi ON GL/EP SUBSTRUCTURE _'_ 0.5[(0.0: i I l I I I I I I 1 J 0 u (0.5) (1.0) (1.5) (INCHES 0 I I I mm 0 10 3O 2O STRIP WIDTH FIGURE 9-54. IMPACT RESISTANCE -2.26N-m(201N.-LB) GARDNER TEST (mm) IN.

80 90 I f FIGURE 9-55. SURFACE IMPACT-RESISTANCE COMPARISONS OF POOR _UALITII' TABLE 9-I 3 EROSION SPECIMEN MATERIAL CONFIGURATIONS Layer No.

Specimen 2 3 5 6 Configuration -501 80 x 7001 0.508 mm 2.032 mm 1.27 mm 0.762 mm l.Ol6 mm 80 x 80 (.020 in) (.080 in (.050 in) (.030 in) (.040 in) Dynapore Perforated (1) (I) (I) (I) SS Sheet -5O3 1.32 mm 0.991 mm 0.991 mm 80 x 700/ 0.508 mm 1.982 mm 80 x 80 (.020 in) (.078 in (.052 in) (,039 in) (.039 in) Dynapore Perforated (2) (2) (2) (2) SS Sheet .025 EB Omit 2.032 mm -5O5 1.27 mm 0.762 mm l.Ol6 mm Perforated (.080 in (.050 in) (.030 in) (.040 in)

Ti Sheet (I) (i) (i) (i)

.025 EB Omit 1.982 mm -5O7 1.32 mm 0.991 mm 0.991 mm Perforated (.078 in (.052 in) (.039 in) (.039 in) Ti Sheet (2) (2) (2) (2) -509 Same as -501 except bond per note 6.

-511 Same as -503 except bond per note 6.

Notes : (1) 120/7251 Glass-Epoxy cloth laminate.

(2) T300/5208 Graphite-Epoxy cloth laminate.

(3) For -501 and -503 configurations, resistance spot weld layer l to layer 2 approximately 6.35 mm (.25 in) O.C. to simulate a diffusion bonded material from the vendor.

(4) Sew, cure, and provide suction holes in layers 3 through 6 prior to attach of layers l and 2 as shown.

(5) Bond layer l (or layer 2 if indicated) to cured substructure with film adhesive on layer 3 contact areas only.

(6) For -509 and -511 configuration, bond layer of 120 fiberglass cloth to internal surface of layer l and 2 assembly before bonding per note 5.

f f Figures 9-57 and 9-58 show typical results of the first series of rain erosion tests. All Dynapore surfaced specimens show damage due to the Dynapore becoming dimpled into the holes in the stainless steel sub-layer. Permanent surface bowing also occurred across flutes. All of the electron beam perforated titanium surfaces show adequate rain erosion resistance. A second set of Dynapore specimens with smaller holes in the sublayer to increase rain erosion resistance was fabricated, configured as shown in Table 9-14. These specimens incorporated a range of smaller diameter hole sizes in the sublayer plus combinations of these sublayers with both Nitronic 50 and 316L Dynapore. These specimens were tested at 179 m/s (400 mph) for 120 minutes, and an additional set was tested at 224 m/s (500 mph) for 60 minutes, An extra set of EB titanium specimens with improved bonding was also tested to check bond strength because of debondlng found previously on some samples.

Final results can be summarized as follows: l . With 80 x 700/80 x 80 or 80 x 700 Dynapore in either 316L or Nitronic 50 material diffusion bonded to perforated stainless steel sublayers, a maximum hole size of 0.381 mm (.015 inch) diameter in the sublayer is allowable to prevent the Dynapore being dimpled into the sublayer holes.

. All Dynapore specimens diffusion bonded to a 0.635 mm (0.025 inch) thick 316L stainless steel sublayer became permanently bowed across either of the 7.26 mm (0.3 in) or 12.7 mm (0.5 in) gaps in the sublayer. Subsequent analysis shows that a 0.83 mm (0.032 inch) thick layer would prevent depressions at speeds up to 179 m/s (400 mph). Curvature of the surface of a wing leading edge would of course normally increase strength compared to the flat specimens tested and impacts would be lessened by the inclination of the surface other than directly at the leading edge.

3. Nitronic 50 wire Dynapore did not provide increased erosion resistance over 316L Dyncpcre.

-1_ I 7.87 mm I I I l f F-- (0._ __ "l _ j 9.65 mm 25.4 mm i i I i (I.0 IN.)

'1 " f / T , I f \ I / _ I I

l

I i " (2.44 tN.) " 62 mm mm 7.49 mm 12.7 mm 12.7 mm .._____23.42 IN')-----_I"-----_I -(_-'30 iN.)

LAYERS I(0.50 IN.) I (0.50 IN.) I (0.92 1_ t _ _t : :_ 7.62 ram.... 15.2 mm 2 _-__._ ........ ! .... (0.30 IN,) I"--(0.60 IN.)"" m)

\' Z,:,LL W,TN s'n'CAST

AS sHoPwN \ 3180 M OR EQUIVALENT W 7 LACES _-6.35 mm (0.25 IN.)

DIA HOLES -- 4 PLACES IN LAYER 4 83 C,_N 2326_ FIGURE 9-56. RAIN EROSION TEST SPECIMEN (SEE TABLE 9-13 FOR DEFINITION OF LAYERS)

-5ol

_q _IN. 20 NUN.

FIGURE 9-57. TESTED DYNAPORE RAIN EROSION SPECIMENS TABLE 9-I 4 EROSION SPECIMENS WITH SMALLER SUBLAYER PERFORATIONS Confi guration Layers 1 & 2 Materials (for other layers see Table 9-13) NS -501 80 x 700/80 x 80 Dynapore plus 0.635 mm (.025 in) SS sheet perforated with 0.381 mm (.015 in) dia. holes @ 1.143 mm (.045 in) pitch.

-503 Nitronic 50(5)plus 0.635 mm (.025 in) SS sheet perforated with 0.381 mm (.015 in) dia. holes @ 1.143 mm (.045 in) pitch.

- 5O5 80 x 700/80 x 80 Dynapore plus 0.635 mm (.025 in) SS sheet perforated with 0.254 mm (.010 in) dia. holes @ 1.524 mm (.060 in) pitch.

-5O7 80 x 700/80 x 80 Dynapore plus 0.635 mm (.025 in) SS sheet perforated with 0.152 mm (.006 in) dia. holes @ 2.54 mm (.I0 in) pitch.

-5O9 80 x 700/80 x 80 Dynapore plus 0.635 mm (.025 in) SS sheet perforated with 0.254 mm (.OlO in) dia. holes @ 2.54 mm (.lO in) pitch.

-511 80 x 700 Dynapore plus 0.635 mm (.025 in) SS sheet perforated with 0.152 mm (.006 in) dia. holes @ 2.54 mm (.lO in) pitch.

-51 3 Nitronic 50 (5) plus 0.635 mm (.025 in) SS sheet perforated with 0.686 mm (.027 in) dia. @ 1.676 mm (.066 in) pitch staggered.

-51 5 0.635 mm (.025 in) 6AL4V titanium with EB perforated (elongated) holes 0.102 mm (.004 in) x 0.203 mm (.008 in) long, @ 1.27 mm (.050 in) between holes, and 2.29 mm (.090 in) between rOWS.

NOTES: (I) 181/7251 glass epoxy cloth laminate.

(2) Sew, cure and provide suction holes in layers 3 through 6 prior to attach of layers 1 & 2 as shown.

(3) Layer 1 to be diffusion bonded to Layer 2.

(4) Bond Layer 2 to cured substructure with film adhesive on Layer 3 contact areas only.

(5) 80 x 700 Dynapore made from Nitronic 50 wire.

/ I . The bowing of the Dynapore across the flutes is significantly increased at the higher speed of 224 m/s (500 mph), This velocity is higher than is normally required for testing commercial transport aircraft.

Electron beam perforated titanium on both glass-epoxy and graphite-epoxy sub-structures was acceptable for rain erosion environments at 179 m/s (400 mph). At 224 m/s (500 mph) slight bowing occurred across the 12.7 mm (0.500 inch) gaps in the sublayer with the flat specimens.

Bonding of EB Ti to either glass-epoxy or graphite-epoxy substructures must ° be carefully controlled to withstand heavy rain enviroments. No bond fail- ures occurred on the Dynapore and titanium specimens with the improved bonds at either test speed.

9.3.5 Compression Crippling Tests Test specimens were designed and fabricated to evaluate the local crippling strength of the candidate leading edge structure. Figure 9-59 shows the specimen design. Spot welding was used to simulate the diffusion bonded joint between the Dynapore outer surface and the perforated stainless steel sublayer.

The substructure design is similar to that used for Dynapore or electron beam porous surfaces. Testing of two specimens indicated a margin of safety in excess of 2 for crippling failures. Figure 9-60 shows a typical specimen tested to failure. The test showed that the leading edge compression design limit would probably be determined by Euler buckling between leading edge rib supports, depending on rib pitch.

9.3.6 Panel Joint Strength Static tests were conducted on 4 candidate configurations for chordwise joints.

The design loads shown below correspond to an ultimate wing strain of 0.102 mm (.004 in). This results in high test loads for test specimens with the stiffer titanium surface.

ffq hl.

FIGURE 9-58. TESTED DYNAPORE RAIN EROSION SPECIMEN _6.35 mm DIA HOLES ___25 IN.)

' i 127 mm I /0.025 mm(0.OOl IN.) TIR (0.50 IN.) / /' _3.18 mm (0.12 IN.)

_- 71.37 mm • (2.81 IN.)

19.56 mm //_SEWN TRUSS NODE -- 4 PLACES 80 BY 700/80 BY 80 DYNAPORE (0.77 IN.), 10.16 mm /SPOTWELD 0.25 O/C TO NO. 216 (04 IN.)

PERFORATED STAINLESS STEEL SHEET. BOND TO SUBSTRUCTURE WITH FM73 OR EQUIV.

-3 PLIES EACH 0.991 mm (0.039 IN.)

T300/5208 GR/EP T300/5208 GR/EP FIGURE 9-59. COMPRESSION TEST SPECIMEN

C

0_'. _'_-_'..',:_ .....

OF p,jC,:_ _.

Outer Surface Flush Bolt Size Compression Load Material 3/8 Dia 5/16 Dia Limit Ultimate N _Ib) N (Ib) Titanium - 1 -505 41,858 (9,410) 62,787 (14,11 5) Dynapore -501 -507 33,273 (7,480) 49,909 (ll ,220) Figure 9-61 and Table 9-15 show the test specimen and test schedule respectively and Figure 9-62 shows the compression test set up with a typical specimen ready for test. The test was designed to identify potential strength, fatigue and smoothness capabilities of joint configurations. Configurations -501 and -507 with Dynapore surfaces sustained their ultimate load without failure. Configur- ation -501 failed at I07 percent ultimate load by delamination of the GR-EP section due to bending adjacent to the simulated rib support. Configuration -507 sustained I07 percent ultimate load without failure (the limit of the test machine).

TABLE 9-15 TEST SCHEDULE JOINT SPECIMENS Dwg.

Dash No.

Type Test No. of Specimens Load/Record -l Static Compression 2 pairs (I) .004 (m/m) strain/check for bearing failures and note joint deflections relative to porous surface -501 (2) Fail ure/Pul t -505 -5O7 ORIG|NAL, :,,-_ `.= OF POOR QUALITY Co¢1_ P'IeL_S S,"o _.)

F'e=_F-oe-$1_='b _- S.

FIGURE 9-60, FAILED COMPRESSION SPECIMEN FIGURE 9-61. JOINT TEST SPECIMENS Configurations -I and -505 with titanium outer surfaces failed at 67 percent and 87 percent of their limit load respectively due to surface debonding. Post test analysis showed some bonding areas to be below-minimum thickness i.e., bond material starvation. In addition, surface cleaning/bonding procedures were suspect.

Specimens -l and -505 were therefore rebonded using an improved procedure for porous surfaces that included: Taping of the outer surface to prevent excessive thinning of the adhesive layer due to extrusion of the adhesive through the perforations.

Venting the autoclave vacuum to atmospheric pressure after the autoclave pressure had reached I03 kp (15 psi) to avoid weakening of the bond due to boiling of the adhesive.

Reducing the time to reach the curing temperature of 121°C (250°F) from one hour to a half-hour, in order to reduce the time spent with the adhesive at a lower viscosity.

Limiting the time taken between cleaning and etching of the titanium surface and application of the FM73 adhesive to one hour and limiting the time before starting the cure cycle to two hours, in order to minimize contamination effects.

After rebonding, the -505 specimen was retested. Failure occurred in the graphite epoxy supporting structure by separation of the stiffener, due to local bending starting near the simulated rib support, see Figure 9-6-3. The failing load was (69,948N) 15,725 pounds or Ill percent of the ultimate load.

There was no evidence of failure of the bond between the titanium surface and the graphite epoxy structure. The final results of static testing were therefore s_tisfactory for both the electron beam perforated titanium and the Dynapore The rebonded -l specimen was used for fatigue testing, see Section POOR QUALITY'

FIGURE9-62. JOINTCOMPRESSION TESTSETUP

FIGURE 9-63. FAILED COMPRESSION TEST SPECIMEN During static testing of the joints, 4 deflection gages were used to determine possible waviness and/or steps across the joint. Two gages were placed opposite each other at about 19 mm (.76 in) from the joint and the other two were located about 150 mm (6 in) from the joint gap on either side.

Waviness of about 0.25 mm (.Ol in) occurred over a length of 350 mm (12 in) with a step of about 50_um (.002 in) at the joint under loading conditions corresponding to steady cruising flight, l he effects of eccentric loading at the ends of the specimens may have increased the measured deflections.

9.3.7 Panel Joint Fatigue The design requirement for fatigue was established as a cyclic loading condition with the average compression strain varying from 0.016 percent to 0.160 percent, repeated for 120,000 cycles without failure.

The fatigue specimens were similar to the static test specimens described previously in Section 9.3.6. They were strain gaged on both sides and were tested at a rate of 7,200 cycles/hour in the test machine shown previously in Figure 9-62.

The rebonded EB titanium surfaced (-l) specimen described in Section 9.3.6 was fatigue tested initially to the required 120,000 cycles, then cumulatively at increasing strain levels until failure occurred after 170 cycles at the required ultimate static strength.

Max. Axial Load MaxT_Strain (Avg_Strain) Cycles at each Load Remarks N (LB) 17,793 ( 4,000) 1,430 (l,170) 120,000 26,689 (6,000) 2,190 (1,804) 20,000 (static limit load) 41,813 (9,400) 3,420 (2,790) 20,000 53,379 (12,000) 4,210 (3,300) I0,000 (static ultimate load) 62,720 (14,!00) 5,840 (4,620) 170 The average strains_shown in parenthese s are the equivalent wing design strains for the upper surface. The maximum strains measured were increased by local bending near the joint due to load path eccentricities. The fatigue strength far exceeded the fatigue strength requirements.

Eventual failure was due to delamination of the outer surface as shown in Figure 9-64, which was taken with the specimen being subjected to an axial compressive load of 17,793N (4,000 Ib).

The following results were achieved on the -501 Dynapore specimen: Max. Axial Load Max._,Strain (Avg./_Strain) Cycles at N (Lb) each Load Remarks 13,256 (2,980) 1,600 (I,025) 200,000 29,047 (6,530) 3,200 (2,361) 200,000 (Static limit load) 35,052 (7,880) 4,000 (2,963) 200,000 44,482 (I0,000) 5,080 (3,530) 33,000 'T' support failure The results show that the -501 joint greatly exceeded fatigue test requirements without failure.

Test results for the -507 Dynapore specimen were as follows: Max. Axial Load Max. LLStrain (Avg_A_Strain) Cycles at Remarks N (Lb) / each Load 18,238 (4,100) 1,600 (I ,065) 200,000 7.94 mm (5/16") 39,144 (8,800) 3,200 (2,536) 63,000 bolt failure The bolt failure originated at the junction of the countersunk head and the shank. Although the use of a smaller bolt resulted in a large reduction of fatigue strength, the -507 joint still exceeded the design requirement.

The overall fatigue test result showed that all of the fatigue test specimens exceeded fatigue requirements. Specimens with the 9.55 mm (3/8") bolt were far superior than that with 7.94 mm (5/16") bolt due to induced bolt bending. The f- results also show that the improved bonding technique for the E.B. perforated titanium surface is satisfactory under both loading conditions.

9.3.8 Panel Development Testing

Fabrication development of the woven steel Dynapore faced fiberglass LFC glove panel included work to meet impact, porosity, bonding and environmental durability requirements. After the panel materials, configuration and fabrication procedure were reasonably well established, mechanical testing of coupons and structural elements was done to determine strength and stiffness properties of the panel constituent materials, both separately and in combinations, representative of the working design. Peel tests after environmental exposure were used to check the Dynapore bond strength during development.

Dynapore (80 x 700/80 x 80) static stiffness and static strain properties were explored. The Dynapore bonded to a perforated "S '' glass sublayer was tested for strain recovery, effective stiffness and fatigue strength. Panel assemblies were tested for bending strength/stiffness, compression strength/stiffness, and shear strength/stiffness with alternative core and materials orientations. Two large 356 x 610 mm (14 x 24 inch) shear panels were tested for buckling resistance and to failure with core orientations of 90 ° and 45 ° to the shear load direction.

Two panel assemblies 305 x 762 mm (12 x 30 inch) in size were compression tested to assess the strain compatibility between a simulated graphite/epoxy wing cover and the porous glove panel attached to it. One assembly had panel fluting at 90 ° and another had fluting at 45 ° to the graphite support stiffeners.

Figure 9-65 shows a 305 x 762 mm (12 x 30 inch) panel before final assembly, and Figure 9-66 shows a panel assembly in the test machine after successful loading to 0.006 strain.

Single fastener joint specimens, representing the initial design for the chord- wise glove panel joint, were tested in static tension and tension fatigue. See Sections 9.3.6 and 9.3.7. Static failure occurred at 185 percent of design ultimate. Fatigue failure at maximum limit load occurred at 356K cycles, which is three times the required life.

f

OF PO01_QUALITY

FIGURE 9-64. FAILED JOINT FATIGUE SPECIMEN COMPRESSION TEST PANEL FIGURE 9-65.

A 254 x 508 mm (I0 x 20 inch) glove panel with a centrally located joint was tested to determine the ability of the surface and joint to conduct lightning discharge currents up to 144 kC 940 KA) magnitude and lightning restrike to 378 kC (I05 KA). The resulting damage and discharge capability were acceptable from a safety aspect, but the panel would need to be replaced in service.

9.3.9 Structural Demonstration Component A large compression panel was tested to verify structural adequacy and fabri- cability of the LFC porous glove panel/graphite epoxy wing box system, when subjected to compression loading to the required strain levels of 0.004. The design incorporates a simulated wing panel with the porous panel attached to the external stringers, designed to have strain compatibility with the wing structure. The panel size was 2070 x 730 mm (81.5 x 28.75 inch) with an area of approximately 1.4 m2 (15 ft2). The test panel was configured for installation in an existing Douglas test fixture for testing on the million pound Baldwin test machine as shown in Figure 9-67.

The design approach was to first design the graphite epoxy J-stiffened compression panel using a simulated rib spacing of 762 mm (30 inch) to control the Euler buckling length_and a stringer height of 38 mm (1.5 inch) with a stringer spacing of 178 mm (7.0 inch). This is representative of wing box structure at the 70 percent span location of the aspect ratio lO Baseline aircraft wing. The ultimate design strain level of 0.004 was chosen because of an anticipated design constraint for composite wing designs due to damage tolerance and other design requirements. The simulated wing panel was designed in the T300/5208 materials rather than the stiffer mixed graphite concept (T-3OO/GY-70 fibers) mentioned in Section 9.1.6 to reduce the test load associated with the strain target for the glove panel.

2,84 OF POOR QUALITy FIGURE 9-66. COMPRESSION TEST PANEL SETUP FIGURE 9-67. LARGE COMPRESSION PANEL TEST ARRANGEMENT The LFC porous glove test panel incorporated the design features of typical longitudinal panel splices at the panel edges. Field fasteners, panel inserts, airflow collection stiffeners and air dams were included to evaluate the structural capabilities of these features and to establish manufacturing feasibility and smoothness of a large panel and the fluted panel stiffening was representative at 45 degrees. A Dynapore porous surface, favored at this time, was used for this test panel. Field fastener spacing was at 15 inches maximum. The combined assembly is illustrated in Figure 9-68 which also shows the strain gage locations.

Computer programmed analyses were used to design both the graphite panel and the porous glove panel. Skin buckling, flange crippling and general buckling were critical elements. Analysis showed that the graphite epoxy panel would require a lod of 1,786 kN/m (I0,200 pounds/inch) to reach .004 compression strain. The overall glove panel analysis was conducted utilizing an orthotropic sandwich panel analysis, named TRUSS especially programmed for the glove panel.

Supplementary calculations covered element crippling and the panel/graphite structure interaction. The analysis predicted that the glove panel would accept 19 percent of the total test load after graphite epoxy material had been introduced to stiffen the glove panel webs and backfacing.

During testing, buckling of the graphite-epoxy simulated wing box structure occurred between the integral stiffeners at 90 percent of ultimate load and failure occurred at 97 percent of the ultimate design load. Figure 9-69 shows the LFC panel surface in the test rig after failure. Figures 9-70 and 9-71 show failures at the inner face of the LFC panel and the mating face of the GR-EP wing box panel respectively, after disassembly.

Initial delamination of the Dynapore surface was observed to occur at a load equivalent to approximately "IG" flight load. This premature failure was found to be caused by pGGr surface cleaning procedures prior to Dynapore bonding.

Subsequent changes to porous surface designs and bonding methods are expected to correct this failure mode.

ORIGIh,tAL PAG[:.' [$ OF POOR QUA_,TY

i;.: , 1,o

O'OVE,EST"_'k" /._ ', 7_.,, 0.28 m 028 m (10.9 IN,) (10.9 IN)

T

oLi

0.69 m (270 IN) _I THROUGH 4 035 rn 113.8 IN.)

r10 THROUGH 13 14 THROUGH 17_ [- 069 m 0.64 m [27,0 IN} 125.1 IN.I

_1

STRAIN GAGE LOCATIONS (18 REQUIRED) FIGURE 9-68. LARGE COMPRESSION TEST PANEL FIGURE 9-69. LFC DEMONSTRATION PANEL FAILURE ' "=k_ FIGURE 9-70. LFC PANEL FAILURE \ o \ • \ FIGURE 9-71. GRAPHITE-EPOXY WING BOX P" q.URE ORIG/_AL p_-_.;- OF POOR QUALITY The average strain gage readings for the GR-EP wing box panel and the LFC panel are shown in Table 9-16, compared to predicted values at increasing loads to failure. Note that the actual strains are all below predicted values at any given load, indicating that the modulus and/or gross-sectional area of the article was higher than anticipated. The results are plotted in Figure 9-72.

This shows that the LFC panel was being strained more than the graphite-epoxy wing box prior to the onset of buckling.

The initial failure occurred in the simulated wing structure with a strain level of 0.0037 in the LFC glove panel. It cannot be determined whether the glove panel would have reached the design ultimate strain level of 0.004 but the limit strain level of 0.0267 was easily exceeded.

TABLE 9-16 DEMONSTRATION PANEL STRAIN HISTORY Strain x 106 Load - % Ultimate Strain 97% 9O I0 20 30 40 50 60 70 80 Locations (Failure) GR-EP 288 624 1024 1372 1684 2019 2349 2683 2890 Inner Panel U_ (Average) r_ LFC r_ ._j _- 388 716 1062 1403 1867 2310 2670 3027 3410 Panel -e-,, c) r_ (Average) c> .._.-} ,., c,r) c> o 390 780 1170 1560 1950 2340 2730 3120 3510 Predicted f

9.4 WINGLEADING EDGE DESIGN INTEGRATION

The initial problem addressed in leading edge design was to define geometry

and construction requirements for integration of the air suction system

and a freezing point depressant (FPD) liquid dispensing system in the same structure. A metallic structure was assumed necessary for impact toughness.

Approaches included: (I) suction slots in locally densified porous material with fluid transpiration through remaining porous areas. (2) All porous material with suction and fluid transpiration through alternate porous strips running spanwise. (3) Spanwise porous strips let-in and alternating with solid, slotted material.

The most acceptable approach appeared to be the all porous nose, with an integrated fluid/air supply structure behind the surface. Various sandwich and machined plate concepts were suggested for fluid distribution. One of the controlling fluid system design criteria was the necessity to withstand 345 kPa (50 psi) fluid pressure behind the porous transpiration surface.

9.4.1 Leading Edge Design - Phase I A section at the 40 percent semispan air collection station of the baseline wing was chosen for the design integration study. The resulting arrangement for upper surface suction only, and with a retractable shieldjis shown in Figure 9-73.

A sandwich construction with a 3.17 mm (.125 inch) cell honeycomb core, 6.35 mm (.250 inch) thick, supports the outer facing of Dynapore backed up by perforated impact sheet. A complex inner surface is grooved and ducted to provide fluid or air suction at the surface. Materials and fabrication procedures were not well defined at this stage, so the drawings should be regarded only as an initial indication of space and volume requirements for the three integrated systems - structural, air and fluid. Aside from fabricability, the main structural design challenges were impact durability of the outer surface and the peel stresses within the sandwich due to the pressurized fluid. The construction concept assumed a one piece seamless outer surface between approximately 5 percent chord on the lower f - surface to the joint at the front spar upper flange.

ORIG'y :!; ....

OF POO,,"_ Q'J/_,,i-f 0.004 / / / 0.003 //" f .--' LFC PANEL /" AVERAGE STRAIN 0.002 APHITE-EPOXY INNER WING BOX PANEL 0.001 0 10 20 30 40 50 60 70 80 90 100 ULTIMATE LOAD (PERCENT) _._._ DEMONSTRATION PANEL AVERAGE STRAIN HISTORY FIGURE 9-72.

_LFC SURFACE HONEYCOMB i CONSTRUCTION PROTECTION FLUID DISPENSER T HONEYCOMB IN DE-ICING AREA TO PROVIDE FLUID PRESSURE BALANCE VIEW 3 TIMES SIZE /--LFC SURFACE "_ / HONEYCOMB _"_j/ CONSTRUCTION / ' LDE-ICING SURFACE HONEYCOMB CONSTRUCTION TYPICAL WING NOSE CAP SECTION D.D f "4 ' FIGURE 9-73. LFCLEADINGEDGE-PHASEI Before continuing with leading edge structural design, the following complementary studies were undertaken (as reported in Section II.0); o Leading edge protection fluid dispensing testing in the wind tunnel to observe fluid flow characteristics and effects on the surface.

o Preliminary environmental contamination and cleaning tests.

o Examination of design parameters to determine required pressure drop and suction flow characteristics of porous surfaces in the leading edge region.

o Leading edge protection shield bug deflection analysis and consideration of fluid dispensing requirements.

o Full scale insect impingement tests on a leading edge shape in the icing tunnel.

Considerable effort was expended in evaluating alternatives to the Dynapore/ perforated glass honeycomb concept for the critical nose cap region of the leading edge. Improved designs included metallic honeycomb or closely-spaced corrugated surface supports with brazing, welding or diffusion bonding techniques to join surface, subsurface and support structure, and organic bonding. At this stage, a cursory ranking of I0 concepts put in first place a design featuring an electron beam perforated titanium surface, diffusion bonded to slot-perforated backing sheet on a corrugated air collection support structure.

It ranked high in toughness, reliability and elasticity.

9.4.2 Leading Edge Design - Phase II A decision was made to concentrate on using the Dynapore/glass Lockcore glove panel construction in the leading edge region. It was felt that continuity with existing porous panel development would be advantageous. To be consistent with the configuration trade study favoring upper surface suction only, Section 5.6.1, the fluid pressure design requirement was avoided by replacing the t fluid protection system with an extendable shield. The objection to chordwise- oriented Lockcore construction, which is not feasible from a fabrication stand- point with curved leading edge contours, was answered by turning the fluted core spanwise, requiring it to taper in cross section with percent chord lines.

Figure 9-74 indicates the first schemeof this type, suitable for a flight test

airplane leading edge. This design featured perforated fiberglass under

Dynapore; the glass being thick enough to accept major impacts safely. The

hinged shield shape although far from ideal from an aerodynamic point of view,

could be used to provide insect impingement protection and avoids the cost and

complication of a variable camber mechanism.

Concept sketches for a production LFC airplane were generated as shown in

Figure 9-75. A "D-duct" accon_nodatessuction for the nose region. Elsewhere

suction air is collected from the spanwise flutes into round collection ducts

running through support ribs. Surface Concept C shows continuous perforations

under the Dynapore. In Concept B, perforations occur only under those Dynapore

areas where strip suction is desired, leaving areas for sewing, fastening or

adhesive bonding in the intermediate blank strips. Surface Concept A suggests

another alternative using larger perforations under small perforations in the

surface materials to improve surface backup for impact resistance.

Figure 9-75 shows a design providing easier air collection from the spanwise

flutes. It features integral collector ducts which also connect the surface

to the supporting ribs. The flutes at the surface would be discontinued at

panel splices to control spanwise inflow-outflow. The airflow between the small

and large ducts is through metering holes to control the flow. Figure 9-77

illustrates an alternative construction that provides a stiffer complete sandwich

structure for the outer surface.

9.4.3 Leading Edge Fabrication

For an initial check on fabricability of the Dynapore/glass laminate construction for the leading edge region, a fiberglass tool was laminated using an existing control surface part to provide the shape. This tool was used for the initial rubber mandrel packing and pressurization fabrication trials. The skin and corrugations were jointed by sewing to strengthen the surface joint.

f.,'- _!F P-t]<,'. '""; >/i'V DYNAPORE --7 /--"E" GLASS

OPEN 1 J_ I\ , ] _-"/- I

/ //:, ..... ...... 7 ........ \ i-

/

/L-POSITION B FRONT SPAR PLANE 83 GE_, 2329_ FIGURE9-74. LFC LEADING EDGE-PHASEII _--- PERFORATIONS IN TITANIUM OR /---0.125 DIAMETER HOLES \ALUMINUM SHEET FOR EXTERNAL /40 PERCENT OPEN _SURFACE /STRIP PERFORATED "S" GLASS \ I /--- DYNAPORE r---DYNAPORE \ __/LFC SURFACE /LFC SURFACE k , _u,.,..,-o.oo5 f,-_, _',_' \ _ _ " _ [ ' 1 _ _ _ ' _ _ _ / ' - - 0 ' 1 2 5 D I A M E T E R H O L E S _ {',, " ' r 0.0064 _'/, _, ' :v _40 PERCENT OPEN HOLE _. "_ t, "' 4' _ _ "_' S _ -Q,.,--k"i -- \ L-o 004 DIA HOLES PERFORATED "S" GLASS ALTERNATIVE LFC SURFACE UNDER SURFACE LAYER SHOWN UNDER SURFACE LAYER SHOWN PERFORATION CONCEPT "A" STRIP CONCEPT "B" CONCEPT "C" /'--SUCTION DUCT _._.__ 7- _'-

METER,NG NOZ_LE_ 2__:_I

.,G.L,'T ..</,, .,,_TP_r?-:: ,. h :(] k_),,i

_SHIELD DEPLOYED/ i., /_"-. " / _ =..j, ,_=j/ _ _--- >_%, ] I , i 1_J "" '"

,. \ \ t -7-------.___.__ _

" _"' "_' "/ SHIELD STOW E D 8_GE N-232¢=3 FIGURE 9-75. LEADING EDGE DESIGN - SEPARATE COLLECTOR DUCTS OF Poor QJALiNr METERING HOLES-_ ,--SUCTION DUCTS '__

/\ CONTINUOUS -_-_-_f'_'S

METERING NOZZLE--_ / \ _-J_,_"YJ'J_/_- _.__-'

,>_._____- --;_. ,i, II

(_><//'<__ \-. \

"J 83.GFh 232_ FIGURE9-76. LEADING EDGE DESIGN - INTEGRAL COLLECTOR DUCTS 83GEN 23267 I FIGURE 9-77. LEADING EDGE BOX DETAIL AT FRONT SPAR The co-cure method of fabrication using 345 kPa (50 psi) autoclave pressure produced promising samples and no problems were apparent with that geometry.

The test samples are shown in Figure 9-78. One specimen included a second duct layer similar to the design shown previously in Figure 9-76.

Further fabrication experience was gained in producing the wind tunnel model, as described in Section 9.5.

9.4.4 Improved Leading Edge Glove The porous surface development was finally concentrated on either a Dynapore surface diffusion bonded to a perforated stainless steel sublayer or electron beam perforated titanium. Figure 9-79 shows a design that is suitable for either surface with integral plenum chambers. The porous surface is self- supporting over the integral suction duct span. This type of construction was used for the initial EB perforated Ti surfaced wind tunnel models.

Figure 9-79 also shows a later version of the perforated strip concept designed to reduce weight. The plenum chamber was formed by inserting an aluminum mandrel below the outer surface and by recessing the silicone rubber flute mandrel to conform to the aluminum mandrel shape. Although this configuration did reduce weight, a satisfactory molded leading edge could not be fabricated without some rework after initial curing. Increasing pressure in the autoclave to force the glass-epoxy structure against the molded surface at the flute nodes was not successful. Unfilled voids still occurred along the nodes, which were unacceptable for the subsequent surface bonding process. The wind tunnel model was repaired by filling the voids with epoxy resin, but this would be unacceptable for a flight structure.

ORIGti"JAL i:'/,:. _._

OF POORQUALI'i"y

FIGURE 9-78. LEADING EDGE - INITIAL CO-CURE ASSEMBLIES

_.," l_'_, ._----- ----

/ _ 0_)2011 (6AL 4V)-_ t (,';;_'".'.\ - \_! _ i POX31' /_f_ II \\ II

_",_. o_o_Guess E ---

FIGURE 9-79. LEADING EDGE STRUCTI' E - Ti SURFACE Recommendations for future Phase II efforts include studies to omit the plenum chamber and revise the flute geometry to incorporate a flat for bonding at the flute nodes. The use of aluminum mandrels for the flute cavity and rubber mandrels at alternate nodes would permit the precise location of flutes and control of molding pressures during initial curing. The need for sewing would be eliminated, saving production time and cost. This type of construction could also be used for the wing box LFC panels.

9.5 SWEPT WING WIND TUNNEL MODEL - POROUS PANELS Fabrication of the LFC panels for the swept wing wind tunnel models resulted in a continuation of design and development work on the leading edge and wing box structure glove panels. The initial design for the LFC panels utilized Dynapore porous surfaces as shown in Figure 9-80. An 80 x 700 Dynapore outer surface was diffusion bonded to an 80 x 80 inner mesh. The Dynapore surface, produced by Michigan Dynamics Company, had an initial porosity of 0.813 m3s-I/m 2 (160 SCFM) at 670 Pa (14 PSF) produced by roll calendaring. This surface was organically bonded to a porous fiberglass substructure designed to reduce the porosity to 0.152 m3s-I/m 2 (30 SCFM) at 670 Pa (14 PSF).

For initial tunnel/model calibration, a non-porous surface model was fabricated using the same design as the porous model, but with a solid fiberglass surface.

An alternative leading edge panel was developed using an electron beam perforated titanium (EB Ti) surface. This design used a perforated strip concept. Each perforated strip contained several rows of holes feeding suction air to the integral channel of the glove panel_as shown previously in Figure 9-79.

To form the LFC surface panels, a high temperature molding tool was designed and fabric_tzd using a master plaster of the desired aerodynamic shape.

Figure 9-81 shows the inside surface of the substructure located in the forming and curing tool. Figure 9-82 is a close-up of the leading edge area showing OF POOR QUALITY /_ DYNAPORE/POROUS GLASS EPOXY SURFACE /---- DYNAPORE/PERFORATED GLASS EPOXY SURFACE I'_---- 12'96----_I _--GLASS-EPOXY TRUSS CORE STRUCTURE FIGURE 9-80. SWEPT-WING WIND-TUNNEL MODEL LFC SURFACE FIGURE 9-81. WIND TUNNEL MODEL SUBSTRUCTURE IN MOLDING FORM TOOL the nesting of silicone rubber mandrels to form the internal ducting flutes.

Figure 9-83 is a view of the cured fiberglass substructure ready for bonding to the porous Dynapore outer surface.

The joining of porous Dynapore sheets in preparation for bonding to the sub- structure is shown in Figure 9-84. A resistance welding technique was used to provide a lap welded joint with the outer surfaces butted together. Joining strips of 80 x 700 mesh Dynapore were used as butt straps in order to provide continuous porosity across surface joint areas. Surface smoothness and edge hold down were primary requirements.

Considerable effort was needed to develop a process to fabricate and bond Dynapore satisfactorily to the porous sublayer. Excessive variability of the porosity of the fiberglass sublayer was a major problem. A porous surface with a porosity of 0.152 m3s-I/m 2 (30 SCFM) at 670 Pa (14 PSF) with no more than 30 percent varability over the surface was finally obtained. This was achieved by selecting fiberglass material that was previously checked for high uniformity of porosity after processing. In the event that the relatively high porosity of 30 SCFM was not acceptable for the achievement of laminar flow in the leading edge section, further porosity reductions could be made as required by using suction area reductions through the use of non-porous spanwise strips. However, the use of these strips was not required. During leak tests on the completed LFC panels, it was found that the flute walls were porous. This was corrected by injection and coating of the plenum walls with polyurethane.

For the alternate leading edge panel, an EB Ti surface panel was used with 0.114 mm (.0045 in) wide by 0.193 mm (.0076 in) long holes elongated spanwise and spaced at 1.27 mm (.050 in) on centers in staggered rows 1.27 mm (.050 in) apart. The strip porosity provided by this structural concept created an area blockage of about 50 percent for the test article as shown previously in Figure 9-79. This blockage resulted in an acceptable suction flow for LFC.

The elongated hole width was the smallest size available from the Farrel Company, Connecticut, at the time of testing. The elongated hole shape was used to reduce the possibility of particles large enough to trip laminar flow being trapped on the surface by suction.

3oo OF POOR QUALITY FIGURE 9-82. WIND TUNNEL MODEL LEADING EDGE AFTER CURING FIGURE 9-83. LEADING EDGE WTM - EXTERNAL SURFACE BEFORE DYNAPORE BONDING The porous outer surface was preformed by rolling prior to bonding. Figure 9-85 shows the completed leading edge after bonding the perforated surface to the fiberglass substructure. It is ready for trim machining and fitting to the wind tunnel model structure.

Some blockage of the porous surface near the ends of the alternate leading edge was detected on the first EB Ti leading edge test articlesdue mainly to bonding material flow. The central section of the leading edge did not have any blocked areas.

Additional porous leading edge insert panels, as described in Section 8.2.2 and illustrated in Figure 8-26, were produced subsequently. The type of construction is shown in Figure 8-28. Far better control of porosity was obtained with this design and wind tunnel results, particularly for the EB perforated surface,were greatly improved.

ORIGII_,_AL _t-_ OF POOR QUALII'Y iiiiiiiiii FIGURE 9-84. DYNAPORE LAP SHEAR JOINT WELDING TECHNIQUE FIGURE 9-85. LEADING EDGE PANEL WITH POROUS SURFACE 9.6 CONCLUSIONS & RECOMMENDATIONS - STRUCTURES l • An LFC wing structure consisting of an outer glove panel supported by a graphite/epoxy primary wing box structure is a satisfactory arrangement.

In addition to reducing wing weight, the use of graphite epoxy material for the primary structure reduces the strain level imposed on the glove panel.

The loads at panel joints are correspondingly reduced.

2. The porous surface can be bonded satisfactorily to a fluted fiberglass substructure to form a stiff LFC glove panel.

.

Integral suction airflow ducting can be provided by the flutes in the glove panel, and in the space between the glove panel and the main wing box, utilizing external stiffeners to create separate ducts.

° Electron beam perforated titanium sheet material is the most desirable porous LFC surface of those investigated, considering strength, damage resistance, rain erosion, uniformity of porosity and ease of fabrication.

5. An all titanium superplastic formed diffusion bonded glove panel is a practical possibility that requires further study.

o A wing aspect ratio of lO is practical but results in a significant aeroelastic weight penalty for adequate roll control stiffness. With 30 degrees sweep, wing bending stiffness is the most significant parameter and the graphite fiber selection and orientation should be selected to maximize this characteristic within practical limits• 7. The aeroelastic penalty for an aluminum wing would be far greater.

8. Wing f]uLLer was not critical.

ORI_31NAL p_ OF POOR QUALITY , Panel joint design must consider local surface deflections due to eccentricities of load transfer, and external smoothness at joints and fasteners. The number of chordwise joints, which are more highly loaded, should be reduced to a practical minimum. Fastening should preferably be from inside to avoid surface imperfections. Joints can be angled relative to the streamlines to reduce the length of surface airflow without suction.

I0.

The glove panel contour can be controlled by using external molding tools.

The finished surface accurately reproduces the mold shape and any slight imperfections in the tool surface are ironed out by the titanium sheet surface of the glove panel.

9.7 REFERENCES - STRUCTURES Douglas Aircraft Company, "Development of Technology for Fabrication of Reliable Laminar Flow Control Panels on Subsonic Transports," NASA CR-145125, October 1976.

(Prepared under Contract NASI-14408).

t I0.0 SUCTION SYSTEM I0.I INTRODUCTION The goal of the suction system study was the design of a simple, light-weight system to remove the required amount of boundary layer air from those areas of the wing surface where laminar flow is to be maintained. To achieve this aim, the suction system was integrated with the aircraft structure as much as practical. A porous or perforated surface was assumed for the wing surface.

Suction air compressors and their power units are located below the wing aft of the rear spar, and the air is exhausted overboard at approximately freestream velocity for minimum drag and suction power required.

As characteristics and performance of various aspects of laminar flow were discovered in the course of the study, the configuration was modified to take advantage of design improvements or to correct newly uncovered problems. For i example, when it was shown that approximately the same drag reduction could be achieved by applying LFC to the upper wing surface back to 85 percent chord and eliminating suction on the lower wing surface, this concept was adopted and the manifolding and compressor installation were changed to provide one- step compression for the suction air.

10.2 REQUIREMENTS Suction flow requirements and surface pressures were based on aerodynamic calculations, discussed in Section 7.1. Typical minimum suction flows (expressed as suction velocity) necessary to maintain laminar flow on the wing are shown in Figure I0-I, together with the corresponding chordwise variation of pressure coefficient. The mass flows were calculated from the velocity, V w, and free- stream density, incompressible relationships having been assumed in the derivation of Vw. For suction system sizing for the advanced LFC aircraft, the calculated mass flows were increased by a factor of 20 percent to provide for local areas of suction greater than the minimum and to allow a margin for off-design operation.

PRECEDING PAGE 8LANK NOT FILMED 5o .

At the beginning of the LFC study, one of the ground rules established was that no redundancy would be provided in the suction power system. The main- tenance of laminar flow on the wing is not a safety item. To provide redundancy for the case of a suction system power unit failure, crossover ducting would be necessary in regions where volume is already at a premium. Also, the size of each power unit would be doubled, which is an additional weight and drag penalty. The system was designed, therefore, with a separate unit in each wing.

In case of loss of laminar flow on one side of the aircraft, the asymmetric drag situation can be handled easily with the standard airplane controls, and the additional trim drag penalty is slight.

In the later part of the study, a contamination-avoidance system was defined that requires application of a liquid to the wing surface during flight operations where potentially contaminating conditions exist. To remove remnants of the liquid before trying to initiate LFC, a positive pressure significantly greater than the suction pressure is applied below the surface. The resulting "clearing" airflow is high compared to the suction flow rate in the opposite direction. The requirement tO provide this "purging" capability affects the system design and may determine the size of some components.

10.3 POROUS SURFACE CHARACTERISTICS The desired flow characteristics of the porous surface were originally defined as: (I) having sufficient pressure drop through the surface during LFC operation so that no outflow of air is induced because of surface pressure variation; (2) having relatively low pressure drop through the surface even at the higher LFC flow rates; (3) having a sufficiently open surface to minimize clogging during low-level flight operation.

ORIG!Ni_L Pi_,k_, r,'_;_

OF POOR QUALI'r'_ The porosity, Q, has arbitrarily been defined as the flow, under standard conditions, in cubic feet per second per square foot with 670 Pa (14 pounds per square foot) of pressure differential across the surface.

In the region of the LFC system, item (I) above is important chiefly in regions where the chordwise pressure gradient is steep and the suction velocity is moderate. Figure 10-2 illustrates the effect of porosity in such a region.

Air travels through the porous surface into the flute or other collection component below, where the pressure is designated Psub" Some of the surface area is blocked by the attachment of the collecting structure. If VC is the local velocity into the surface and s is chordwise distance along the surface, ideally: -_, .I _ ] cls = . . \)tl, where Vth is the theoretical suction required, determined by aerodynamic criteria.

As the surface pressure at point 1 is higher than at point 2, f

> \fC

I z To prevent outflow, Psub must be equal to or less than P2" The maximum allowable porosity, Qmax' at which outflow is avoided without the suction flow exceeding the theoretical value is illustrated on the left-hand side of Figure 10-2. Here Psub = P2 and V2 = O. The dashed line represents the case where the porosity is greater than Qmax" To prevent outflow, Pp, with the result that Psub I The right-hand side of Figure 10-2 shows that velocity profile resulting when the surface porosity is less than Qmax" Here \ ORIGINAL P,_C_Z _$ OF POOR QUAt !TY --1.2 DESIGN POINT CQ FOR CHORD --5.05 m (16.58 FT) --1.0 --0.8 CP N , CQNX 10 s --0.6 --0.4 --0.2 l = 1 I ! I l I 0 10 20 30 40 50 60 70 80 90 PERCENT CHORD _._ ;L_ ....

FIGURE 10-1. WING SURFACE PRESSURE&SUCTION REQUIREMENT SURFACE -- PRESSURE DISTRIBUTION

\

Pt PRESSURE ----'j _V t FORQ<QMAx FOR Q > QMAX I I VELOCITY FOR Q = QMA;' THROUGH SURFACE "" _VTHEOR POROUS OR PERFORATED SURFACE (a) Q = QMAX ANDQ>QM_ (b)Q< QM_ B3_,E _, 23215 EFFECT OF SURFACE POROSITY ON SUCTION VELOCITY VARIATION FIGURE 10-2.

J This condition gives a smoother velocity profile than that achieved with the more porous surface and increased tolerance to off-design conditions.

The widths of the open area and of the blocked portion of the surface both affect the required porosity. Increasing the width of the open area subtended by a flute makes a lower porosity (less open surface) necessary, to match the greater differential in surface pressure that is subtended by the flute.

Increasing the percentage of blocked area results in a greater allowable porosity, as the airflow per unit of surface, or V£, is increased, thus increasing the pressure drop through the surface.

During the study of contamination-avoidance procedures, it was found necessary to provide for "purging" the porous surface of the liquid used to prevent contamination before initiating suction for LFC. The pressure differential necessary to purge in a reasonable time was found to be about 6900 Pa (I psi), see Section 10.7. This pressure differential is an order of magnitude greater than the pressure differential, in the opposite direction, during the suction mode. Surfaces with high porosity have very large airflows during purging.

The necessity of a porosity that minimizes the effect of the purging requirement on the size of suction system components becomes a fourth consideration on the above list.

For the flute widths and amounts of blockage that are acceptable from structural considerations, a porosity of 0.061 to 0.076 m3/s/m 2r 12 to 15 SCFM/ft 2) at 670 Pa (14 PSF) at the leading edge gives the condition where no outflow exists with the flows approximately equal to the theoretical requirements. If more open porosity were used, more suction would be requiredjincreasing the suction airflow above that required for LFC in order to increase the pressure drop through the surface and avoid outflow. A porosity of about 0.15 m3/s/m 2 (30 SCFM/ft 2) at 670 Pa (14 PSF) is proposed for the surface above the wing box and aft of the rear spar. Figure 10-3 shows typical flow characteristics for a porous surface.

f- 0.134 m 3 . s--l/m2 (STD AIR) AT z_P - 670 Pa SURFACE MATERIAL WITH pOOR OF QLjALITY 0 F_ TEST DATA Wa kg Asurf ' • s--l/m2 I II IIII11 I I I llJJll I _Jllll _ _ _Jll_ 0.01 10 100 1,000 10,000 100,000 83-GEN-23217 oz_P "°,Pa P SURFACE MATERIAL WITH 26.4 SCFM/FT 2 ATZ_P= 14 LB/FT 2 0.5 0.1 -- D

y

Wa/Asurf AVER OF TEST DATA LB/SEC/FT 2

/

---- EXTRAPOLATED

/

0.01 -

/

/

/

/

I I IIIIII I I tlllllll 0.001 I I I I I l lllll 1 III I lill 1.0 10 0.1 100 1,000 O 83_GE N-232)8 a-_ P -._.., LB/FT 2

@

FIGURE 10-3. TYPICAL FLOW CHARACTERISTICS- POROUS SURFACE 10.4 DUCTING It was desired to have a simple, light-weight, and low-maintenance suction system. To achieve this, the ducting was designed to make use of the aircraft structure wherever possible. In sizing the ducting, the object was to obtain a reasonable compromise between duct size and pressure drop through the system.

The pressure drop in the system affects the size and weight of the compressor and its drive unit, and the amount of fuel consumed by the LFC system. Figure 10-4 shows the effect on these parameters of the pressure loss in the system. A pressure at the compressor face of 80 percent of the minimum surface pressure was taken as a goal in the design of the suction system. This provides an allowable 20 percent pressure drop in the system for air from the wing areas where both the surface pressure and the suction requirements are quite low.

As was shown in Figure I0-I, in areas where the suction quantity requirements are higher, the surface pressures are generally higher, resulting in an increased allowable pressure drop where the greater pressure drop is likely to occur.

At the beginning of the study, both chordwise and spanwise directions were considered for the initial air collection, as shown conceptually in Figures 10-5 and 10-6. The chordwise method of Figure 10-5 requires a large percentage of the available space ahead of the front spar for ducting the collected air to the compressor and ducting much of the air across the front spar for its entire length is not structurally efficient. The spanwise method makes use of external stiffeners on the main box surface to provide integral spanwise ducting and was selected for the advanced LFC aircraft, although it does require holes in the main box for ducting the suction air to the suction pump below the wing.

The suction flow collection system is integrated with the wing structure, as shown in Figure 10-7. The boundary layer air is drawn through the porous or perforated surface into flutes which are bonded to the surface material. The flutes provide both structural stiffness and flow passages. The air travels through the flutes to an opening that leads to an opening into an auxiliary duct. The main flow channels are formed by the surface panel, the integral stiffeners and the wing box skin.

OF POOR QUALITY 3.0

\

\

\ 2.5 "\

\

\ '\ 2.0 '\ f,---COMPRESSOR WEIGHT RELATIVE \ COMPRESSOR "\ REQUIREMENT - "_,_ \'\, _o 0.5 I I I I t I , 0.5 0.6 0.7 0.8 0.9 1.0 PRESSURE AT COMPRESSION FACE/MINIMUM SURFACE PRESSURE EFFECT OF SUCTION SYSTEM PRESSURE DROP ON COMPRESSOR FIGURE 10-4.

REQUIREMENTS (UPPER-SURFACE-ONLY CONCEPT) OUTER WING SECTION _IFRONT SPAR CHORDWISE BAYS-_BA Y __ REAR SPAR SECTION A-A (NOT TO SCALE) SUBSECTION--/ ..-----___ _, f POROUS SURFACE

\_, _ \\_ _\\ _ v,_w

@

FIGURE 10-5. CHORDWISE AIR COLLECTION SYSTEM

ORIGIP_AL r._. .... r,_.

OF POOR OOALII"Y SECTION OF OUTER WING.

T UPPER SURFACE FLOW CHANNELS TRAILING E "_'_._ J (16) DGE SECTION A-A (NOTTO SCALE) SECTION B-B .._.. POROUS SURFACE _ B PLENUMT_ /"POROUS SURFACE SPACERS 7 Y/_'//////////_/,¢'/////////////////l_ _/_/.j_/_//////////////////////J v////A FLO W CHANNEL , / -7---FLOW METERING _ B/_uu_;:RING _, z_ METERING \ =,_,=7-""_/METERING ( ...... _ BARRIER NOZZLE -- BARRIER NOZZLES _ CI_IANI_EL/ (TYPICAL) (TYPICAL) 7.GEN.22807 FIGURE 10-6. SPANWISE AIR COLLECTION SYSTEM DYNAPORE OUTER SURFACE LAYER PERFORATED FIBERGLASS LAYER STREAMWISE FLOW DIRECTION GLOVE PANEL SUCTION FLOW \ FRONT SPAR COLLECTOR CHANNEL METERING NOZZLE MAIN BOX SKIN INTEGRAL STIFFENER TYPICAL UPPER SURFACE CONSTRUCTION AND SUCTION SYSTEM FIGURE 10-7.

The depth of the wing box stiffeners controls the area available for spanwise ducting of the suction air. The depth of the stiffeners as determined by structural requirements, was assumed initially, and the pressure drops along the integral spanwise ducts were calculated as a function of the spanwise station of suction air collection. The pressure drop was calculated using conventional pipe flow relationships with an assumed minimum friction coefficient of .002. The duct flows are relatively low in the main box region and the Mach number varies from about 0.05 to 0.15 in the spanwise integral ducts.

In the spanwise collection system, it is necessary to provide holes in the upper wing main box skin to conduct the suction air from the spanwise ducts to the manifolds leading to the compressor inlets. The size of these holes is related to the number and location of the suction engines and is discussed in Section 10-6. Figure 10-8 shows a typical detail of the arrangement provided. To keep the structural penalty associated with the cutouts to a minimum, the flow was accelerated to a maximum Mach number of 0.3 in the ducts leading to the manifold. A minimum bend radius of twice the duct inside diameter was maintained throughout the system.

The channels formed by the wing box structural members provide adequate flow area for the region from the front spar back to about 62 percent chord, but beyond this station the suction requirements increase very rapidly. Structural members with greater flow area were devised to provide more flow area in these regions.

For the high flows in the vicinity of the rear spar, the structural/ducting arrangement shown in Figure 10-9 was proposed for the concept with suction on both surfaces. An auxiliary spar provides one boundary, the rear spar another, and the upper and lower wing skins are the remaining boundaries of a region dedicated to the collection of the suction air and its ducting to the vicinity of the suction compressor.

OR_C--f,",h._k p,,,,,,_ ,...

' ,_ ! e - OF POOR Qu._Lj]y EWING SKIN :y__STRINGER _____ POROUS SURFACE AUX DUCT SKIN ELLIPTICAL WING SKIN ORIF CUTOUT TO COMPRESSOR INLET SGEN-_ MANI FOLD DUCT FIGURE 10-8. TYPICAL WING-SKIN CUTOUT FOR SUCTION DUCT f-- /_ UPPER SURFACE DUCT i R SPAR PLANE "-_ LOWER SURFACE DUCT AUX SPAR PLANE 8-GEN-22961 FIGURE 10-9. DUCTING AFT OF REAR SPAR For the upper-surface-only configuration with suction extending to 85 percent chord, the concept shown in Figure I0-I0 provides large flow areas aft of the rear spar. The rear spar, at 67 I/2 percent chord, is farther forward than the 70 percent chord location with suction on both surfaces, but there is still more than adequate fuel volume in the LFC wing. To obtain the desired flow control, four ducts may be desirable rather than the three shown.

As discussed in Section II.0, a system has been proposed to avoid contamination of the porous surface by applying a liquid to it during takeoff and climb.

To remove any residual liquid before LFC operation is initiated, a require- ment was defined to provide positive pressure air under the surface for a short period. Because of the relatively high pressure differential required for this operating mode, the airflows through the surface are much higher than those obtained during the LFC mode. Depending on the suction design flow and Mach number, some elements of the suction system can be sized by this purging require- ment rather than by suction requirements. This is discussed further in Section 10.7.

I" i ¸

10.5 SUCTION PUMP SYSTEM

Air is pumpedfrom the boundary layer by means of compressors. Characteristics

and configurations for this part of the suction system were investigated, and

the type of compressor and its driving unit were selected for the advanced

LFC aircraft. Two units per airplane, one on each wing, were determined to

be compatible with the structural and airflow requirements of the concept.

10.5.1 Compressor Design

The general operating requirements for the compressor are: (I) normal operation at very low inlet pressures; (2) inlet distortion tolerance; (3) low noise levels in the critical I000 to 2000 hertz range; (4) pressure ratios and airflow capacity to be determined in the study.

A maximum operating altitude of 12,190 m (40,000 ft) was assumed. Typical pressure relationships for the suction system are shown in Figure I0-II. The lowest pressure on the wing upper surface at 12,190 meters (40,00Oft) altitude was calculated to be 12.5 kPa (261 Ib/ft2), while the lowest pressure on the wing lower surface is 17.96 kPa (375 Ib/ft2). For the initial baseline aircraft, it was assumed that a suction distribution system could be designed with a 20 percent drop in total pressure from the lowest pressure on the wing surface.

This includes a AP of 3 percent through the wing surface and a AP of 5 percent from the end of the collection duct to the compressor face. The resulting pressures at the compressor face are I0 kPa (209 Ib/ft 2) for the upper surface air and 14.36 kPa (300 Ib/ft 2) for the lower surface air, compared with a free- stream static pressure Po of 18.8 kPa (393 Ib/ft 2) at 12,190 meters (40,000 ft) altitude.

10.5.1.1 Pressure Ratio. For suction on both wing surfaces, to achieve a compact installation, the upper surface air is compressed to the pressure level of the lower surface air in a "boost" compressor. The two airflows are then combined and compressed in the main compressor before being exhausted overboard.

The pressure relationships discussed above indicate a pressure ratio of 1.44 for ',<_ i ORIGI:",IAL P_.._; _S OF POOR QUALITY

I

STRUT/_ /,,,: 8-GE N 22°J6G FIGURE 10-10. DUCTING AFT OF REAR SPAR (SUCTION ON UPPER SURFACE ONLY) Pa (1,000 M° = 0.8 12,190 m (40,000 FT) (LB/FT 2) Pt (REF) (600) o -- TOTAL PRESSURE TO EXHAUST AIR AT V = V ex o (500) Po (400) _" LOWER SURFACE ABSOLUTE ;YSTEM PRESSURE (MINIMUM) DROP PRESSURE t LOWER SURFACE (300) COMPRESSOR UPPER SURFACE INLET A (MINIMUM) SYSTEM &P ___._.____ UPPER SURFACE (200) COMPRESSOR INLET (loo) 83-GEN 23219 FIGURE 10-11. TYPICAL SUCTION SYSTEM PRESSURES the boost compressor. The determination of how much the suction air should be compressed in the main compressor before exhausting it in the freestream involves the size of the equipment and the amount of fuel required. At one extreme, the air can be exhausted with a very low velocity, with the attendant penalty of the "ram drag" of the boundary layer air. This results in additional fuel consumption by the main propulsion engines to overcome the drag. The other extreme is to exhaust the air at a high velocity, thereby generating some thrust allowing the main propulsion engines to be throttled back. (The main propulsion engines are sized with the LFC inoperative, so that any thrust available to the aircraft from the suction system does not reduce the size of the main propulsion engines.)

Figure 10-12 shows how the total fuel burned and the size of the drive unit for the compressors vary with the pressure ratio of the main compressor, for a typical cruise condition. The relative exhaust area required for the suction air is also shown in Figure 10-12. At pressure ratios below about 1.7 the necessary area becomes very large. The figure illustrates that fuel consumption is f \ insensitive to pressure ratio over a compressor range of about 1.7 to 2.2. Above a pressure ratio of 2.2, the fuel consumption of the compressor-drive power system increases faster than the benefit of the thrust generated by the higher exhaust velocity. Below 1.7, the increased fuel flow of the main propulsion engines to overcome the ram drag becomes a factor. At a pressure ratio of 1.88, the suction air exhaust velocity equals the freestream velocity. Using this pressure ratio as a design point leaves a margin on each side for off-design operation within the range of minimum total fuel consumption.

10.5.1.2 Compressor Sizin 9. The airflow capability that determines the diameter Wa _t of the compressor is usually expressed as at , where e t is the total temperature divided by standard sea level temperatures, and a t is the total pressure at the compressor face divided by standard sea level pressure, or the total pressure in atmospheres. The temperature of the air into the compressor is assumed to be the same as the boundary layer recovery temperature (recovery factor : 0.85).

The airflow is calculated from the LFC requirement by integrating the suction requirements over the porous surface: Wa : p_gV Is CQdAs or Wa = P_g Is Vwdas w _ is independent of altitude.

For a given Mach number, CQ, and Cpmin, a

_T

Most of the study was conducted using values of CQand Cpmin corresponding

to a design point condition of Mach 0.8 with the lift coefficient (CL) required

for the initial LFC cruise. However, as discussed in Section 7.1.6, both CQ and Cpmin change at off-design conditions so that before specifying a compressor for the suction system, the desired range of operation must be considered.

For an aircraft in airline service, a fixed-area exit for the compressor is desirable. This imposes some limitations oncompressor capability. Figure 10-13 shows a typical compressor map. At I00 percent of rated inlet flow, the exit area is sized for the design point condition at a compression ratio of 2.66. The compressor inlet pressure is 0.53 of the free-stream static pressure to which it is assumed the compressor exhausts. The operating line "A" shows how the airflow and pressure ratio vary at this condition if the compressor RPM is changed.

As discussed in Section 5.3.7 and illustrated by Figure 5-16, the normal operating procedure would be to hold altitude and Mach number during cruise. The wing lift coefficient would then fall until sufficient fuel was burned to allow a step climb to the new cruise altitude, reverting to the original lift coefficient.

Figure 10-13 shows the effect on compressor operating conditions of reducing the lift coefficient from a design CL of 0.56 to a CL of 0.44. The estimated suction requirements at this point are compared with those of the design point in Table I0-I. Operating line "B" on the compressor map, Figure 10-13, shows typical compressor operation associated with the reduced CL.

Other off-design conditions, such as off-design Mach number, must be considered when specifying compressor characteristics, particularly if the compressor has been significantly oversized. Care must be taken to avoid both the stall (surge) and the choking regions of the compressor.

O_,,Gii"_AL F/-_C.C. __ OF POOR QUALITY r_.

/

1.olo m

RELATIVE TOTAL 1.005 FUEL FLOW _TOTAL FUEL FLOW i Vo 4.0 3.0 RELATIVE SIZE OF COMPRESSOR DRIVE AND SUCTION AIR EXIT AREA 2,0- REA TO EXHAUSTSUCTION AIR _,._"_E (RATING) OF 1.0- / _.__._PRESSOR DRIVE V°x - V° I I I I .0 1.5 2,0 2.5 3o0 MAIN COMPRESSOR PRESSURE RATIO FIGURE 10-12. EFFECT OF MAIN COMPRESSOR RATIO i t 3,8 ¸ OPERATING LINE ASSUMES: 1. CONSTANT EXHAUST NOZZLE AREA INLET TOTAL PRESS. = 0.53("A") 2.

EXIT STATIC PRESS. = 0.62 ("B") 3. 'qpoly = 0.88 COMPRESSION 3.0 RATIO CL (DES. PT.)

C L -- 0.44 50 60 70 80 90 100 110 INLET CORRECTED AIR FLOW RATE (PERCENT) 83"GEN23222 FIGURE 10-13, TYPICAL COMPRESSOR MAP ORIG|NAL FAG'_ |S OF pOOR QUALI"P/ TABLE I0-I OFF-DESIGN REQUIREMENTS Constant Altitude & Mach No.

Relative Wa_ Relative Press @ 6 Suction CPN _Psys Compr. Face Compr. Compress, CL Air Flow min approx. Static exit Press. Face P.R.

0.56 1.00 -I 20% 0.53 1.0 2.66 0.44 0.75 -0.8 15% 0.62 0.64 1.95 10.5.1.3 Compressor Type. Both axial and centrifugal compressors were considered for the advanced LFC aircraft suction system. Their characteristics are compared qualitatively in Table 10-2. The centrifugal type has many character- istics that would make it attractive for application in an LFC suction system.

However, the pressure-ratios of 1.44 and 1.88 discussed previously are considerably lower than the typical single-stage pressure ratio of a centrifugal compressor.

For LFC on both wing surfaces, an attempt was made to take advantage of the centrifugal compressor's high pressure ratio to reduce the size of the required equipment. It was proposed to "overboost" the upper surface air and combine the exhaust in an ejector with the air from the wing lower surface, as shown schematically in Figure 10-14. (Although a single-entry impeller is shown, a double-entry design might be used.)

TABLE 10-2. COMPRESSOR COMPARISON Centrifugal Relative to Axial Efficiency (Polytropic) A = 0 to -3 percent Max Dimension Larger Diameter* A depends on pressure Weight ratio - no relative indication Lower Noise Distortion Tolerance Better Pressure ratio/stage Approx 4.0 vs 1.5 Better Installation Flexibility Cost Low Maintai nabi lity More Rugged Off-Design Operation Wider Range *The increase in diameter over the axial depends on the extent of off-design capability.

Figure 10-15 shows the effect on the main compressor size and fuel flow, assuming no mixing losses in the ejector. The reduction in equipment size was not significant, even with the optimistic assumptions. The installation of the large diameter centrifugal pump and an ejector with a reasonable mixing length proved difficult. Another disadvantage is the "screech" associated with ejectors, which would introduce a potentially strong noise source.

The study indicated a pressure ratio of 1.44 for the boost compressor, to raise the pressure of the upper surface air to that of the lower surface air.

This ratio is suitable for an axial compressor, and the geometry of an axial flow compressor lends itself to a concentric installation with the exhaust from the compressor combining with the lower surface air at the inlet to the main compressor. The most efficient pressure ratio for the main compressor has been shown to be in the range of 1.9, which is also lower than would be provided by a single-stage centrifugal compressor. Axial compressors are therefore proposed for the LFC advanced transport. The design should incorporate features

i

to provide distortion-tolerant operation. Blade chord lengths should be relatively long to increase inlet Reynolds number, and the blade-vane spacing should be selected to minimize noise generation.

The above study was completed during the initial period of the contract when suction on both upper and lower wing s_urfaces was proposed. The decision to use suction on the upper surface only removed one advantage of the axial flow compressor: a single compressor with a compression ratio of about 2.7 replaced the two low-pressure-ratio compressors. The compression ratio of 2.7 is in the range of that of a single stage centrifugal compressor, although the dimensions of the centrifugal type are still a disadvantage. As the compressor type is not a major item at this point, the study was not repeated for the upper-surface only case, and an axial compressor is used for the suction system of the advanced LFC aircraft with suction on the upper surface only.

ORIGINAL OF POOR PAGE f_ QUALITy S FRONT SPAR WING LEADING PLANE "BOOST" COMPRESSOR EXTERNAL AIR ----.

INLET TO GAS CENTRIFUGAL TURBINE I COMPR ESSOR PRIMARY NOZZLE FOR _ EJECTOR TO UPPER SURFACE INLETS MAI TO LOWER TURBINE SURFACE INLETS TURBINE EJECTOR EXHAUST (OVERBOARD) F LOW "'MAIN" COMPRESSOR REAR SPAR PLANE )ARD EXHAUST FROM COMPRESSOR WING TRAILING EDGE FIGURE 10-14. SCHEMATIC SHOWING USE OF "BOOSTED" UPPER SURFACE AIR TO DR IVE EJECTOR ( SCHEMATIC: MAIN BOOST SURFACE AIR UPPER COMPR'L"" COM p R.L______ Ve x LOWER SURF, AIR 3.0 MAIN COMPRESSOR PRESSURE RATIO MAIN COMPRESSOR PRESSURE RATIO 2.0_ 1.0 Vo VIX 1.2 RELATIVE POWER 1.1 REQUIRED RELATIVE SIZE OF MAIN COMPRESSOR 1.0 0.9 0.8 I I I I 1.0 1.5 2.0 2.5 3.0 BOOST COMPRESSOR PRESSURE RATIO FIGURE 10-15. EFFECT OF BOOST COMPRESSOR PRESSURE RATIO 10.5.1.4 Compressor Efficiency. For the power-required calculations, the polytropic efficiency was assumed to be 88 percent. Present day compressors of large engines have efficiencies of this order and higher. By the 1995 time period, the level of 88 percent should be achievable even with small compressors operating at low inlet pressures. A change of 1 percent in polytropic efficiency would change the power requirement by approximately 1.2 percent.

10.5.2 Power Unit For Compressor A study was made of power sources to drive the suction-system compressors.

A concept with separate engines installed in wing pods was assumed for the initial baseli-ne aircraft.

The order of magnitude of the power required is significant when different types of power sources are evaluated, as some types are not suitable for certain power ranges. The estimated suction power requirements for a typical LFC aircraft are shown in Table 10-3. (Although only the concept with suction on both the upper and lower surfaces was under consideration at the time the power-d_ive study was performed, a typical upper-surface-only power requirement is included in the Table.) For a 381 m2 (4100 ft 2) wing area, a total power per airplane of 752 kW (I010 hp) is required at 10,668 m (35,000 ft) altitude.

Two turboshaft engines rated at approximately II00 kW (1500 hp) each at sea level will provide this level of power. For the baseline airplane performance, the specific fuel consumption of an engine of this size in the 1995 era is assumed to be 0.07 gs-I/kw (0.4 Ib/hr/hp) under cruise conditions, based on performance of current engines and estimates of performance improvements.

An alternative means of providing power of the required magnitude is by power extraction from the main propulsion engines, either mechanical power or bleed air. With the main propulsion engines in the tail, mechanical power extraction is not practical. Ducting the low pressure suction air from the wing tip to the fuselage, and then aft to a compressor in the tail region is also not practical. It would result in excessive pressure drop, very large space requirements, and severe structural problems.

OF POOR QUALITY

Table I0-3 Typical Suction Power Required 10,608 m (35,000 ft) M o : 0.8 Suction on Upper & Lower Surface to 70% Chord Sw : 381 m2 (4100 ft2) Power Power Suction Press Compr Airflow Face Airflow per Airplane CPR* per Airplane 165 kW 30.7 w/gs -I 1.27xi04 I_: 1.44 5.35 kg/sec Upper Surface (200 hp) (265 Ib/ft2) (18.7 hp/Ib/sec) (II.8 Ib/sec) 1.83xi04 Po.

Lower Surface 4.54 kg/sec (382 Ib/ft2) (I0 Ib/sec) 59.3 W/gs-I 588 kW 1.83xi04 l.88 Upper & Lower 9.9 kg/sec (36.1 hp/Ib/sec) (790 hp) (382 Ib/ft2) (21.8 Ib/sec) 763 kW TOTAL (lOlO hp) Suction on Upper Surface to 85% Chord Sw = 288 m2 (3100 ft2) Suction Power Power Airflow Press Compr CPR* Airflow per Airplane Face per Airplane 660 kW 1.27xi04 P_ 91.2 W/gs"I 2.66 7.24 kg/sec Upper Surface (880 hp) (55.5 hp/Ib/sec) (265 Ib/ftZ) (16 lb/sec) *For Vexhaus t : Voe

A comparison was madebetween using separate turboshaft engines and turbines

driven by engine bleed air. A bleed-air-only system and a system where fuel

is added to bleed air in a burner ahead of the turbine were also considered.

Except for small amounts of bleed for de-icing purposes, engine bleed air used

for commercial aircraft applications is normally restricted to a temperature of

about 500°K (900°R). At higher temperatures than this, fire-wall type protection

must be provided around the ducting, along with a sensing system for leak

detection. Table 10-4 summarizes the comparison of the performance of a bleed-

only system with a bleed temperature of 500°K (900°R) and one with a bleed air

temperature of 683°K (1230°R). The high-temperature bleed system has only a

slightly lower overall fuel consumption than the lower-temperature system.

The bleed-only systems had about 30 percent higher SFCthan the turboshaft and

the size of the bleed ducts was incompatible with the very limited space available

behind the rear spar. For these reasons, the bleed-only system was eliminated

from further consideration.

Figure 10-16 shows power available with air only and with combustion, as a function

of turbine pressure ratio (PR) and inlet temperature. The fuel-air ratios

required to achieve the turbine inlet temperature of a bleed-burn system with

500°K (900°R) initial air temperature are also plotted in Figure 10-16. With

a bleed and burn system, high specific powers can be obtained with fairly low

turbine pressure ratios (low bleed-air pressures) at higher turbine temperatures.

The bleed-and-burn system was compared to a turboshaft engine at a design-

point condition. The same turbine inlet temperature of 1444°K (2600°R) and the

sameturbine efficiency were used for each system. The assumptions used are

listed in Table 10-5. Table 10-6 lists the station cycle parameters calculated

for each system. The effect on aircraft fuel consumption is given in Table 10-7.

(The SFC penalty for bleed varies with engine cycle and operationg condition,

but the value used is typical of an energy-efficient engine operating at about

90 percent of max cruise. The SFC penalty of a bleed system increases with

decreased throttle setting, with a constant value of bleed extraction.) The

analysis shows a lower overall fuel consumption for the separate turboshaft

engine.

ORIGINAL p_'_, L OF POOR QUALITY Table 10-4 Effect of Bleed Temperature for Bleed-only System 500°K (900°R) 688°K (!230°R) Bleed Temp 855 (124) 241 (35) P_.- (psia) Bleed Pressure @ engine, 731 (I06) 207 (30) P_ (psia) Bleed Pressure @ Turbine, Turbine Pressure Ratio 394.6 (240) 205.5 (125) Power/airflow, kW/(kg/s) (hp/Ib/sec) 2% 1.07% ASFC for one percent bleed For aircraft with total fuel flow : 1264 g/s (10,030 Ib/hr/airplane) and total air flow = 72.5 kg/s (160 Ib/sec)[S w = 381 in 2 (4100 ft2)]: Power Required = 753 kW (I010 hp) 1.90 (4.2) 3.66 (8.08) Airflow to Turbine, kg/s/airplane (Ibs/sec/airplane) 2.6% 5% Bleed Flow 65.5 (520) 67.6 (536.6) AWl due to bleed, g/s (Ibs/hr) .0876 (.517) .090 (.531) BSFC, gs-lkw (Ib/hr/shp) 4.8 (1.9) duct 12.7 (5) Dia pipe (per side), cm (in) 9.9 (3.9) duct shroud 250 (550) 215 (475) Est. Weight of Ducting, kg (Ibs) Table I0-5 Cycle Comparison- Assumptions Turboshaft compressor ratio = 14.8 Compressor adiabatic efficiency = 0.82 Turbine polytropic efficiency : 0.90 Burner efficiency - From DAC Cycle Program Fuel heating value = 4.28 x 107 J/kg (18,400 BTU/Ib) 0.5% burner AP Nozzle Cv = 0.98 Turboshaft exhaust velocity = Vo (no net thrust) Bleed-and-burn turbine pressure ratio = 8.0 B1 eed Ai r 500°K (900°R) 241 kPa (35 psia) @ exit of bleed port @ 10.7 km (35,000 ft) altitude (typical value corresponding to 500°K for adv. technology turbofan) 207 kPa (30 psia) @ burner Table 10-6 Turboshaft & Bleed-and-Burn Cycles I0,608 m (35,000 ft) Mach 0.8 Turboshaft Bleed-and-Burn 36.1 kP:L (5.23 psia) Compressor Pt 2 in m 247°K (444.5=K) Tt 2 534 ¢_ (77.4 psia) 207 k P_ (30 psia) Burner Pt 3 in 590°K (I063°R) 500°K (900°R) Tt 3 944°C (1700°F) 854°C (1537°F) aT .02455 .0270 qb " f/a Burner 0.95 0.939 qb .0258 .0288 f/a 531 kP_ (77.0 psia) Compr Pt4 1444°K (2600°R) Turbine in Tt4 195 _(m_" (28.3 psia) 206 kP_ (29.9 psia) Power Pt5 ll70°K (2106°R) 1444°K (2600°R) Turbine in Tt5 27.4 k_ (3.93 psia) 25.7 _P_. (3.73 psia) Power Pt 6 Turbine Out 933°K (1679°R) 761°K (1370°R) Tt 6 AT 409°C (736°F) 511°C (920°F) Power/airflow 495 W/g's -I (301 hp/Ib/sec) 644 W/g.s -I (392 hp/Ib/sec) Power Turbine BSFC .052 g-s-I/kw (.309 Ib/hr/hp) .045 g.s-i/kW (264 Ib/hr/hp) 1.147 1.075 Pressure Ratio Nozzle 237 m/s (779 ft/sec) 192 m/s (629 ft/sec) Vexhaust .... J Table 10-7 Comparison of Fuel Required by Turboshaft and by Bleed-and-Burn System Assumed for Main Engines: No-bleed SFC : 16.7 g.s'I/kN (0.59 Ibs/hr/Ib) I% increase in bleed - 1.07% increase in wf Engine fuel/air = 0.0174 Fuel Consumption for Turboshaft: (from Table _ -_) wf = O.052.(kW req), g/s [wf = O.309-(hp req), Ibs/hr] Fuel for Bleed-and-Burn: wf = fuel for turbine + fuel for bleed - savings for exhaust thrust bleed air flow - kW___, g/s [= _-_, Ibs/hr] savings = SFC x exhaust velocity x bleed air mass flow wf = 0.045 (kW req) + 0.029 (kW req) - .005 (kW req) = 0.069 (kW req), g/s [wf : 0.264 (hp req) + 0.171 (hp req) - .029 (hp req) : 0.406 (hp req), Ib/hr Comparison: Bleed-and-Burn Fuel Turboshaft Fuel : 1.3 _F P©OR QUALITY A qualitative comparison of the two systems is summarized in Table 10-8. The turboshaft engine has the advantage of being independent and self-contained, with the attendant merits of ease of installation, maintenance, and ground check. The bleed-and-burn system requires a slightly smaller turbine, and eliminates the requirement for a compressor for the power drive. The major disadvantage of the bleed-and-burn system is the requirement for a pressurized, relatively high temperature duct running from the main engines in the tail area, through the fuselage, and down part of the wing span. This must be done without compromising the safety of the aircraft.

When the time comes to build an LFC aircraft of a given size, and to choose _ a power source for its suction-air compressors, the principal criterion will be either the direct operating cost or the return-on-investment. Another factor is hardware availability. At this future time, detailed design and cost studies will be made of all available engines including turbos ha fts of the required size and turbines that might be incorporated in a bleed-and-burn system. For this conceptual study, however, the turboshaft engine shows three definite advantages: o Lower fuel consumption o Better operational flexibility o No hot air ducting through wing and fuselage.

For these reasons, the decision was made to use a turboshaft engine to drive the compressor in this study.

OF POOR QUALll'Y Table I0-8 t ComParison of Suction _ualitative mower Systems Item Turboshaft Bleed-&-Burn System Ground Check Self-contained r:equi res high-temp., press, gas supply r- ° Effect on ilain :lone Penalties increase as _ngl ne main engine throttled.

If LFC bleed + a/c bleed exceed normal engine bleed quantity limits, special engine devel.

required Inlet, compressor, 9ucting through fuselage ;4eight starter and part of wing Slightly smaller pod for Drag Slightly larger dimensions for t,_rbo- turbo-machinery machinery :lone Bleed air ducts from ;4ing Volume fuselage to burner Could be "off-the-shelf", Special equipment with Availability depending on size dev, iopment & certification aircraft program.

Conventional modded Instal lation Leak-detection provision turboshaft wilere duct near critical structure or equipment.

Blow-out provisions if duct runs through low- volume, sealed compartment.

I 10.6 SUCTION ENGINE ARRANGEMENT In the wing box, the suction flow collection system is integrated with the structure, as shown in Figure 10-7. The boundary layer air is drawn into the main air ducts formed by the suction panels, integral stringers, and wing skin. The air flows spanwise to the suction engine location, where it is ducted from the individual channels into manifolds merging into inlets for the suction compressors. The number and location of the suction engines must be such that the flow in the ducts does not undergo excessive pressure loss, and the size of the cutouts required in the wing structure to remove the air is compatible with the structure dimensions and allowable stress levels.

10.6.1 Number of Suction Engines With the small compressors and engines that are required by the LFC aircraft suction system, there are advantages in having fewer, larger engines rather than more smaller engines to provide a given total capacity. Reynolds number effects and clearance and minimum gage considerations combine to make the larger units more efficient and proportionally smaller and lighter. For the LFC aircraft, there is also a weight penalty for each collection station where the air must be ducted from the spanwise flow channels, through a dry bay, and manifolded into inlets for the suction "pumps" (compressors). A drag increment is incurred, likewise, for each engine pod. When direct-operating cost (DOC) is considered, keeping the number of engines to a minimum is important as both initial and maintenance costs are sensitive to the number of engines. In particular, maintenance labor costs are almost directly proportional to number of engines, with size having very little effect. Because of this, the number of engines is an important factor in airline acceptance.

The number of suction engines affects the suction-engine-out situation but because safety was not affected, a decision was made early in the LFC study not to provide redundancy or cross-ducting because of the large amount of additional weight, volume, and complexity involved. With one suction engine/ side, laminar' flow would be lost on that side only in case of a suction engine shutdown. Laminar flow would be maintained on the other wing, and normal flight, with an increase in drag over half the wing surface, would continue following throttle and trim adjustment. With two suction engines/side, laminar flow could be maintained on approximately three-quarters of the wing surface if one outboard engine were shut down. Failure of an inboard engine would result not only in loss of laminar flow on the quarter of the wing surface sucked by that engine but also on a portion of the adjoining outboard wing surface. As the shutdown rate per engine is independent of engine size, the increased fuel consumed because of loss of laminar flow from suction- engine failure is roughly independent of the number of engines.

The concept of one suction engine per airplane was rejected for two reasons: (I) for safety and for low noise levels in the cabin, jet engines in the passenger region of commercial transports should be avoided whenever possible; and (2) preliminary calculations showed large penalties from increased duct sizes and pressure drops, resulting from ducting air the entire semi-span.

10.6.2 Location of Suction Engines With one suction engine/side, the location of the engine must also be considered.

Placing each engine in the wing root has the same disadvantages as the centrally located single engine mentioned above. A midwing location may require a dry bay to provide space for removing suction air from the wing box ducts, combining all the suction flows, and ducting the air to the compressor. For the design mission of the LFC advanced aircraft, ample fuel volume remains despite the inclusion of dry bays. A logical position for the location of the dry bay, compressor, and suction engine is at the "wing break" point, which is at 40 percent of the semi-span for the baseline configuration. The structural discontinuity in the flap system is avoided. Flow distributions and pressure drops were calculated for the wing box suction air as a function of location along the span. Figure lO-17(a) shows typical cumulative flows from each direction. Over the greater part of the surface between the front and rear spars, the required suction flow per unit area of wing surface is quite low, of the order of 3 x 10 -2 m/s (0.I ft/sec), and is independent of the wing chord.

With the wing blade height selected by structural considerations, the calculated pressure drop in this region varies with span as shown in Figure lO-17(b).

Pressure drop was calculated by conventional "pipeflow" relationships. The dashed line in this figure represents the pressure if the pressure drop were assumed to be twice the conventional driction drop. Even with this assumption, the pressure at the 40 percent semi-span meets the goal assumed in the definition of compressor and power requirements.

The area requirements for removing air from the wing box ducts were calculated as a function of Mach number and spanwise location, with the results shown in Figure 10-18. The location of the engine just outboard of the.40 percent semi- span station results in approximately equal cutout area requirements for the air from the inboard and from t#e outboard wing surfaces.

Layouts were made of the ducting and engine installation with the engines located below the wing just outboard of the wing break. The inboard bulkhead of the "dry bay" is located at 40 percent semi-span, the wing break point. For the wing geometry used in the layout (AR = 12), the engine centerline is at 41 I/2 percent semi-span. Figure 10-19 is a profile view of the installation study showing the suction pump, engine and duct manifold arrangements for LFC suction on both wing surfaces and for suction on the upper wing surface only. In addition to the multiple holes in the wing box upper surface, a large hole is required in the lower surface for ducting the suction air to the compressor. The compressor rotors are aft of the rear spar for safety reasons and to facilitate ducting air to the inlet with a minimum of severe turns. The drive unit is axially coupled aft of the compressors, and the installation enclosed in a faired pod under the wing. The aft mounting of the turboshaft is advantageous both for safety and for minimizing the noise transmitted to the laminar wing boundary Iayer.

OF P=_ J o: L_ ,,L;,"V W/gs-] HP LB/SEC 400-

,oo

/ "BLEED & AIR / BURN" 300- / TURaINE/ / TURBINE / PowEy / , POWER 400 - __t /" f ACT_L/--_,- . -- 0.03 200- 300 - __/,_'.././" - 0.02 FUEL-AIR 200 - 100%COMBUSTOR EFF. -- 0,01 100 ,. I I 0 400 600 800 1,000 1,200 1,400 OK t I I I 1,000 1,500 2,000 2,500 OR TURBINE INLET TEMPERATURE 8_E_,_J22_ FIGURE 10-16. TYPICAL "BLEED & BURN" SYSTEM RESULTS (500°K BLEED AIR) CQ Vw -'G"%o=13 x 10 - 4 LOCAL AIRFLOW TOTAL AIRFLOW Pa I 13,000 - I LB/FT 2 SURFACE 260- PRESSURE 12,000 - 240- PRESSURE 11,000 220 - • / _EST, FOR f = 2fo=:.eflo w 10,OO0 ',/ (b) DUCT PRESSURES 20O I 12,190 m (40,000 FT) 9,O0O I I = = i I 20 40 60 80 10O DISTANCE FROM FUSELAGE C L (% SEMI-SPAN) a._E.2m_ FIGURE 10.17.

SUCTION AIRFLOW IN WING-BOX DUCTS r AR =12 Sw =334.5 m2 (3,600 FT 2) V w =0.03 m/s (0.1 FT/SEC) IN. 2 cm 2 ASSUMED: - I. 12 EQUAL CHANNELS BETWEEN MDUC T FRONT AND REAR SPAR i 2. PRESSURE DROP BASED ON 0.25 / 60% WIDTH AVAILABLE '%,,,_ORAIR FLOW / 3O

\ /

INBOARD FLOW CUTOUT AREA

"_',. _ ,>,,.. 0.35

OUTBOARD CUTOUT i" " 0.35 I I I I 35 40 45 50 DIVISION POINT (% SPAN) 83_E,2322, FIGURE 10-18.

TYPICAL AREA FOR WING-BOX-DUCT CUTOUTS (UPPER SURFACE)

C

_ .L - ._ ,-. _' _C _CC\ 1 _._<' _',_;,_ .,---_ __

__ .-_ _ _s__.:_. ]I -_

LAMINARIZATION TO 70 PERCENT CHORD UPPER AND LOWER SURFACE " '....... II TM rr"!_ LAMINARIZATION TO 85 PERCENT CHORD UPPER SURFACE ONLY ._.,_,,_ FIGURE 10-19. SUCTION SYSTEM MANIFOLDING INTEGRATION

C

The alternative LFC airplane configuration with suction on the upper surface only but with the suction continued further aft shows that ducting air from the surface back to 85 percent chord requires the suction pump and engine to be located further aft, but the overall arrangement is greatly simplified.

Refer also to Figures 5-48 and 5-49.

10.6.3 Number and Location - Conclusions The study layouts demonstrate that a configuration having one engine/side, with the engine located just outboard of the "wing break", is a feasible concept for the LFC basline aircraft. This configuration results in acceptable pressure losses in the ducting system, and is compatible with the proposed wing structure. Moving the engine outboard results in more pressure drop and larger cutout areas for the flow from the inboard region, as well as reducing the vertical dimensions in the wing for installing the ducting to the engine.

The pressure drop and area increases would be even greater if the engine location were moved to the wing root region. For these reasons the decision was made to use two suction engines per aircraft located each side at the wing break.

10.7 POSITIVE-PRESSURE PURGING SYSTEM During the development of the surface-contamination-control system (Section II), a procedure was defined for applicazion of a liquid to the surface during takeoff and climb for contamination avoidance and ice protection. Some of the liquid remains on, or in, the surface, and would impede the flow of LFC air if allowed to remain when suction is initiated. Tests demonstrated that a pressure differential of 6900 Pa (one psi) applied across the porous surface would remove the liquid and return the surface to approximately its original porosity in one minute. Tests on perforated materials indicate that a lower pressure increment may be sufficient for clearing these surfaces. A lower pressure requirement reduces the magnitude of the situation discussed below, which is based on the original 6900 Pa (one psi). See Section 11.6.2.4.

Ramair is a convenient source of positive pressure, having sufficient pressure

to supply the required differential for purging all LFC porous surfaces except in

the attachment line area. Pressurized cabin air can be used as a source for

this part of the system.

The greater pressure drop across the surface during the "purging" moderesults

in an order-of-magnitude greater airflow. Unless the Mach number in a duct or

other component is very low in the suction mode, the positive-pressure flow

will "choke" before the surface has been cleared. A comparison of typical

suction and "purging" conditions is madein Table 10-9. In the suction case,

the values of pressure coefficient, CPN, and suction flow quantity, CQ, are

representative of those in the wing box region behind the front spar. The second

column shows the flows encountered with the application of 6900 Pa (one psi)

pressure differential across the surface at 7,620 m (25,000 ft) altitude. In

this condition, the critical flow quantity, w_-6_, which determines area require-

ments, is almost eight times its value during LFC operation. If some part of

the suction system is designed for even a moderate Mach number, such as the wing-

duct cutout, which is designed for 0.3, the surface area over which the positive

pressure can be applied is appreciably smaller than that from which the suction

air is drawn. A system of valving can be devised to permit sequential purging

of the wing box surface where it is not desirable to increase flow areas to

accommodatethe positive-pressure flow requirements. This is illustrated in

Figure 10-20. The upper schematic shows the LFC modewith continuous suction

applied to the whole LFC surface. The lower schematic illustrates how the

purging flow to the porous panels can be reduced. Groups of valves could be

opened in sequence to limit the duct Mach number. The number of sequences needs

to be kept to a minimum to avoid excessive time being taken for purging.

For wing surface regions, where the suction flow is high, such as aft of the

rear spar, the ratio of purging flow to suction flow is reduced and continuous

purging could be achieved with acceptable duct Machnumbers. Figure 10-21

shows that with a suction CQof 1 x 10-3, the purging flow Mach numbers are

acceptable even with moderate Mach numbers in the suction mode.

CF-!_S_:/'I_ :°L .....

4J (,..} C.J ,."-".

4-J 4- .._ "---. _ _-- CS * ¢_" c_J P_ v v C_J -¢'T _'1 Ud m CO t_ Oh f¢3 L_ 0 0 C It oB,.,- <:_ ._a 4- ¢.-. ¢,,j U (M ,(.j .._ I .I._ ¸ I 0 o _, " _ <::1 o I 0 0 L.¢_ P,,. kr) v v v v v vv_ O0 X C_J

E

°

aJ II ¢_-) ._ 7 or) u')t M _d p-- ¢:_ v 0 G (J p') U _ °r- U 1.

b'l p- l..

J_ 4-- I- t- O S- _J U _J 4-- E e'- _J o _- IlJ $-

il

U £,.

I J_ ORIG,_TJAL F_C:'_ _ OF POOR QU_LETY LFC MODE

_l_ __._

POROUSSURFACE. __.

l_._l_.__._l_l

PANEL FLUTES[ } l I I i{ 11 ][__ COLLECTI)R I CHANNELS LFC SUCTION PANEL NOZZLES E ASSEMBLY

t

INTEGRAL DUCTING

t

DUCT THROUGH WING SKIN DRY BAY = TO COMPRESSOR MANIFOLD _GEN-26372 f_ PURGING MODE

.tttttt.ttttttt. tttttt

POROUS SURFACE ,

I I I

t t t

PANEL FLUTES L Jl J L J [ II II _J COLLECTOR

CHANNELS ]

I LFC SUCTION NOZZLE/VALVES PANEL ASSEMBLY

t

INTEGRAL DUCTING DUCT THROUGH WING SKIN DRY BAY RAM AIR MANIFOLD 1 TM 9-GEN _h_3 7 ] FIGURE 10-20. SCHEMATIC OF SUCTION SYSTEM _r ,,_*e _ OR;GINAL Pk_L_. iS OF POOR QUALITY 1.0 0.8 DESIGN SUCTION FLOW CQ =1.29×10-4 0.6 MACH NO.

DURING PURGE 0.4 CQ = 1 x 10-3 0.2 CONDITIONS: SEE TABLE 10-9 0 I I 0 0.05 0.10 0.15 0.20 0.25 SUCTION DESIGN MACH NO.

FIGURE 10-21. EFFECT OF SUCTION DESIGNMACH NUMBER ON PURGING MACH NUMBER WITH POSITIVE PRESSURE 10.8 SUCTION/PURGING CONTROL The controls for the suction system can be more fully defined after completion of the planned development, test, and flight demonstrator activities. More information is required on such items as the permissible amounts of over- and under-suction, the effect of spanwise differences in suction quantities, actual pressure drops in suction-system components, etc. The general control concept now proposed for the advanced LFC aircraft provides chordwise pressure variation by means of valves between the ends of the main flow ducts and their inlet to the collection manifold. Calibrated nozzles are used at the end of each collector channel to take care of spanwise variations in main-duct pressures.

If the sequential application of positive pressure to purge the surface proves to be necessary, selector valves can be incorporated with the control nozzles.

The selector valves could also be used to prevent flow through the LFC system during flight with LFC off.

The control systems should be designed to increase flight crew work load as little as possible.

10.9 SUMMARY - SUCTION SYSTEM Design studies and analysis indicate that the system proposed for pulling LFC air from the wing surface through a porous or perforated skin into a collection system in which the flow ducts are incorporated with the wing structure is practical. A dry bay, located on the outboard side of the wing "break", contains a manifold for collecting air from the spanwise ducts and conducting it to the inlet of a compressor. A pod beneath the wing encloses the suction compressor and its power source, a turboshaft engine, as well as part of the manifold.

Flow requirements have been estimated to ensure their compatibility with the wing structure. A method has been devised to provide positive pressures beneath the wing surface to remove any residual fluid before LFC operation is initiated.

While some suction system requirements need to be defined quantitatively, or more accurately, no major problems can be discerned that should discourage the proposed orderly development of the concept.

" "" ll.O ENVIRONMENTAL SYSTEMS ll.l INTRODUCTION The ultimate goal of the LFC program is to develop an aircraft design that will achieve laminar flow during cruise flight conditions in an economical and practical manner. One foreseeable problem that could interfere with the attainment of this goal is the contamination of the wing leading edge due to insects, airborne particles, and/or ice. The wing must therefore be protected from any surface accumulations that would create a turbulent boundary layer during normal cruise; laminar flow would not be required during climbout or during an icing encounter.

Early studies have shown that roughness on the wing leading edge due to the residue of impacted insects can be sufficient to create a turbulent boundary layer. Possible approaches are to: (a) Add a protective system to the aircraft that will either prevent insects t from impinging on critical LFC surfaces, prevent adhesion and buildup of roughness after insect impingement, or remove insect contamination.

(b) Reduce the insect aerial population near airports to an acceptable level and do not incorporate a contamination prevention system.

(c)

Accept the reduced aircraft performance resulting from insect contamination and do not incorporate a contamination prevention system.

An economic trade-off analysis could determine which approach would result in the most profitable mode of operation. However, this evaluation assumes that a contamination avoidance system and some form of wing ice protection is required.

A major concern in the evaluation of the various contamination avoidance concepts is whether the ice protection and suction requirements can be met in the space available in the wing leading edge, In view of the limited space envelope and PRECEDING PAGE BLANK NOT FILMED viscosity on pressure drop at low temperature and pressure. The functional dependency must be determined experimentally and is influenced by the condition of the air flowing through the porous media. The major concern with regard to the contamination avoidance system is the effect of temperature on viscosity and surface tension of liquids. These parameters affect the performance of spray nozzles and porous media, the ability of liquid to cover a surface and flow aft, and the pressure required to purge liquid from a porous suction surface.

Current large commercial aircraft use hot air to provide ice protection. A cursory review of the space requirements indicates that this concept may not be feasible. Another change from normal ice protection design is the consider- ation of runback. De-icing and anti-icing systems presently used on commercial aircraft can be permitted to melt some of the ice cap. This water flows aft along the wing until it freezes. The frozen runback remains on the surface because it is aft of the area that is heated by the ice protection system. This condition is intolerable for an LFC aircraft where the roughness could cause transition of the boundary layer.

Some of the considerations included in the feasibility evaluation are listed below: (a) Effectiveness of the concept in protecting the wing against contamination.

(b) Compatibility of the contamination avoidance system with an ice protection system, the aerodynamic performance of the wing, the structural integrity of the wing and materials involved, and the requirements of agencies such as the Environmental Protection Agency (EPA).

(c)

Compatibility of the contamination avoidance system with the LFC suction system with regard to (a) clogging of the suction passages and (b) physical space limitations imposed by the suction system installation.

(d)

Ability of the contamination avoidance system to be manufactured and maintained to meet the maximum allowable roughness requirements of the LFC system.

35O

the possible interaction between systems, consideration must be given to ice

protection when evaluating the contamination avoidance system. Since the

contamination avoidance and ice protection systems must provide coverage of

the same area, and in many cases can use the sametechniques, it would be

advantageous to use a common system for the two functions. Further definition

of the suction system space requirements may make this necessary.

In addition to the design of active contamination avoidance and ice protection

systems, this section of the report reviews the peculiar effects of other environmental contaminants on LFC aircraft design and operation. Dust, sand, exhaust products, and other airborne contaminants can be drawn into, or settle on, the suction surface. The suction material must be relatively insensitive to these factors and maintenance methods must be established to restore the original system characteristics.

Erosion due to rain, hail or sand has caused severe roughening of the wing leading edge for airlines operating in semi-tropical areas. The structural implications for the wing surface are considered elsewhere. However, erosion also is con- sidered in the selection and design of the contamination avoidance and ice protection systems.

Another environmental factor that influences LFC aircraft operation is the frequency and severity of atmospheric ice crystal encounters and their disruption of laminarization. If ice crystals in the cruise altitude regime caused loss of LFC with significant frequency, the cruise altitude would need to be changed.

This section includes a discussion of the problem and a review of presently available data. A program is recommended to provide the basis for a quantita- tive analysis of the effect of ice crystans on the economics of LFC aircraft operation.

The effect of operational pressures and temperatures on the suction and con- tamination avoidance system performance is also evaluated in this section.

For the suction system, consideration is given to the effect of density and C'_ i ¸ (e) General practicality of the design including such factors as system weight, availability of materials, and limitations due to manufacturing techniques.

(f) Compatibility of the contamination avoidance system with the

aircraft environment, especially considering rain, ice, and sand erosion.

(g) Operational considerations such as ground support requirements, periodic maintenance, flight restrictions and operational procedures.

A review of each concept with respect to these factors provides an indication of feasibility and also a measure of confidence in the ultimate successful implementation of the concept.

11.2 CONTAMINATION AVOIDANCE AND ICE PROTECTION SYSTEM REQUIREMENT The major considerations for the contamination avoidance system are the distri- z \ bution and characteristics of airborne insects, the extent to which smoothness must be maintained, compatibility with other LFC systems and structure, and compatibility with normal operation and maintenance procedures.

(a) Insect Population and Characteristics - The contamination avoidance system must prevent contaminants that will trip the boundary layer from adhering to any portion of the wing in which laminar flow is to be maintained. For the LFC aircraft to be economically viable, this objective must be achieved for the large majority of flights. One method of achieving this goal is to use a shield forward of the wing that may also act as a high-lift device.

To analyze the effectiveness of a shield, it is necessary to calculate the trajectory of an insect subjected to the aerodynamic forces of an approaching aircraft. To accomplish this, it is necessary to know the drag-to-weight ratio of the range of insects that are expected to be encountered. Figure II-I, taken from Reference II-I, presents the frequency that insects of various weights were encountered during experiments conducted at very low altitudes. Reference ll-l also states that 54 percent of the roughness generated due to insect impingement occurs during the ground run, about 33 percent during climb to 305 m (lO00 feet) and the remaining 13 percent while con- tinuing the climb to 1524 m (5000 feet). Hence, the values shown in Figure ll-l for insect distribution near the ground are very significant to the study of insect impingement. Figure ll-2 shows the ballistic coefficients of twelve species of small insects with wings removed.

The ballistic coefficient represents the ratio of drag-to-weight of the insects tested. It can be seen that for the range of insect weights that are expected to be encountered (Figure ll-l), the ballistic coef- ficients range from O.l to 0.2. Wings would only increase this value slightly.

Most insects are confined to the so-called terrestrial zone (from ground level to 91.4 m (300 feet) although for insects occurring up to 1500 m (5000 feet), the insect density appears to be a continuous function with C respect to altitude. Figure 11-3 taken from References II-I through 11-6 illustrates both of the foregoing statements. The large variation in insect population found by different investigators is also apparent.

References 11-7 and 11-8 present data relative to the height of insect deposits on a wing leading edge. The data indicate that insect deposit can range in height from 0.381 mm (0.015 inch) to 0.762 mm (0.030 inch) near the stagnation point. The height range decreases rapidly to a value of 0.152 mm (0.005 inch) to 0.302 mm (0.008 inch) within a distance of a few percent of the chord from the peak value, and then decreases more slowly with increasing distance from the leading edge. Figure 11-4 is a crossplot of the data for constant values of E (deposit height).

The region in which protection is required to avoid deposits greater than 0.102 mm (0.004 inch) high, is forward of about 8 percent chord on the upper surface. For the airfoil used in the study, this coincided with a tangent of about 7 degrees to the surface.

20-

I I OF r'-_."_'-_-, .... _,J #.Li;'_" m

FREQUENCY 10-

(PERCENT)

L

r---.1 ,_, ,--.r-I 0 1.0 2.0 3.0 INSECT WEIGHT (mg)

"i ' ' ' i

0 1 15 2.0 2.5 3.0 35 40 AVERAGE BODY LENGTH (ram) 83.GEN 23159 FIGURE 11-1. INSECT FREQUENCY VERSUS WEIGHT -NEAR THE GROUND (FROM "ROUGHNESS DUE TO INSECTS" BY W. S. COLEMAN) FT -1 m-1 NOTE: K = (CD,12_)(p/p i) 0.06 _X p = AIR DENSITY _^ CD IS IN TERMS OF BODY VOLUME (4) 0.2- _ 0 'II_ AND AIR DENSITY X 0 _ CD== zD_/pV2_

BALLISTIC

COEFFICIENT _ _ Pi INSECTDENSlTY

0.04 _, _ LEGEND

(K)

0.1 -- '_z_ _ o CROP FUL, 0.02 A 0- 0.0 I 1 I 0 10 20 30 INSECT WEIGHT (rag) _N-23160 FIGURE 11-2. INSECT BALLISTIC COEFFICIENT (K) VERSUS WEIGHT (FROM "ROUGHNESS DUE TO INSECTS" BY W. S. COLEMAN, USING SMALL INSECTS WITH WINGS REMOVED) ORIGINAL PAGE' IS OF POOR QUALITY 3o0 - 10.0 - 8.0 2oo 4.0 lOO 5O 2.0 2o INSECT DENSITY 1.0 INSECT DENSITY (INSECTS PER 0.8 (INSECTS PER mS) 106 FT3) O.4 LEGEND 0.2 GLICK - 1926-1931 t'l JOHNSON- 1948 FREEMAN - 1938 0.1 HARDY & MILNE - 1932-1935 I I I 500 1,000 0 Z,soO 2,OOO B3GEN 23161 ALTITUDE (METERS) FIGURE 11-3. INSECT POPULATION DENSITY VERSUS ALTITUDE 15 MIN HEIGHT OF INSECT DEPOSIT DISTANCE FROM LEADING EDGE (x/c) (PERCENT CHORD) 0 1 2 3 4 5 6 ANGLE OF ATTACK (DEGREES) .-=,,.2,,., FIGURE 11-4. EFFECT OF ANGLE OF ATTACK ON LIMIT OF INSECT IMPINGEMENT (NACA 66-009 AIRFOIL AT Rc = 6.9x106 ) FROMREFERENCE11-7 It is apparent that the insect density varies greatly due to geo- graphical and seasonal factors. Therefore corresponding operational requirements will also vary. A contamination avoidance system that is effective at all altitudes below 305 m (lO00 feet) should be adequate in a temperate climate but some data indicate that protection up to 152a m (5000 feet) may be required under semi-tropical conditions.

Each of the candidate systems is evaluated on its ability to provide contamination protection up to 305 m (IOO0 feet) altitude as a minimum and up to 1524 m (5000 feet) as a design goal.

(b) Roughness Criteria - The permissible roughness is a function of cruise altitude, the chord_ise distance from the attachment line (similar to stagnation point of two dimensional flow), and the type of roughness.

A maximum allowable height of 0.IO2 mm (0.004 inch) was used for system evaluation and preliminary design, based on Reference ll-9.

(c)

Ice Protection/Contamination Avoidance System Compatibility - The ice protection system must not allow water to run back into an area from which it cannot be removed and couldsubsequently freeze. All thermal systems (such as hot air and electrical heating) permit some runback that will trip the boundary layer and are unacceptable. The ice pro- tection system must be certificable by the FAA and must meet the require- ments of FAR 25. Laminar flow is to be maintained after encountering continuous maximum icing condition or intermittent maximum icing condition as defined by FAR 25.

(d) Space Compatibility - The contamination avoidance system must be designed within the space constraints of the leading edge box and be compatible with the space requirements of the structure and the suction system. Of particular concern are the volume requirements of a Krueqer-type shield that protects the leading edge against contamination which must also be properly sized and located for aerodynamic performance. Another potential conflict of requirements is the liquid dispensing versus suction area requirements, especially in the region of attachment line travel.

(e) Suction System/Contamination Avoidance System Compatibility - It is highly desirable to use the same surface material and liquid for the contamination avoidance system and for the ice protection system.

To effect this integration, the porosity requirements of the two systems and the spreading characteristic of the liquid must be compatible.

Also, the liquid must not clog the porous surface after low altitude application or a method must be devised to clear the porous surface.

(f)

Aircraft Operational and Maintenance Compatibility - The contamination avoidance system should not require special flight procedures that will seriously degrade performance or affect safety. Consideration must also be given to crew workload, worldwide availability of materials, environmental pollution, and ground maintenance including the ability to fabricate, install and replace all system components.

11.3 POSSIBLE SYSTEMS CONSIDERED The selection of several candidate concepts was the first step in arriving at a feasible contamination/ice protection system for the LFC aircraft. The task included: Compilation of all conceivable methods that could be employed to protect the LFC aircraft from the effects of contaminants and/or ice. This includes methods obtained from a review of LFC literature, similar industrial applications, and brainstorming sessions.

Formulation of a theoretical basis and conceptual design for each conceivable method. In some cases, this would include a rough order of magnitude analysis to assess the effect of the parameters involved.

Evaluation of each method to determine the critical technical areas that require development in order to assure feasibility.

Initial evaluation of the various concepts to eliminate infeasible approaches in order to concentrate the effort on those that offer the greatest probability of success and/or the most advantageous solution to the problem. This provided a list of candidate systems for more detailed design studies.

The various contamination avoidance concepts considered fall into five main categories: mechanical, liquid, electrical, disposable covers and surface materials. Each category is discussed separately since many of the problems involved are common to all concepts in a given category.

ll.3.1 Mechanical Systems Mechanical systems that will result in a wing surface tha meets the smooth- ness criteria for laminar flow have been subdivided into three types: shields, covers and scrapers. Generally, the mechanical systems require some mechanism to deploy, operate and stow the apparatus; this may require rigorous control of tolerances to meet the smoothness requirements.

Designing an ice protection system compatible with mechanical cover or scraper systems is not simple, but a preliminary design for a possible electric de- icing system is presented in Reference ll-lO. There is concern, however, regarding the implementation of the electrical heaters into the LFC suction surface and about the power requirements of such a system.

ll.3.1.1 Retractable Shield - Reference ll-ll describes wind tunnel tests of a shield concept in which a curved plate was extended forward through a slot on the upper surface. A degree of insect protection was obtained using this approach; however, aerodynamic and mechanical problems were created relative to increased drag, pitching moment, weight, and wing surface smoothness.

If suction is not required on the lower surface, a shield similar to a variable cambered Krueger (VCK) could be deployed. The primary purpose of the shield would be to prevent impingement of insects on the fixed wing. Secondary benefits, such as providing ice protection and high-lift, could also be obtained.

To protect against contaminants, the shield would be extended during all

aircraft operation below 1524 m (5000 feet) including taxi, takeoff, climb,

descent, approach, and landing. Full ice protection would require the shield

to be extended during all icing conditions unless wing leading edge ice

protection were provided independently. Extension of the shield could require

special operating procedures with possible restrictions on speed. A thorough

study of the operational impact of shield deployment during icing conditions is necessary to determine the acceptability of this approach or the need for supplemental ice protection.

An alternative is an internally stowed shield. The rigid shield would normally be stowed in the wing leading edge cavity forward of the front spar. It would be deployed through a slot in the lower wing surface near the leading edge.

The size of the shield would be restricted by the space available forward of the spar after the installation of the suction system and structure.

_ Figures ll-5 and ll-6 show two possible mechanical shield leading edge protection systems.

11.3.1.2 Retractable Cover - This concept may also be retractable. A cover differs from a shield in that it is in direct contact with the protected surface.

A thin foil cover such as 0.061 mm (0.002 inch) thick mylar, could be extended across the leading edge. After flight through the contaminated region, the cover could be retracted. A continuous drive could be used to pull the cover along the wing from a stored roll in the fuselage.

The exposed drive mechanism would need to meet the aero smoothness requirements.

11.3.1.3 Mechanical Scraper - Early designs and tests of a variety of mechanical scrapers using wires and felt pads indicated that this method might be feasible (References 11-8 and II-II). In these designs, some type of external frame was used to support the scraper. The felt pad, which would be more compatible with an LFC surface, was propelled along the leading edge of the wing by either mechanical or aerodynamic forces.

ORIqi[,_ALPAGE IS

OF PO,.O:._IQUALITY I I LFC PANEL LD DEPLOYED ELD STOWED \ ; GEN 7274 " FIGURE 11-5. PRELIMINARY CONCEPT OF INTERNALLY STOWED SHIELD I __ FIGURE 11-6. PRELIMINARY CONCEPT OF RETRACTABLE SHIELD

II.3.1.4 Recommended Approach Mechanical System - The most practical

approach for a mechanical system is the use of a Krueger-type shield with laminar flow on the upper wing surface only. A high lift Krueger flaD is currently being used on commercial aircraft and a shield of this type could provide substantial protection against contaminants and ice. Analyses and tests will be required to determine if supplemental protection is needed.

A completely retractable shield may also be feasible for an LFC aircraft with laminar flow on both upper and lower wing surfaces.

With an adequate shield, the wing suction system would be completely free from contamination protection requirements; and with the shield performing as a high lift device it would also improve the aerodynamic performance of the aircraft. The shield could also be designed to meet the ice protection requirement, further simplifying the suction system design.

II.3.2 LiQuid Systems Liquids could be used to protect surfaces against deposits by acting in any one of three basic ways: (a) A film of liquid over the surface could cushion the impact of the contaminant and maintain the deposit in a semi-liquid state preventing adhesion to the metal surface.

(b)

A liquid could be used as a washer to dissolve the excrescence and/or the bond between the deposit and aircraft surface.

(c) A liquid spray could be used to erode the excrescence by the impact force of the impinging droplet.

These methods can be used separately or in conjunction with each other; they may be used continuously, intermittently, or after the encounter in the case of a washer or spray system.

The liquid system can be used for both insect protection and ice protection

if the fluid acts as a freezing-point-depressant (FPD). Freezing-point-

depressants have been used in anti-icing and de-icing applications for the

protection of propellers, wings, cowls and other surfaces for many years.

The fluids used for FPD are usually mixtures of alcohol (ethyl, methyl or isopropyl) and ethylene glycol. The most fully developed method of distributing the FPD on a wing leading edge is through a porous surface, although the use of spray nozzles is also feasible.

For LFC, two potential problems are common to all liquid systems. (1) Clogging of the suction orifices with liquid or contaminants (even when the suction system is not active) will require special procedures during flight to expel the liquid from the porous materials. Periodic ground cleaning procedures may also be required to remove any cumulative residue. (2) Runback of the liquid and contaminants to surfaces further aft on the wing where any residue could be deposited if not removed by aerodynamic forces. Laboratory and wind tunnel tests of the various concepts with candidate LFC surface materials are required f.

to evaluate these problems for a specific design. In general, the liquid would need to have a low enough viscosity for it to move quickly across surfaces and avoid these problems. Additional considerations in the selection of a suitable liquid are wetting ability to provide good spreading over the surface to be protected, low evaporation rate to maintain the protection over the complete impingement area, a low freezing point, and a low surface tension to aid purging of the pores.

11.3.2.1 Liquid Film - With sufficient continuous application of a liquid film over the wing leading edge, contaminants impinging on the Wing would be cushioned and carried aft by the liquid to a point at which they would be blown off the wing by aerodynamic forces. One method of applying the liquid film is by a means similar to that used for the TKS fluid de-icing system (Reference 11-12). The TKS fluid de-icing system consists of a fluid reservoir, pump, filter, metering unit and liquid distribution panel at the surface. The distribution panel is a porous metal section that is an integral part of the wing and substitutes for the wing leading edge. By means of this porous leading edge, the fluid is exuded over the surface in and on both sides of the stagnation region. The de-icing fluid spreads in a limited region on dry surfaces and is miscible with water to enhance spreading when flying through a moist atmosphere. If system activation were delayed until after formation of an ice cap, the fluid would spread throughout the ice cap to de- ice the surface. The fluid reduces the freezing point of the ice resulting in a slush that is blown away (Reference ll-13). The TKS de-icing system is currently used on the Short Brothers Skyvan SC.7 and the Hawker Siddeley HS125 executive jet aircraft.

Other methods of introducing a freezing-point-depressant that appear to be feasible include the use of other types of porous panels, leading edge orifices, or spray nozzles. It may be possible to use the porous LFC suction surface to distribute the freezing-point-depressant. The final choice of the type of fluid distribution system will be influenced by the space available and the roughness criteria. Figures 11-7 through 11-9 depict some of the schemes considered for a liquid leading edge protection system.

The performance of a liquid film protection system was investigated by Coleman (Reference II-II). Coleman reported a_definite increase in the flow requirement at higher ambient temperatures. In the normal icing temperature range of 30-35°C about 0.069 Kg/s/m2 (0.85 Ib/min/ft2) of water would be required to keep the surface clean. At 50°C the required flow rate is increased by 50 percent.

These flow rates result in a feasible system weight for a reasonable area of coverage. Coleman used a solution of water with 2 to 30 percent detergent added to assist in distributing the fluid over the surface to be protected.

Some of the potential problem areas of a liquid-film/freezing-point-depressant system for contamination and ice protection are: (a) Space limitations - The requirements of the suction system in terms of both the surface area in which suction is required and the ducting and support structure within the wing severely limit the space

C

available for fluid distribution panels. If space limitations negate PAGE IS ORIGINAL OF POOR QUALITY ITION CHAMBER PRESSURE CHAMBER DISTRIBUTION MEMBRA POROUS SKIN MANIFOLD CONTINUOUS WEL STAGNATION POINT VARIATION ; AIR FPD FLUID B3_IEN-23]_5 FIGURE 11-7. FREEZING POINT DEPRESSANT LEADING-EDGE DISTRIBUTION STRIPS WITH ALTERNATE AIR AND FPD PASSAGES

C

SLOTTED LEADING EDGE POROUS OVER FPD PASSAGE t S_H INTEGRAL PASSAGES _ _ LFCAI R PA_SSAGE \--_----FPD PASSAGE -- _ - _ =====_ 7-GEr_ 22749 PRELIMINARY CONCEPT OF FREEZING-POINT DEPRESSANT FIGURE 11-8.

DISTRIBUTION SYSTEM the use of porous panel distribution, the liquid film concept may still be feasible using a spray type distribution system. Mounting the spray nozzles on a Krueger-type shield is possible.

(b)

Leading Edge Erosion - An early problem associated with TKS porous panels was that water hitting the pores produced an explosive erosion effect. Several layers of steel calendared wire cloth presently used, resist erosion better than previously used materials and appear to be adequate for in-service use. Rain erosion tests were run on Dynapore and electron beam perforated surfaces. See Section 9.3.4.

The calendared wire cloth used by TKS is much coarser than that investigated as an LFC surface.

11.3.2.2 Washer System - A washer system could apply a cleaning fluid to a contaminated surface to weaken or dissolve the bond between a deposit and the aircraft surface. This system may be used alone or as part of a liquid-film system that also cushions the impact of an insect or a water spray that erodes the deposit. The unique feature of the washer system when compared with a liquid- film is that the washing fluid may be dispensed in a single application after the contamination encounter. This may result in a lower weight system. However, the more stringent requirements of a cleaning fluid will tend to make it more difficult to find a fluid compatible with a freezing-point-depressant ice protection system. Feasibility of the washer system depends primarily on locating a good cleaning agent that is also a freezing-point-depressant. Lachmann has demonstrated the feasibility of one type of washing system in Reference 11-9.

11.3.2.3 Erosion by Water Spray - When a liquid drop collides with a solid surface it exerts a localized pressure and then flows out radially from the point of impingement (Reference 11-14). In the case of a water drop colliding with a glass plate, the flow velocity immediately after the impact has been found to approach ten times the value of the impingement velocity (Reference 11-15).

The radially flowing liquid then exerts a shear force on the surface of the solid over which it is running and a pressure force on any protrusion that it encounters. The combination of impact pressure, flow shear force, and flow pressure acting on a protrusion can be made to exceed either the cohesive or adhesive bond strengths and remove the contaminant.

Lachmannreported the results of insect removal by water spray in a wind

tunnel (References ll-8 and ll-9). The tests evaluated the effect of jet

size, water pressure, wind speed, and fluid properties. A two percent

solution of liquid soap in water was found to work well probably due to the

combined effect of erosion and washing as described previously. Lachmann

concluded that "it is feasible to wash fly deposits off a wing leading edge

by water jets and that the quantity of water required is .... little more than ....

1.5 kg/m (l Ib/ft) length of leading edge...".

Tests by NASAusing a JetStar aircraft were less encouraging; hard-shell portions

of insect remains were not eroded. It would seem that a more comprehensive investigation is required.

II.3.2.4 Recommended Approach Liquid System - Considering the functional, operational, and compatibility requirements defined in the INTRODUCTION, a liquid system that would appear to have the best change of success for con- tamination avoidance of the wing leading edge would include the following features :

C

(a) Distribution - spray nozzles, mounted on the shield, to minimize the complexity of the wing leading edge design and the conflict of space requirements with the suction system, although dispensing the liquid through the porous surface would use less liquid and is preferable if a practical design could be achieved.

(b) Fluid Properties - must include some form of freezing-point-depressant, contain a wetting agent, have low viscosity, and low surface tension.

(c) Flow Control optimal operation would include variable or intermittent

flow capability and ability to maintain a constant flow over a reasonable temperature range.

11.3.3 Electrically Powered Systems A number of electrical/magnetic phenomena could be utilized for removing insect deposits, ice or other contaminants. The Russians have developed an Electro-lmpulse (El) de-icing system to shed ice. Ultrasonic vibration has been used to clean contaminated surfaces. Home ovens are being self- cleaned by pyrolysis and electrical heating is a possibility. Magneto- hydrodynamic forces have been used to divert flow fields. Each of these concepts was evaluated for applicability to the LFC contamination protection requirement.

11.3.3.1 Electro-lmpulse - This de-icing concept is based on the forces developed by an induced magnetic field in the aircraft skin. A pulse of electric current is passed through a coil located in close proximity to the skin to be de-iced. The magnetic field created by the ceil induces eddy currents in the conductive skin panel. The interaction of the magnetic field and induced currents causes mutual repulsion between the skin and coil (Reference 11-16), creating a hammer blow effect that sheds the ice formed on the surface.

The major problem areas anticipated with the use of the E-I system are (a) the space limitations due to the suction system requirements (b) the requirement for the E-I system to deflect the skin coupled with the expected rigidity of the LFC outer panel, and (c) the possibility that the E-I system will not shed insect deposits. Generally, the E-I system ice shedding capability increases as ice cap thickness increases. The small mass of contamination may not produce the forces necessary for shedding.

Although direct use of the E-I system for insect protection may not be feasible, the E-I principle may be used with a coating. The coating may be a simple protective film or an electrically conductive film. A simple protective film would be removed in a manner similar to de-icing. The high energy impulse would accelerate the skin and the coating would break loose due to inertial forces. An electrically conductive coating would be subjected to an electro- magnetic repulsion force in addition to the inertial forces and may, therefore, require a lower energy pulse for removal.

11.3.3.2 Ultrasonic Vibration - A commonly used method of industrial cleaning is the immersion of a contaminated part in a liquid bath that is subjected to ultrasonic vibration. This cleaning method uses the "brushless scrubbing" effect of cavitation produced by the ultrasonic pressure waves. The process is especially effective for intricate parts and/or difficult to remove contaminants (Reference ll-17).

A conceivable inflight use of ultrasonic vibration would be to provide a large volume of fluid over the wing and simultaneously vibrate the surface. Practical considerations would include the thickness of the liquid layer required to sustain cavitation, the length of time that the liquid would need to be applied, and the physical installation of the transducers.

In view of the large quantity of liquid necessary to sustain cavitation, the difficulties expected in designing the transducers, and the space limitations expected to be imposed by the suction system, an in-flight ultrasonic cleaning system does not appear to be feasible. However, there may be justification for £ consideration of ultrasonic cleaning for ground use.

II.3.3.3 Thermal Decomposition - Contamination due to insect deposits can be removed by the application of heat. If the aircraft surface is heated sufficiently, either the bond between the insect deposit and aircraft would be weakened or the insect deposit would be thermally decomposed. In either case, the surface would be cleaned in a manner similar to that used for self- cleaning ovens.

Two major considerations must be evaluated before pyrolysis can be said to be feasible: compatibility of thermal decomposition requirements with material temperature limits and available electrical power. Self-cleaning ovens are usually operated at temperatures above 455°C (850°F) which is well above an acceptable temperature for aluminum. The power required to maintain this temperature level also appears to be excessive even if segmental cyclic heating is used. Based on these considerations, thermal decomposition of insect deposits does not appear to be a practical approach for insect protection.

11.3.3.4 Magnetohydrodynamics - Theoretically, MHD can be used to deflect the flow field and thereby control insect impingement. Two considerations have ruled this approach out immediately: the high magnetic field strengths required in practical applications and the need for an electrically charged fluid to produce the force.

11.3.3.5 Recommended Approach - Electrical System- None of the electrical systems appear to be feasible by themselves as an insect protection system.

The electro impulse system might be feasible, however, if used in conjunction with a coating.

11.3.4 Disposable Covers A number of investigations have been made of covers or coatings applied to the wing leading edge prior to takeoff and removed after the insect encounter.

These covers may be categorized by the method of removal (dissolving, electrical, mechanical, or thermal) and by the form of the cover after removal (large sheet, small pieces, liquid or vapor). Four major concerns are common for these systems: (a) contamination of the environment, (b) effect of the disposed cover on the aircraft, (c) ground support and (d) the cover does not eliminate the requirement for an ice protection system on the fixed wing after the cover is removed.

"I 11.3.4.1 Temporary Coating - A coating applied on the ground could protect the surface during takeoff and climb. The coating would be automatically released and would not interfere with cruise performance of the aircraft.

Three categories of temporary coatings were investigated: polishes, controlled adhesion coatings, and subliming coatings.

A wax coating, polish or a release agent containing silicone, oils or teflon will reduce the adhesion of contaminants. These materials could be used in conjunction with a fluid system but fluid spreading would be difficult.

If highly effective, they could conceivably be u_ed without supplemental insect protection. A number of coatings are available as listed in Table II-I.

The major concern with a polish type coating is that the buildup of the coating may block or interfere with the suction perforations. Polishes would not be compatible with a porous suction surface.

8@9 .Cb aP :> -r- r_'_ >_c = c 0

o

C_.rh __._-_ 0'-- _ 0 U _.j •

o

0 0 0 -_ 4 "_ _'- © 0 CU G; _ aP _3J oO ¢D Z I-- 0 C_ _E I "_'- ,_- -I_ C_ Oh 0 c- _:_ 0 or-- N- _c_ e--

C

_h

¢D LI_ 0 O0 Oh 0_.- I_1_1 0 o") ,-- .J 0 _ "_- C_ I •_- 0 "_ 0 0_.- 0 _ _ I-- e- I .4_ ORIGINAL PA@E I_9 OF POOR QUALITY "0

o

o_,,.

¢_) _J U .t- _- "0 _D I.l- _J ot--

=

-'s 0 c_ O I ._ or-- ,,= ,-- i.=.,.,.

c- 4- c- ._ i'I:1

_ _o o

If) -r _C I-- '-I- A controlled adhesion coating could be applied prior to takeoff. Shear forces could be used to peel or flake the coating from the surface. Con- sistent adhesion control would be difficult as well as the problems previously described for polishes.

A subliming coating is conceivable which will have evaporated due to low pressure and airflow prior to reaching cruise altitude. This type of temporary coating is discussed under "Thermally Removed Coatings".

11.3.4.2 Chemically Removed Coating - A film may be applied to the wing leading edge prior to takeoff and washed off after climbing through the insect envelope. The film could be a gel or viscous fluid and remain in place for the initial climbout. Insects would impinge on the film and be trapped without impinging on the wing surface.

Tests run by Coleman (Reference 11-18) indicated that this system is probably feasible. Glycerine, glycerine and gelatine, and soap in methanol were tested. Glycerine by itself did not completely protect the surface. A mixture of 60 percent glycerine, 30 percent gelatine, and I0 percent wetting agent provided complete protection. Soap in methanol required much more water for removal. Conceivably, the washer system used to remove the soluble film could also be used for ice protection.

Two potential problem areas need to be explored before feasibility can be established: (a) find a material that is easily dissolved and will not clog the suction pores and (b) ensure that the insect residue is blown off the wing surface and does not adhere during its travel across the wing surface.

Although this method appears to be feasible, the complex ground operating procedures required to apply the coating reduce its attractiveness. This type of system should be considered further if some of the more desirable approaches prove impractical.

II.3.4.3 Electrically Removed Coating - As discussed previously, a coating

may be removed by accelerating the skin or coating by means of the Electro Impulse (E-I) principle. Although no experimental work has been performed on this system, it is conceivable that a metalized sheet or coating could be applied prior to takeoff. After climbing above the insect envelope, the cover could be removed in one of several ways. If an E-I system is used for ice protection, the same system can be used to fragment a coating into small pieces which would be dispersed by aerodynamic and wind forces. A sheet- type cover, similar to that used by Lachmann (Reference ll-9), could be removed by a single E-I coil located at the wing leading edge near the root.

Pulsing the E-I coil could be made to repulse the corner of the cover and initiate a tear along the leading edge. Aerodynamic forces could be made to complete the ejection of the cover in two halves. A metalized cover could also be held in place by magnetic force. After climbout above the insect level, the magnetic attraction could be interrupted and aerodynamic forces or electromagnetic repulsion used to remove the cover.

11.3.4.4 Mechanicall_ Removed Cover - The contamination sensitive surfaces could be covered during takeoff and initial climbout. Upon reaching a pre- determined altitude which is well above the insect envelope, the cover would be jettisoned. Lachmann described such a design in Reference 11-9 and Coleman reported work as long ago as 1945 by Smith and Higton (Reference II-II) on this concept.

A variety of materials has been studied for this application including paper, tracing linen, cellulose sheet, plastic spray, and lacquers. To date, the greatest success has been achieved using a self-adherent, thin cellulose fiber matting. The cover was slit along the leading edge by a cutter that was subsequently retracted into the fuselage. Aerodynamic forces removed the cover.

One concern relative to the use of this system is whether the Environmental Protection Agency would allow ejection of the cover over land. A biode- gradable cover would probably solve this problem; however, a cover that meets this additional requirement has not been found to date. Ground main- tenance requirements are also a concern.

11:3.4.5 Thermally Removed Coating - The leading edge can be protected against contamination due to insect impingement by the application of a coating that is removed by heat. A layer of ice formed on the leading edge and subsequently removed by a thermal de-icing system is a feasible example of such an approach. Other general subcategories are:

(a)

A flammabl_ coating that is removed by an ignition source, e.g., an electric spark. The method was mentioned by Coleman (Reference II-II) and rejected on the ground of safety.

(b) A volatile coating that is boiled off or evaporates in flight.

This method was also mentioned by Coleman who reported unsatis- factory results during early tests. Although this method is probably feasible using thicker coatings and/or different materials, practical considerations minimize its attractiveness.

(c)

A coating that melts, decomposes, or sublimates; e.g., an ice cap.

Coleman experimented with this type of system and concluded that a layer of ice 6.3 to 9.5 mm (I/4 to 3/8 inch) thick should suffice to protect the wing from insects. On extremely hot days, a thicker layer would be required. An integral intensive cooling system in the wing leading edge would be needed. Sustaining an ice cap on the wing without cooling equipment would appear to be impractical due to the typical delays in takeoff clearance at many high density airports.

An ablative material can be envisioned that could be applied quickly on the ground and would be stable up to at least 93°C (200°F). A thermal ice protection system could be used to remove the ablation material at high altitude. Although this method is probably feasible, it does not appear attractive.

11.3.4.6 Recommended Approach - Disposable Cover - Disposable covers have three major disadvantages: contamination of the atmosphere, possible aircraft damage, and ground maintenance requirements. Soluble and electrically removed covers appear to have a low to moderate chance of success. Further investigation and testing is warranted to assess their practicality if the more promising approaches prove to be infeasible.

11.3.5 Special Surface Coatings for Contamination Avoidance A permanent coating may be applied to the wing leading edge to either prevent contamination or to facilitate contamination removal. Three types of surfaces have been considered: (a) icephobic or anti-stick materials that have low bond strengths with ice or insect deposits, (b) elastic materials that cushion insect impact, and (c) hydrophylic materials that are easily cleaned.

11.3.5.1 Anti-Stick Coatings - A low bond strength between the ice or insect deposit and the wing surface can be achieved using any one of three basic approaches: (a) use of a superslick or icephobic material that possesses inherent low bond properties, (b) use of additives that cause a surface to exude oils or similar substances, and (c) use of a very hard, smooth surface.

It has been postulated that aerodynamic forces would shed ice and other deposits from such surfaces. A number of coatings including Teflons, silicone elastomers, fluorocarbon elastomers, and greases (see Table II-I) have been tested in icing tunnels. To date, no coating has been found that provides auto-release of the ice formed under these similated conditions (References 11-19 and 11-20) although some of the coatings did reduce the forces required to release the ice, compared with an untreated aluminum surface.

Four other major concerns exist in the use of anti-stick coatings: (I) Is the coating compatible with the roughness and waviness criteria for LFC? (2) Is the coating sufficiently resistant to rain, dust, and ice erosion to be practical?

(3) Is the coating compatible with the suction system, i.e., can the coating be applied and not interfere with holes, slots, or porosity? And (4), will an anti-stick material shed insects? Limited experience on the Jetstar aircraft indicates poor insect shedding characteristics.

In view of the high probability of a negative answer to most of these questions, it is concluded that an anti-stick coating by itself is not a feasible insect protection system. However, it appears that anti-stick materials could be used in specialized applications with other insect/ice protection systems such as washers, shields, or vibrators. Their presence may increase the primary system efficiency enough to result in an overall weight and cost saving.

11.3.5.2 Elastic Coatings - Elastic coatings proved to be effective in cushioning the impact of insects at low velocity. By avoiding disintegration, the elastic coating prevents contamination by reflecting the insect from the surface. Three millimeter-thick, solid rubber and foam rubber with shore hardness 10-35 proved to be effective at velocities up to 56 m/s (108 kn) (Reference 11-21).

Some of the problem areas anticipated with the use of elastic coatings are: (I) erosion at moderate to high speeds, (2) ineffectiveness of insect protection at high speeds, and (3) incompatibility with suction and ice protection systems. Like anti-stick surfaces, elastic coatings do not appear to provide adequate insect/ice protection by themselves, but may prove useful in con- junction with other insect protection concepts.

11.3.5.3 Hydrophylic Coating - A hydrophylic material has an affinity for water so that it is easily cleaned with a water or alcohol rinse. One such coating is under development (Table II-I) but to date no contamination tests have been conducted. Data are limited due to the newness of this concept.

Further advances in hydrophylic materials development should be monitored to establish usefulness of this concept. The Jetstar program has investigated two hydrophylic coatings with results about the same as for the "superslicks" 11.3.6 Contamination Protection - Initial Conclusions The preliminary evaluation of the various concepts that could provide con- tamination and/or ice protection for the LFC aircraft wing leading edge indicates that a number of approaches offer promise of being feasible.

Three approaches were selected as offering the best chance of success or having unique characteristics that are desirable for the LFC aircraft.

(l) A shield that normally forms part of the wing lower surface that

can be extended to a position in front of the leading edge. The shield would contain a liquid system for contamination and ice protection.

(2) Spray nozzles located in the wing leading edge or in the aft face

of a shield, that distribute a washer/deicing fluid over the impingement area. The distributed liquid is required to inhibit the adhesion of insects and act as a freezing-point-depressant.

(3) A liquid system similar to (2) above except that the spray nozzles are replaced by a porous distribution panel in the wing leading edge.

11.4 SHIELD PRELIMINARY DESIGN AND PERFORMANCE EVALUATION Investigations were conducted to determine the compatibility of the various shield concepts (rigid, folding and variable camber) with the space constraints and the performance requirements of the LFC aircraft. This section presents the results of design and analytical studies and the resulting design recom- mendations.

11.4.1 Preliminary Design Studies Feasible installations of the three types of shields are shown in Figures II-I0 through 11-12. Figure II-I0 shows a rigid shield with a simple hinged deploying mechanism. Although this concept limits the shield size and location of the trailing edge, the simplicity and compactness of the design may make it suitable for early flight tests and in areas of limited space. This concept would provide the least aerodynamic benefit to the aircraft. Flexible hoses are used for the liquid supply lines and are routed along and supported by the hinges. An alternate design using rigid tubing and swivel fittings is also feasible.

C_ ORIGINAL P/-tL. ;:." ;_ OF POOR QUALITY LFC PANELS

SPRAY NOZZLES

POP-UP TYPE ALTERNATIVE FLUSH SPRAY NOZZLE CONCEPT FPD SUPPLY MANIFOLD FIGURE 11-9. PRELIMINARY CONCEPT OF FREEZING-POINT DEPRESSANT SPRAY ON FIXED LE (1) SHIELD SURFACE MAY BE USED FOR FPD DISTRIBUTION (2) ELECTRICAL IMPULSE DE-ICING MAY BE USED ON TIP I I

I

1,3-GEN_231 _ I I FIGURE 11-10. RETRACTABLE SHIELD - DEPLOYED FOR INSECT/ICE PROTECTION OF pOOR Q,I,_i_L'="Y I_GEN-23131B FIGURE 11-11. FIXED CAMBER SHIELD (EXTENDED) --INSECT PROTECTION FLUID SHIELD_ _ __i-_----- 03,GEN-2_I ] 42 FIGURE 11-12. VARIABLE CAMBER SHIELD (EXTENDED) The fixed-camber, folding-shield concept (Figure II-II) permits the use of a larger shield with increased protection for the upper wing surface. The complexity of the linkage and drive mechanism will require a more extensive development effort than the rigid design.

The variable-camber shield shown in Figure 11-12 combines the good shielding qualities of the folding shield with high lift capability and is the most likely arrangement to be used in a commercial application. The length of the shield shown in the Figure is 12 percent of the chord length. The contour of the shield is similar to the high lift device designed for a Douglas Advanced Commercial Aircraft. This design was used for many of the preliminary shielding effectiveness analyses.

If ice protection of the shield is required for aerodynamic or stowage con- siderations, two methods are feasible. Flexure of the variable-camber devices would provide some de-icing capability as on current commercial applications.

The preferred method is the use of a freezing-point-depressant applied through a porous distribution panel as described in Section 11.5.

11.4.2 Contamination Avoidance Effectiveness Analyses As the laminar flow aircraft encounters an insect or other potential source of contamination, the insect will be affected by the flow disturbance of the approaching aircraft. The initial position and motion of the insect relative to the wing determine whether or not the insect will impinge on the wing and create a source of turbulence. This problem is very similar to the problem of ice accretion on a wing leading edge, for which a number of analytical solutions are available (References 11-22 through 11-24).

11.4.2.1 Computer Programs - The most applicable tool available at Douglas for this type of analysis is computer program H9PB, "Water Impingement on Two- Dimensional, Multi-Element Airfoils". This program has the capability of analyzing the trajectory of a droplet in the flow-field induced by an airfoil composed of as many as ten elements. An iterative, stepwise solution is used to solve the differential equations of motion in cartesian coordinates. The coef- ficient used in the equations to described the drag-inertial characteristics of the droplet is CD Re , where 24 K CD is the drag coefficient Re is the Reynolds number based on air properties, droplet diameter and droplet relative velocity 2 r 2 Vo _w K is the inertia parameter 9 C _a g from Reference 22 Coleman, in Reference II-25, derives similar equations for insect trajectories using a drag-inertial coefficient defined as C k Vd, where: To C is the chord length k is the insect ballistic coefficient in (length) -I units Vd is the insect relative velocity Vo is the free stream velocity To permit insect trajectory analyses, computer program H9PB was modified so that the insect ballistic coefficient (k) and airfoil chord (C) would be substituted in appropriate places for the droplet parameters.

Flow field velocities and body coordinates were generated by a separate computer program using the Neumann method for angles of attack of 0 ° and 90 o .

These data are input to program H9PB which modifies them for the angle of attack being considered. Special techniques are provided to obtain sufficient accuracy of the flow field near body surfaces.

The output of computer program H9PB includes a tabulation of the coordinates of the droplet (or insect) relative to the airfoil as the particle moves towards the airfoil. Particle and air velocities are also tabulated. To facilitate the evaluation of the data, a graphics computer program was developed to display the trajectory output data of H9PB.

II.4.2.2 Shielding Analysis - Three existing airfoil designs, available from the Douglas Advanced Commercial Aircraft (ACA) program, were used as the basis of the study. (1) A two-element section at the 90 percent span wing station was used to evaluate the wing tip and zero flap conditions. (2) A four-element section at the 80 percent span wing station was used to evaluate the effect of full flaps (50°). And (3), a model at 13 percent with a flap angle of 27 I/2 ° was used to evaluate the inboard airfoil which has a much blunter leading edge that extends above the shield.

Another major variable in the analysis is the angle of attack. Contaminants may impinge on the wing leading edge during the takeoff roll, rotation, and o climbout. During this time, the aircraft angle of attack may vary from 0 ° to 9 o . To determine the range of airfoil angle of attack, it is necessary to add the wing incidence angle. Figure 11-13 shows the wing twist used to determine the local incidence angle. Note that the wing tip is at -2 o when the wing root is at +5.5 o . Angles of attack of -4 o , _o and 15 o were used to bracket the anticipated range of wing section angle of attack with respect to the free stream. Other parameters that were varied during the analysis included insect, the ballistic coefficient (k), the airfoil chord length (C), and the aircraft velocity (Vo).

A number of runs were made to correlate the results of the insect trajectory analysis with results of droplet analyses under identical conditions. A 900 micron droplet trajectory was found to be almost identical to an insect trajectory with a k of 0.66/m (O.2/foot). An insect having a k of this value weighs about 0.8 mg. A spherical droplet weighing 0.8 mg would have a diameter of 1150 microns.

11.4.2.3 Trajectory Analysis -

(a) Flaps Retracted - The two-element airfoil section at the 90 percent wing span location was chosen to represent the wing tip and the main wing area after flaps have been retracted. An insect ballistic coef- ficient of 0.66/m (O.2/ft) was used. This represents an insect weighing about 0.8 mg which is heavier than about 90 percent of the insect population. Two angles of attack were selected: 0 ° to represent the worst condition for impingement above the shield and +15 o to represent the worst condition below the shield.

Figure 11-14 presents the results of the trajectory analysis at 0 ° angle of attack. A small amount of impingement is noted above the shield and none below. The trajectory analysis predicts that insects starting at a position between 0.030 and 0.035 chord lengths above the leading edge will impinge on the wing. This represents a shielding effectiveness of about 94 percent.

The fact that insects impinge on the wing above the shield does not necessarily mean that the resulting contamination will cause turbulence.

Based on data from Reference 11-25 it can be seen that insects that impact at an angle of less than about 7o with respect to the surface do not leave a residue greater than 0.102 mm (0.004 inch) height. The maximum angle of impact from Figure 22 is 5o which implies that the insect residue will be less than 0.102 mm (0,004 inch) height, and should not cause transition.

Figure 11-15 presents similar results at an angle of attack of 15 o .

Since the shield is designed for LFC on the upper wing surface only, impingement as shown is acceptable.

(b)

Landin 9 Flaps - A four-element airfoil section model at the 80 percent wing span location was available for this analysis from the Douglas ACA program. The flaps were extended to 50 o , representing the approach and landing situation. Three angles of attack (_) were considered: LU WING r_ INCIDENCE ,,=.1 U.I --U3 (iw) 0 (DEGREES) u.

O IA,I C3 m or) --5 I I i I I I = J = 0 10 20 30 40 50 60 70 80 90 100 (PERCENT) SEMISPAN ('Q) 83GEN 23154 FIGURE 11-13. LFC WING CHARACTERISTICS ACA WING AT 90-PERCENT SPAN LOCATION CHORD LENGTH = 22.2 m (87.4 IN.)

INSECT AERO COEF (K) =02.00 PER FT AIRSPEED = 74.6 m/s (145 KIAS) ANGLE OF ATTACK = 0.0 DEGREES .+ START POSITION 0.035 0.030 -1.10 -120 B3-GEN-23136 FIGURE 11-14. INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF -4 ° and 0 ° representing the landing roll out phase which is the worst case for impingement on the wing upper surface and +15 ° representing the approach and flare out phases during which impingement under the shield is critical.

Figure ll-16 presents the results of analyses at m = 0 ° and insect ballistic coefficient (k) of O.l. The results indicate that insects will not impinge on the upper surface under these conditions.

_ _40 Figure II 17 presents similar results at _ = . Impingement is possible only with very dense insects (k > O.l).

Figures If-18 through ll-20 present the results of an analysis at = +15 ° illustrating the effect of k varying between O.l and 0.35.

The concern is that the lighter insects would be forced upward by the airflow between the shield and the wing and impinge near the wing leading edge. In cruise, the attachment line point is in the region of ± 0.6 percent of a chord length from the leading edge and, therefore, any point above the -0.6 percent point should be considered as the "upper surface" and must be protected.

Figure ll-18 shows no critical impingement for large or dense insects.

The other figures in this group show that light insects that start in a narrow band (0.5 to l.O percent chord wide) could impinge near the wing leading edge.

If it is assumed that the wing without the shield would intercept insects in a band lO percent of chord wide, the Krueger is about 90 percent effective in shielding the leading edge under the foregoing conditions. At lower angles of attack or flap setting, the shield would be more effective.

It is not essential to have fully effective shielding in the landing configuration because the leading edge could be wiped clean before takeoff.

ACA WING AT 90-PERCENT SPAN LOCATION INSECT AERO COEF (K)= 0.200 PER FT CHORD LENGTH = 2.22 m (87.4 IN.)

ANGLE OF ATTACK = 15.0 DEGREES AIRSPEED = 74.6 m/s (145 KIAS) C _ 0_._: _L "_......

OF POOR QUaLiTY .& B3-GEN-23139 FIGURE11-15. INSECT TRAJECTORIES-SEA LEVEL, TAKEOFF • ACA WING AT 80-PERCENT SPAN LOCATION INSECT AERO COEF (K)=0.100 PER FT CHORD LENGTH = 2.47 m (97.1 IN.)

ANGLE OF ATTACK = 0.0 DEGREES AIRSPEED = 74.6 m/s (145 KIAS) START posITION -0.055 -0.070 83-GEN23143 FIGURE 11-16.

INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF ACA WING AT 80-PERCENT SPAN LOCATION INSECT AERO COEF (K) 0.100 PER FOOT CHORD LENGTH=2.47 rn (97.1 IN.)

ANGLE OF ATTACK -4.0 DEGREES AIRSPEED=74.6 m/s (145 KIAS) \ ORIG',NEL F'I_;:: _, i_ OF POOR QUALITY START POSITION 0.26 0.25 B3GEN 220_0 FIGURE 11o17.

INSECT TRAJECTORIES - SEA LEVEL TAKEOFF, --4-DEGREE ANGLE OF ATTACK ACA WING AT 80-PERCENT SPAN LOCATION INSECT AERO COEF (K) 0.100 PER FOOT CHORD LENGTH 2.47 m (97.1 IN.)

ANGLE OF ATTACK 15.0 DEGREES AIRSPEED 74.6 m/s (145 KIAS)

J

/ f" 1 / J

Y

• - -_._._.._ C FIGURE 11-18. INSECT TRAJECTORIES - SEA LEVEL TAKEOFF, 15-DEGREE ANGLE OF ATTACK ACA WING AT 80-PERCENT SPAN LOCATION INSECT AERO COEF (K) = 0.250 PER FT CHORD LENGTH = 2.47 m (97.1 IN.)

ANGLE OF ATTACK = 15.0 DEGREES AIRSPEED = 74.6 m/s (145 KIAS) I I = 30% 40%_ c OF POOR QUALITY 83"GEN-2314 l FIGURE 11-19. INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF ACA WING AT 80-PERCENT SPAN LOCATION INSECT AERO COEF (K) = 0.350 PER FT CHORD LENGTH =2.47 m (97.1 IN.)

/ANGLE OF ATTACK = 15.0 DEGREES AIRSPEED = 74.6 m/$ (145 KIAS) 83-GEN-23140 FIGURE 11-20.

INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF

(c)

Takeoff Flaps - A four-element airfoil section model of the 13 percent wing span location was available for this analysis from the Douglas ACA program. The flaps for this model are extended 27 I/2 o to represent the takeoff configuration. The inboard wing section selected is much more blunt than the section outboard of the suction engine and is therefore more susceptible to insect impingement.

Figure 11-21 presents an insect trajectory at _ = 0 ° and k = 0.2.

It is apparent that insects would impinge on the upper surface at an angle steep enough to create protuberances greater than 0.102 mm (0.004 inch) in height. Figure 1!-22 shows a similar condition with the angle of attack increased to 15 °, Again, impingement on the inboard leading edge region is predicted. The band in which insects will impinge on the leading edge is about 1 percent of the chord length.

11.4.2.4 Effectiveness Analysis - Conclusions - Based on the results of the trajectory analyses, it is concluded that a Krueger-type shield will provide F substantialcontamination protection in the critical stagnation region. The analysis indicates three potential areas of concern: In the inboard area where the wing thickens, impingement could occur (I) on the wing upper surface to an extent that would disrupt laminar flow unless supplementary liquid protection were provided.

(2) In the wing tip area where the angle of incidence is negative, the

computer program predicts that impingement could occur on the upper surface during the ground roll. However, since upwash and ground effects are not included in the analysis, there is a high probability that the extent of impingement would not pose an operational problem.

(3) The computer program predicts impingement by light particles travelling under the shield at a high local angle of attack. Considering wing twist and thickness, the inboard wing is more susceptible to this condition. However, the deposit left by a light particle is not expected to be thick enough to disrupt laminar flow.

ACA WING AT 13-PERCENT SPAN LOCATION INSECT AERO COEF (K) = 0.200 PER FT CHORD LENGTH = 2.47 m (97.1 IN.)

ANGLE OF ATTACK = 0.0 DEGREES AIRSPEED = 74.6 m/s (145 KIAS) ORIGIT,6:,L [-=_.._. ,_ OF POOi_ QbAi_liTY START POSITION 0.02 0.01 83_N-23133 FIGURE 11-21. INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF

C

ACA WING AT 13-PERCENT SPAN LOCATION INSECT AERO COEF (K) = 0.200 PER FT CHORD LENGTH = 10.16 m (400 IN.)

AIRSPEED = 74.6 m/s (145 KIAS) ANGLE OF ATTACK = 15.0 DEGREES 0.6% I I I I 4O% 10% 20% 30% J 'rii',.

FIGURE 11-22.

INSECT TRAJECTORIES - SEA LEVEL, TAKEOFF ll.a.3 Wind Tunnel Test of Shield Effectiveness To determine the practical effectiveness of the shield concept, insect impinge- ment testing was done in the Icing Tunnel Facility at NASA Lewis Research Center.

The model was a DC-9 wing tip on which a representative fixed leading edge shield was installed. Insects were released upstream from a series of containers set at an angle to the plane of the wing so that a number of streamlines were represented along the span. This arrangement is illustrated in Figure II-23.

The test results supported the theoretical analysis. Angles of attack ranged from -4 degrees to +7 degrees. At an angle of attack of -4 degrees, insect contamination of the upper surface occasionally exceeded a critical height of about O.l mm (.004 inch). This indicated that a more extensive shield or a supplementary liquid protection system may be required in the wing tip region, where negative angles of attack would exist during taxiing and at takeoff roll.

II.4.4 Shield Operational Use The shield would be extended on the ground prior to brake release to provide protection from airborne contaminants and ice during the ground roll and climbout. Normally, the shield would be retracted at or below 1524 m (5000 feet) on clini_out, depending on flight test experience. If icing conditions are anticipated at higher altitudes, shield retraction could be delayed. The shield would normally be re-extended during descent at altitudes below 1524 m (5000 feet) or under icing conditions.

Normally, icing conditions are not encountered at the cruise altitudes associated with LFC aircraft operation. In the very unlikely event that icing conditions are encountered during cruise that result in the loss of laminarization, the flight could be continued in the turbulent mode or the shield could be extended and the wing de-iced. The actual operating procedure will vary depending upon the experience of the particular airline operator. The possibility of encountering such icing conditions would vary with route structure and seasonal conditions.

G: ll.5 SHIELD ICE PROTECTION SYSTEM Except for infrequent occurrences, icing conditions are only encountered at the lower altitudes of commercial aircraft operation. Generally, this limits consideration of ice protection system performance to takeoff, initial climb- out, descent, landing and holding during either departure or landing. Large accumulations of ice are the result of continued exposure of the surface to icing conditions for long periods of time. The most severe icing encounter found during the analysis of the DC-9/DC-IO fleet was a 15 minute departure hold followed by a 30 minute landing hold.

The combination of the Shield/Liquid-Film ice protection system is capable of providing protection during all of these flight conditions. If operated as a de-icing system, a fluid capacity for three cycles per flight should be adequate. For an anti-icing system a capacity for one hour of operation should be sufficient. For operational flexibility, the liquid storage capacity should be sufficient for two flights.

A schematic diagram for a liquid-film ice protection system is shown in Figure 11-24. Additional porous panel dispensing units would be fed in parallel using multiple flow-metering valves as required.

The supply tank, pump, filter and check valve would be mounted on a panel in the pressurized area of the fuselage. The supply tank would include a sight gage and a remote indication of fluid level. A slight positive pressure could be provided by the pneumatic system (or other suitable source) if desirable for pump operation.

Typically, a piston type positive displacement pump with an integral relief mechanism, driven by a 28 volt d.c. motor, would be used and the flow rate would be pre-set on assembly. The system would be designed so that the normal operating pressures fall within the constant flow region of the pump character- istic. Due to the increase in fluid viscosity, operation at low temperatures (i.eo, below -40°C) will give rise to higher pressures, and under these conditions the pump output flow rate falls so that the pressure does not exceed a pre-set value.

ORIG!NAL PAGE Ig OF POOR QUALITY \ : SIMULATED SHIELD INSECT INJECTION TUBES (5)

%

WIND TUNNEL FIGURE 11-23.

INSECT IMPINGEMENT TESTS IN LEWIS ICING TUNNEL AIR PRESSURE __FILTER/RE LIE F PORT FLUID

_u,_, FLU,D

II ..... I _ OUANTITY U I T INDICATION PUMP - CONSTANT FLOW WITH INTEGRAL RELIEF

LJ'""."-"°"'"'"'°""'°"°"

I1.---..-.-.( _ PRESSURE WARNING SWITCH SPRING-LOADED iI_ ,..4 CHECK VALVE It_ = TU It II I INBOARD II II I SHIELD i] FLOW II : ,L it METERING

II ', /IIVALVE"

,°-.-.I-° ,!11

_ OUTBUA SHIELDU POROUS j PANELS 113_EN-231G2 FIGURE 11-24. SCHEMATIC DIAGRAM OF LFC L.E. ICE PROTECTION SYSTEM A fine filter (nominally 0.8 micron) is required to prevent blockage of the porous panels by contaminants introduced with the fluid. A bypass and a differential pressure indicator are unnecessary. The pressure drop across the filter would vary widely due to the viscosity variation as a function of temperature. A spring-located check valve would minimize loss of fluid during maintenance operations or during flight with the pump OFF.

It is envisaged that the porous panels would provide an active region as shown in Figure II-25 and would be divided into individual panels about four feet long.

The system can be operated in either the de-icing or anti-icing mode. Generally, anti-icing requires about twice the flow rate needed to de-ice and may require a longer period of operation. For example, three de-icing cycles could be achieved in fifteen minutes compared to an estimated one hour capability required for an anti-icing system. Using these ground rules, the weight of fluid required for anti-icing is eight times that required for de-icing. To provide ice protection capability for two flights, a typical LFC large transport aircraft would require about 82 kg (180 Ib) of liquid based on a de-icing system and 653 kg (1440 Ib) based on an anti-icing system. A combined system that anti- ices the wing in frontof the engines and de-ices the remainder of the wing would minimize the liquid required to provide adequate protection. De-icing forward of the engine is not acceptable due to the possibility of large ice particle shed from the wing entering the intake.

11.6 LIQUID-FILM PRELIMINARY DESIGN AND PERFORMANCE EVALUATION Numerous studies have shown that once insects adhere to the wing surface, it is extremely difficult to remove them. The liquid-film concept prevents adhesion by coating the wing leading edge with a thin film of liquid. The liquid cushions the impact, prevents adhesion, and carries the insect debris aft along the wing and off the trailing edge.

Two methods were considered as a means of applying the liquid to the wing: (1) a porous distribution panel through which the liquid was forced, and

(z) spray nozzles mounted on the aft face of a Krueger-type shield and

directed toward the wing leading edge.

The term liquid-film will be used to denote the former concept. Performance and design of the spray system is discussed in Section ll°7.

11.6.1 Preliminary Design Studies The liquid-film system can provide integrated ice protection and contamination control. A freezing-point-depressant (FPD) that has good spreading properties and other properties compatible with aircraft use (non-toxic, low flammability, etc.) is required to protect against both insects and ice. Other properties such as viscosity and surface tension may be critical to ensure that insects are swept off the wing and that the liquid can be cleared from the small pores of the porous suction surface. The recommended liquid formulation is discussed in detail in Section 11.6.3.

The system schematic would be identical to that of the shield ice protection system shown in Figure II-24, except that the flow and liquid capacity would be increased. The description of system components and functional operation presented in Section ll.5 also applies. The unique features of the liquid- film system for the LFC aircraft are associated with (1) variable flow rates to meet the requirements of both ice protection and contamination avoidance

with reasonable quantities of liquids, (2) integration of the liquid-film

distribution panels with the LFC suction panels in the stagnation region, and

(3) selection of a FPDfluid compatible with an LFC aircraft using porous

suction surfaces.

Analyses of the trajectories of insects indicate that the upper surface dispenser

must protect against direct impingement for distances slightly less than I0

percent of the chord length whereas the lower surface dispenser must provide

protection to at least half the chord length.

A unique problem exists in the attachment line region when attempting to keep

the wing leading edge completely wetted. The liquid will stream away from

the attachment line and, therefore, a dry area will tend to form whenever

the attachment line lies in an area outside of the dispensers. For this

reason, liquid should be dispensed onto the entire region through which

the attachment line travels. On the LFC aircraft the region through which the

attachment line travels also requires suction. These conflicting requirements necessitate either (a) a system that can alternatively provide liquid dispensing or suction capability for the same surface area or (b) a system that meets the suction requirements in a series of narrow spanwise strips permitting liquid dispensers between the suction areas. All concepts that were developed to interconnect the suction and liquid systems were felt to be impractical from the reliability and maintainability points of view.

With separate liquid and suction systems, spanwise liquid distribution channels could be molded into the laminated glass substructure. A single chordwise liquid distribution channel could then be used to connect the inlet with the spanwise channels.

For a takeoff through a contaminated area, the liquid-film system that protects the attachment line region would be turned on prior to taxiing to the runway.

Gravity and wind would ensure a fairly complete initial coverage of this portion of the wing while the aircraft is in the static condition. During the ground roll the position of the attachment line will depend on flap position, wing incidence with respect to the fuselage, and the effect of the ground plane.

It is expected that some portions of the leading edge would start to dry, but during rotation all regions that have started to dry would be completely rewetted. Similar changes in angle of attack as gears and flaps are retracted would again rewet any dry areas. Attachment line travel during acceleration and oscillations of the attachment line during climbout would continue to rewet areas that started to dry. It is expected that a fluid that does not dry in twenty seconds could keep the attachment line region wetted.

Nigh flow dispensers would be used to form a liquid-film over the wing leading edge aft of the attachment line region using the highsinitial-flow/low-sustaining flow method developed during tests described in Section II.6.4. During the initial phase of the ground run, the leading edge would be completely wetted by the use of a high flow rate for a short time (less than ten seconds).

The surfaces would then be kept moist by a low sustaining flow. At a pre- selected time or altitude, possibly 1520 m (5000 feet), high flow would be used to wash off all debris over the trailing edge of the wing. The regulator supplying the high-flow system would require two pressure settings to provide the capability for the foregoing operation.

In order to ensure a smooth surface in the critical leading edge region, the entire surface would be of one piece construction from the upper surface at the front spar, around the leading edge to about 6.5 percent of a chord length aft of the leading edge on the lower surface. This subassembly would consist of the nose "D" duct with the suction and liquid-film ducting installed and the upper suction surface with spanwise metering ducts. The upper and nose regions would be attached to the main structure using internal fasteners. The remainder of the leading edge cavity surface would be provided by a lower panel assembly extending to the front spar. The nose section would contain the spanwise collector ducts in addition to the suction surface and metering ducts.

The lower panel could be removed for maintenance and to gain access to the internal fasteners attaching the upper panel. Internal panels in the "D" duct would provide access to the liquid-film system and nose ducting.

11.6.2 Experimental Evaluation of the Liquid-Film Concept Tests were conducted in the Douglas Low Speed Wind Tunnel to investigate four key aspects of the liquid-film concept: o Liquid Distribution: the ability of the liquid to completely cover the wing leading edge.

o Removal of Debris" the ability of the liquid to transport contaminants off the wing.

o Residual Film: the degree of surface roughening caused by the liquid residue and its effect on laminarization.

o Clogging and Clearing: the effect that the liquid has on the suction surface porosity and methods of clearing a porous surface if it becomes clogged with liquid.

The flat plate test model used previously for laminar flow evaluation of porous surface specimens in the wind tunnel was used for these tests. Two of the 3 m x 1.2 m (4 ftx 1 ft) specimens were modified by adding three 13 mm x 124 mm (5 in x 4.9 in) liquid dispensing plenums to the underside of the specimens. The two specimens were designated as follows: -517 50 x 250 Dynapore (a Dutch weave cloth made of 316L stainless steel wire with 50 wires per inch warp and 250 wires in the fill direction. Bonded by sintering and callendered to produce a smooth surface and obtain the desired resistance to airflow.)

-541 Perforated titanium with 0.102 mm (0.004 inch) diameter holes, spaced 1.02 mm (0.040 inch) on-centers, in a square pattern.)

11.6.2.1 Liquid Distribution - To ensure that the liquid-film, contamination- avoidance system will prevent insects from adhering to the wing leading edge, it is necessary to completely cover the region in which impingement occurs. Wind tunnel tests were conducted to explore potential problems and possible solutions in three areas: (I) the formation of rivulets, (2) the uniformity of distri- bution and edge effects, and (3) the avoidance of dry areas in the range of chordwise travel of the attachment line (i.e., the stagnation region for two dimensional flow).

The runback of de-icing liquids tends to form into rivulets in a short chord- wise distance depending on liquid characteristics and flow rate. To assure complete coverage of the wing leading edge, it would be necessary to reapply the liquid at a chordwise distance less than that in which rivulets would form.

One objective of the wind tunnel tests was therefore to determine,the distance (in the flow direction) over which the FPD liquid would provide complete coverage and to determine the effect of fluid flow rate on this distance.

To ensure uniform fluid distribution, the resistance to the fluid flow across the surface must be significantly greater than the fluid pressure drop in the spanwise distribution channel. A de-icing system of similar design (Reference ll-12) manufactured by TKS Corporation of England uses a porous sublayer of high pressure drop Provic sheet to ensure uniform spanwise distribution. Since the flow was interrupted at the end of each channel, the tests also evaluated the effect of this interruption on the ability to provide complete liquid coverage of the wing.

The final area of concern is the attachment line region. To ensure complete protection, the TKS de-icing system distributes liquid over the entire region of attachment line travel. This is done to prevent a dry area when the attachment line falls between two dispensing plenums. This ideal implementa- tion method may not be achievable on the LFC aircraft due to conflicting suction system and structural requirements. An acceptable compromise system is one that permits spaces between dispensing channels if it can be shown that the space will not dry out during the time period in which the attachment line falls into this region. One purpose of the test was to determine the time taken for the surface to dry out after being fully wetted.

Tests were conducted on the perforated titanium and Dynapore panels to evaluate the effect of liquid flow rate, tunnel dynamic pressure (q), and tunnel-to- suction plenum differential pressure, on the ability of a liquid to completely cover the test panel aft of the liquid distribution plenum. The liquid used during these tests was lO0 percent ethylene glycol which is a typical freezing point depressant (FPD) that possesses good wetting capabilities.

Table ll-2 summarizes data on the effect of differential pressure between the suction plenum and local ambient on the ability of the liquid to coat the wing surface. The data show that when the suction plenum was open to laboratory ambient the flow quickly broke into rivulets. This was because the differential pressure forced the liquid through the porous material instead of allowing it to flow aft. The data from the run in which the plenum-to-local-ambient pressure was equalized demonstrates improved coverage 178 mm (7 inches) aft of the dispenser.

Based on these result it was concluded that the suction plenum pressure should be equal to or greater than local ambient pressure to achieve the most efficient liquid distribution.

Tunnel dynamic pressure (q) was varied from 478.8 to 2394.0 Pa (lO to 50 psf) to determine the effect of q on the distribution of the liquid. For the same flow of liquid, a definite improvement in coverage was noted as q was increased.

A number of test runs were made to determine the liquid flow rate necessary to provide complete coverage. It was found that a high initial flow rate was required to wet the entire surface. The flow rate was then reduced incrementally and the effect on surface coverage was observed. Figure 11-26 presents repre- sentative results which indicate that a sustaining flow of 0.251 cm3/s for a 127 mm span (0.92 in3/min over 4 inches) was the minimum flow necessary to maintain surface wetness after an initial flow rate of about 1.092 cm3/sec (4 in3/min). The results also indicate that the flow necks to an extent dependent on the flow rate. To prevent dry areas, the plenums must be over- lapped and, if possible, the ends of the dispenser should be provided with a slightly higher flow rate than for the basic panel.

r'.," C) C.O I-- r,,-" r_ Z ._.1 o,J I-- ! Z Z L_ 0 r_ LO L.LJ I-- la- --J la_ r,,"', _0 I-- LL_ I--- F-- I, C_ .J ORIGINAL PAGE IS /" OF POOR QUALITY / WRP ACTIVE WIDTH ,._E,,-2_,, FIGURE 11-25. FIRST APPROXIMATION OF ACTIVE POROUS PANEL WIDTH REQUIRED ON SHIELD FOR ICE PROTECTION PANEL NO. 1 PANEL NO. 2 PANEL NO. 3 IN

L_

I IJ

l

I

,I

I

II

Jl31N.

i

I

q = 2,393 Pa (50 PSF) TEMPERATURE RANGE 28°C TO 38°C I i _,iN.

FLOW RATES

:l

I, 1.25 ¢m 3/s (4.57 IN.3/MIN) -----0.55 ¢m3/s (2.03 IN.3/MIN}

:i

I 10.25 cm3/s (0.92 IN.3/MIN)

II

I

!I

I 6 IN.

! ,

I

I

t ! I _71N" I

I

83-GE N-23| 63 FIGURE 11-26. FPD LIQUID FLOW TESTS IN LOW-SPEED WIND TUNNEL PERFORATED TITANIUM PANEL - ONE PLENUM During the tests, it was observed that the porous surface would stay wet after the glycol flow was discontinued, for periods of 20 seconds to one minute.

The time during which the surface would remain wet decreased when water was added to the glycol. The maximum time that flow can be disrupted will depend upon the viscosity and evaporation rate of the final fluid formulation.

11.6.2.2 Removal of Debris - The liquid-film concept of contamination avoidance requires that the liquid first prevents adhesion of the insect to the wing surface and second lubricates the surface so that insect remains or other contaminants are transported off the wing. Particles larger than 0.01016 cm (0.004 inch) in height have been shown to affect boundary layer transition and therefore the liquid-film system must prevent their adherence and remove them from the wing.

To evaluate the removal of contaminants, glass beads 0.15 mm (.006 inch) diameter were used to simulate debris. Initial tests indicated that aerodynamic forces would blow particles off of a dry wing but would not remove particles from a damp or wet surface. Additional runs showed that the application of a glycol at a high rate for a short time caused most of the particles to be transported off the 0.305 m x 1.22 m (I foot x 4 feet) panel but did not completely remove the particles from the 3.05 m (I0 foot) long test section. The ability of water to permit removal of particles from both the perforated titanium panel and aluminum afterbody was demonstrated. Based on observations during the test runs, it was evident that the high viscosity of the glycol was the predominant factor in causing particles to adhere to the surface. The evaporation rate appeared to have a minor influence. Three conclusions can be drawn from the results of testing: (a) Aerodynamic forces will not maintain the level of cleanliness required for an LFC aircraft.

(b) To ensure complete removal of contaminants from the wing, the fluid viscosity must be less than that of I00 percent ethylene glycol.

(c) A viscosity slightly greater than that of water is acceptable.

11.6.2.3 Residual Film - When the ethylene glycol evaporates from the surface, a film remains that could appear rough to the freestream. Tests to evaluate this residue were conducted. Visual inspection indicated that the film of glycol that remained on the perforated titanium was extremely smooth. The residue on the 50 x 250 Dynapore was slightly rougher and on smooth aluminum, beads of glycol formed. During clogging/clearing test runs on panel LFC-17 (80 x 700 Dynapore), visual inspection indicated that the residue was extremely smooth with the glycol diluted with 40 percent water or 30 percent water/lO percent alcohol.

Measurements of the chordwise distance to the transition point as a function of suction rate indicated that the residual glycol film had a negligible effect on the extent of laminar flow.

11.6.2.4 Clogging and Clearing - Whenever a liquid covers any portion ofthe porous material on the wing surface, the possibility exists that the liquid will lodge in the pores and reduce the surface porosity. This situation could be encountered (I) during flight or ground operations in rain, or (2) when the liquid-film con- tamination-avoidance or ice protection system is operated. Hence, the liquid could be either water or the contamination-avoidance liquid which will probably be some mixture of glycol, water, and additives to inhibit corrosion.

To ensure proper operation of the suction system during cruise, the liquid must either be prevented from entering the porous surface or must be cleared out of the pores before it freezes or the viscosity increases substantially. Con- ceptually, a system can be envisioned in which the suction manifolds are always pressurized slightly to prevent the liquid from entering the pores.

A more practical approach is to provide a high positive pressure in the suction system for a short duration after encountering the liquid to clear the porous surface.

The results of early tests in the wind tunnel indicated that liquid does

degrade the surface porosity severely,resulting in up to 82 percent reduction

in suction airflow. Attempts to clear the panel using the suction system

in a reasonable time were unsuccessful with suction levels as high as 958 Pa

(20 psf). Further exploratory tests confirmed that unreasonably high suction levels would be needed to clear trapped liquid from the porous material.

Tests also indicated that water was expelled far more easily from the porous surface than ethylene glycol which required a relatively high pressure. As a result of these findings, glycol mixtures were used for subsequent clogging/ clearing tests and pressure was used to expel the liquid.

Further testing was done in the wind tunnel to include the effect of aero- dynamic forces on clogging and clearing. Test panel LFC-17 (80 x 700 Dynapore) was used, which incorporated the most recent fabrication and design concepts.

The panel was not modified to add liquid distribution plenums. Instead, the liquid was applied externally upstream of the test section using a spray tube.

The liquids tested were 60 percent ethylene glycol/40 percent water, and 60 percent ethylene glycol/30 percent water/lO percent alcohol and suction plenum pressures ranged from 96 to 3352 Pa (2 to 70 psf). Figure 11-27 presents results in which a glycol mixture was cleared from the porous specimen in about fifteen minutes at a pressure slightly below 3352 Pa (70 psf). The pressure tapered off during the test because of reducing flow restriction. The unclogged porosity of the test panel was 0,61 m3/s/m 2 at 670 Pa (120 SCFM/ft 2 at 14 psfg), At this point it was decided that a more comprehensive investigation was needed. Two approaches were taken: (1) Specify a liquid formulation that best meets the requirements of the contamination-avoidance/ice protection functions and possesses the physical properties allowing it to be easily expelled from a porous surface. The results of this investigation are presented in Section 11.6.3.

(2)

Systematically evaluate clogging and clearing of porous panels to determine the influence of surface porosity, overall porosity, and liquid properties on the required clearing pressure and time, as follows: The basic wind tunnel setup previously described in Section 8.1.1 was modified to allow filtered plant air to pressurize the plenum under the porous test specimen and a 280 mm x 1190 mm (II inch x 47 inch) aluminum panel insert was modified to permit the installation of 152 mm x 305 mm (6 inch x 12 inch) porous panels. Candidate contamination-avoidance liquids were sprayed on the test section from a pressurized supply tank. The physical properties of these liquids are presented in Table 11-3 and Figure 11-28.

The evaluation of the various construction techniques and liquid formulations followed a standardized procedure. Each specimen was airflow tested in the "as-manufactured" condition to establish the overall porosity with no airflow over the surface. Next a calibration curve of airflow versus pressure drop was established at the test conditions, wind tunnel q = 479 Pa (I0 psf). This calibration curve was used as a reference to calculate the percent clogging due to the liquid. With the tunnel operating at a q of 479 Pa (I0 psf) and with a slight positive pressure of about 249 Pa (I inch of water) maintained in the plenum, the candidate liquid was sprayed upstream of the test specimen until the entire specimen was coated with liquid. The plenum pressure was then increased to the test value and time measured as the pressure passed through 2758 Pa (0.4 psig). A constant plenum pressure was maintained as the panel cleared. Flow versus time was recorded to indicate the effectiveness of the plenum pressure in clearing the liquid from the porous surface. Characteristics of the various porous surface specimens that were tested are presented in Table 11-4. Initial pressure drop versus flow measurements were made on each specimen. The value of the flow per unit surface area at a differential pressure of 670 Pa (14 psf or 2.69 inches of water) was used as a measure of the openness of the porous material and is termed "porosity".

9J _.- _ _._ L_ ^ Or-- ,p, E v_ ° ," 0 c_ ' _ X I--" b,'z,, G, I-- o'_ L,L C:n I C, .j o E --J I--- r',,, O o4 _1_ * _J X I'-'-- "'" r..m c_ ' _2 I--- C A OJ U e"'" u fO U L °_-,.

v 0_==.

>..

v P_ U U "_ In (",a __ 4_ v o Q U ._J ¢M e-- U E in, .,..- 4-) C 4-: ,, E Q '-- 0 U op,, I--- ",- -X E 0 >_ 4-) -P ul U '_1 N Q 43J q- L n= X _j t_ o,-- .'X OF POOR QUALITY NOTES: 1. WIND TUNNEL OPERATED AT q = 10 PSF = 1,915.2 Pa 2. PLENUM PRESSURE 4.2 PSF WHILE APPLYING LIQUID (4.2 PSF = 201.1 Pa) 3. LIQUID APPLIED EXTERNALLY (60% GLYCOL, 30% WATER, 10% ALCOHOL) 4. PLENUM PRESSURE INCREASED AS SHOWN.

Pa 4,000- 3_- 3,000- PLENUM POROSITY LB HR FT 2 10 2.500- TUNNEL LB/FT2 (PSF) 2,000- 1.500- i -TO- 1,000- 500- 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME (MINUTES) _GE. 23,,5 FIGURE 11-27.

CLOGGING AND CLEARING OF POROUS MATERIAL (SPECIMEN LFC 17 IN LOW-SPEED WIND TUNNEL)

C

4oo 2oo PROPYLENE GLYCOL METHYL ETHER 40% WATER 4o 60% ETHYLENE 2O VISCOSITY GLYCOL CENTISTOK ES 40% WATER 33% ETHYL ALCOHOL 67% WATER FREEZE POINT WATER I I i -20 0 20 40 60 IN) OF I I I I I I I 0 10 20 °C

O

FIGURE 11-28. VISCOSITY VERSUS TEMPERATURE FOR CANDIDATE LIQUIDS OF pOCi'_ QJA LI+Pt" r_. 0"_ • c-_ _ C_ I--- L_ •._ LL.

O_ ,.,_ t-,,--, Z I_.;J e,_ 0 0 c',J 0 0 _ 0 0 0 CD 0 C_ 0 0 0 0 0 _..

m!

I---

%'

L,U t--- Z c" (_ _._ _ _._ _-J .._ _ _ _ _ CZ: 0 _ rO r_ 0.; fO _ (_ t'_ --J r_ Z C',.J CxJ Od 0 0 0 0 0 (..9 Z I •,-- >._ .,-- .,-- >_ .,-.- .,-- :_, .,..- --J

C

ma I'-- I--- eY 0 0 0 r..-_ ' ,_.1 0 0 0 '_" _ _ _ _ 0 0 0 _ _ _0 0 0 0 ._J _--Od L_ or- LL_ Z Z t_ I'-- I--- ¢'_ OJ OJ _ _ P_ CO CO CO Z _d •"4 _ _ 0 0 0 0 0 0 E __.o

C

Prior testing had indicated that the primary force that prevented the liquid from being expelled from the porous material was due to capillary action.

It was not clear, however, to what extent viscous forces also retarded the clearing process. It is assumed that for a given liquid the capillary forces are controlled by the pore size at the surface whereas the viscous forces are controlled by the flow resistance across the entire specimen including both the Dynapore and the epoxy-filled glass. The porosity of the Dynapore provides an indication of the pore size at the surface.

Figures 11-29a and 29b present results from the clearing tests of the 152 mm x 305 mm (6 inch x 12 inch) specimens at two pressure levels. The results show that surface porosity has a definite influence on the ability to clear the liquid from the material whereas no definite tendency is seen as a function of overall porosity.

Figures ll-30 and ll-31 show the relationship between the pressure required to obtain an arbitrary level of clearing versus surface porosity and overall porosity, respectively. Although there is a great deal of scatter in the data, the results show that clearing pressure is affected by surface porosity. No such dependency is seen as a function of overall porosity (note the data point at .0508 m3/s/m 2 (lO SCFM per sq. ft.) in Figure ll-31. Assuming that capillary forces are the dominant factor, the clearing pressure would be inversely proportional to the surface pore diameter (which can be represented by the square root of the surface porosity). The line on Figure ll-30 which is drawn with a slope of -I/2 supports this premise. Based on these data, it appears that capillary forces (and hence liquid surface tension and surface porosity) dominate the clearing process and that viscosity has no consistent effect.

To further explore the effect of liquid properties on the required clearing pressure, various liquid formulations were tested. Typical results are shown in Figures 11-32 through 11-34, A dramatic reduction in the pressure required to clear the specimen is noted when the liquid is changed from aqueous C) Oti,_G_5 ..... " : ' LIQUID: 60% ETHYLENEGLYCOL/40% WATER PRESSURE: 12.4/kPa (1.8 PSIG) APPLIED FOR 2 MINUTES PERCENT 80 O OF ORIGINAL / OVERALL POROSITY (P) SCFM/FT l (DRY) POROSITY 70 [] AFTER 2 MINUTES 0 10< P(17 [] 24< P<40 [] 60 [] 0 IN)< P< loo AVERAGE 1 I 5O 0 (100) (200) (300) (SCFM/FT 2) I I t I 0 0.5 1.0 1.5 m3/s/m 2 L_GEN Z31_ SURFACE POROSITY FIGURE 11-29.A EFFECT OF POROSITY ON LIQUID CLEARANCE AT 12.4 kPa PRESSURE LIQUID: 60% ETHYLENE GLYCOL/40% WATER PRESSURE: 15.5 kPa (2.25 PSIG) APPLIED FOR 2 MINUTES PERCENT OF 80 ORIGINAL

B OVERALL POROSITY (P) SCFM/FT =

(DRY) POROSITY o 0 10<P<17 AFTER 2 MINUTES 70 [] 24<P<40 0 8o<e< 1oo AVERAGE I I 0 (100) (200) (300)(SCFM/FT 2) I I I I 0 0.5 1.0 1.5 m3/s/m2 SURFACE POROSITY "_"_'_'" FIGURE 11-29.B EFFECT OF POROSITY ON LIQUID CLEARANCE AT 15.5 kPa PRESSURE 6- BY 12-INCH SPECIMENS, 80 BY 700/80 BY 80 DYNAPORE INITIALLY (PSIG) CLOGGED WITH 60% ETHYLENE GLYCOL/40% WATER (4.5) (3.5 D (2.5 PRESSURE REQUIRED TO (1.5) 10 kPa OBTAIN 80% OF ORIGINAL POROSITY IN 2 MINUTES (1.0) LEGEND ® z_3 PLIES 194 [] 1 PLY 120 5 2 PLIES 120 (0.5) I I I I _ J I I ; (SCFM/FT AT (50) (100) (150) (200) (300) (5O0) 14 PSP) I l I I I I t I I .......

0 0.5 0.75 1.0 1.75 225 m/s/m AT 670 Pa DYNAPORE SURFACE POROSITY FIGURE 11-30. EFFECT OF SURFACE POROSITY ON LIQUID CLEARANCE

C

6- BY 12-1N. SPECIMENS. 80 BY 700180 BY 80 DYNAPORE INITIALLY CLOGGED WITH 60% ETHYLENE GLYCOL/40% WATER (PSIG (5.0) Pa N/m2 30,000- LEGEND (4.0) O 3 PLIES 194 (3 1 PLY 120 (3.0) 20,000" Q 2 PLIES 120 D (2.5) Q PRESSURE REQUIRED ® TO OBTAIN 80% ® OF ORIGINAL POROSITY Q (1.5) 1o,o00- IN TWO MINUTES (1.0) S_O0- mS/s/m2 OA 0.5 AT 670 Pa 0.1 0.2 0.3 (o.sL I I I I I I SCFM/FT = (20) (30) (40) (60) (1o (so) (go) 113GEN,2]147 AT 14 PSF OVERALL POROSITY

©

FIGURE 11-31. EFFECT OF OVERALL POROSITY :-)F F_-..;_.

/ ',.

SPECIMEN: 164 SCFM80 by700180 by80 DYNAPORE/1-PLY 120/IMPACTLAYER LIQUID: 60% ETHYLENE GLYCOL/40%WATER PERCENT OF ORIGINAL (DRY) POROSITY 6O CLEARING PRESSURE A 12.4 kPa (1.80 PSIG) 15.5 kPa (2.25 PSIG) i i 5 10 15 TIME (MINUTES) NOTE: EACH 6- BY 12-1N. SPECIMEN WAS SATURATED WITH LIQUID IN THE LOW-SPEED WIND TUNNEL AT A q OF 479 Pa (10 ps/!. POSITIVE PRESSURE WAS APPLIED AND THE TIME TO REACH A PRESSURE OF 3.1 kPa (0.45 PSIG} WAS MEASURED FIGURE 11-32. TIME TAKEN TO CLEAR DILUTED ETHYLENE GLYCOL SPECIMEN: 164 SCFM80 by700/80 by80 DYNAPORE/1.PLY 120/IMPACTLAYER LIQUID: 33% DENATURED ETHYLALCOHOL/67%WATER IO0 NpaGPRESSURE (1.1 PSIG) 8O Pa (0.9 PSIG) PERCENT 60 OF ORIGINAL (DRY) 4O POROSITY

20 /

J = | ' !

00 1 2 3 4 5 TIME (MINUTES) NOTE: EACH 6- BY 12-1N, SPECIMEN WAS SATURATED WITH LIQUID IN THE WIND TUNNEL AT Aq OF 479 Pie (10 PSF). POSITIVE PRESSURE WAS APPLIED AND THE TIME TAKEN S.tGENZ3Zg'_ TO REACH A PRESSURE OF 3.1 kP= (0.45 PSIG) WAS MEASURED TIME TAKEN TO CLEAR DILUTED ETHYL ALCOHOL FIGURE 11-33.

ethylene glycol to an alcohol or propylene glycol methyl either (PGME) mixture. Figure II-34 also shows that a porous surface that meets the porosity and smoothness requirements of the LFC aircraft can be cleared after being saturated with liquid by applying 6.894 KPa (one psi) of internal pressure for one minute.

Table 11-5 summarizes the pertinent liquid properties and required clearing pressures. A relationship is apparent between clearing pressure and surface tension. No such relationship is apparent between clearing pressure and viscosity.

To further investigate the ability of positive pressure to clear the surface of contamination avoidance/ice protection liquid, combinations of three fluids on three porous surfaces were tested. Porosity was measured as a function of time and pressure for each combination and the qualitative pattern of fluid removal was determined visually. The test results provide an indication of which materials and liquids would have the best clearing properties in flight. Specimens measuring 127 mm x 254 mm (5 inches x I0 inches) were fabricated from the following materials: Materials Specimen -503 80 x 700 Dynapore plus perforated stainless steel sheet - 509 80 x 700/80 x 80 Dynapore plus perforated stainless steel sheet.

-515 Titanium perforated with 0.102 mm x 0.203 mm (.004 inch x .008 inch) holes.

The liquids tested were 60 percent ethylene glycol - 40 percent water; 57 percent propylene glycol - 43 percent water; and 60 percent propylene glycol methyl ether (PGME) - 40 percent water.

Clearing pressure was supplied through a plenum directly beneath the specimen and an airstream with a dynamic pressure of 220 Pa at approximately 23°C was directed over the specimen parallel to its surface.

TABLE 11-5 PROPERTIES OF TEST LIQUIDS 33% Denatured 60% Ethylene 60% Propylene Glycol/40% Glycol Methyl Ethyl Alcohol/ Water 67% Water Ether/40% Water 52 34.5 36,7 Surface Tension (dynes/cm) 5.6 2.75 4.4 Viscosity (cs) (1.8) (0.67) (0.54) Clearing Pressure* (psi) kPa 12.4 4.62 3.72

C

*Pressure required to obtain 80% of the initial (dry) porosity in two minutes.

The test results are summarized in Figures 11-35 and 11-36, which show the time to reach 98 percent porosity versus pressure. The porosity changed very slowly with time above 98 percent porosity, making determination of time to I00 percent porosity largely arbitrary. Approximate times to I00 percent porosity were 5 to 15 seconds longer than the time to 98 percent porosity for all materials with PGME; ethylene glycol and propylene glycol were not I00 percent cleared from any specimen in less than 2 minutes for the pressure ranges of 0 to 14,000 Pa for -503 and -509 and 0 - I0,000 Pa for -515. No data are shown for ethylene glycol or propylene glycol on the -503 specimen or for propylene glycol on the -509 specimen because 98 percent clearing did not occur within two minutes with any of these combinations for pressures up to 20,000 Pa.

The results indicate the -515 specimen and PGME are the material and liquid combination with the best clearing characteristics. This is consistent with observation of liquid removal from the surface. Most of the liquid was removed by run back along the surface. The Dynapore specimens, however, retained liquid in their surface structures, inhibiting complete run back.

Evaporation accompanied final clearance of the PGME from the Dynapore specimens, but the ethylene glycol and propylene glycol did not evaporate.

Because of the rapid drop in vapor pressure as temperature is lowered, final clearing of PGME by evaporation cannot be relied upon in flight conditions.

Run back will be the primary mechanism for liquid removal. A smooth surface, such as perforated titanium, and a liquid possessing low surface tension and low viscosity, such as PGME, are therefore recommended as having the best clearing characteristics. The optimum clearing pressure can be determined in a flight test program, where the low temperature effects of higher viscosity and higher surface tension and the effects of a higher airstream velocity will be taken into account.

SPECIMEN: 164 SCFM 80 by 700/80 by 80 DYNAPORE/1-PLY 120/IMPACT LAYER LIQUID: 60% PROPYLENE GLYCOL METHYL ETHER/40% WATER 10080 '_p RESSURE P,_,-, -,. C j.:';_:'r{ OF .... ": I I I I ./i.--- 6.2_Po(o._ PsiG) [ I I L_L/--- 5.s2 kP,io.8Ps,G) I I.L..'_...-_ 4.e3 kP, (o.7 ,,siGl

PERCENT

OF ORIGINAL (DRY) POROSITY 40 2O 0 1 2 3 4 5 TIME (MINUTES) NOTE: EACH 6- BY 12-IN. SPECIMEN WAS SAt URATED WITH LIQUID IN THE LOW-SPEED WIND TUNNEL AT A q OF 479 Pa (10 PSIG). POSITIVE PRESSURE WAS APPLIED AND THE TIME TAKEN TO REACH A PRESSURE OF 3.1 kPa (0.45 PSIG} WAS MEASURED.

FIGURE 11-34. TIME TO CLEAR DILUTED PGME o PGME--515 IO0 + PGME--503 L_ PROPYLENE GLYCOL --515 8O + TIME TO 98% 6O POROSITY (SEC) +4"d.+ 2O I I 0 I I I I I 1 I 16 18 20 0 2 4 6 8 10 12 14 PRESSURE (kPa) 83_..EN-23152 LIQUID CLEARANCE TIME VERSUS PRESSURE FIGURE 11-35.

r 11.6.3 Liquid Formulation The liquid used for contamination-avoidance and ice protection must meet many requirements to ensure proper system operation and compatibility with the overall aircraft operation. Some of these requirements have not been fully defined at this time.

(a) Freezing Point - Unless all of the contamination avoidance liquid in the wing tubing is purged after each application, the liquid in the unheated wing will be cooled by the ambient temperature which at cruise altitudes can be as low as -70°C. Allowing for a 25°C ram temperature rise, an aircraft skin temperature of -45°C can be expected. The ice protection system will be required to operate following a descent into icing conditions after being cold soaked during cruise to this temperature. During descent, the tubing and liquid in the wing will warm up somewhat. The system may also be required to operate during cruise to eliminate small amounts of ice that could disrupt laminar flow. It is therefore a design goal that the ice protection liquid have a freezing point of -45°C or less to permit operation during cruise.

(b)

Surface Tension and Wettability - These two interrelated properties of the liquid affect (I) the ability of the liquid to form a film over the entire wing surface without breaking up into rivulets, and (2) the pressure required to clear a porous surface that has been clogged with liquid. To enhance its wettability, the liquid should possess a low surface tension and low contact angle (e) between the liquid and metal.

The pressure required to clear a clogged porous surface is related to the pressure exerted by capillary action which is proportional to the product of the surface tension and the cos e. A liquid with a good wettability (i.e., low e) will be difficult to clear from a porous .r surface unless it also has a low surface tension. Since good wettability is essential, the selection of a liquid with the lowest possible surface tension is necessary to minimize the pressure required to clear the liquid out of the porous material.

During operation in rain or icing conditions the liquid may be diluted with water, therefore, it is required that the product of surface tension and cos o be minimized for all aqueous solutions of the contamination-avoidance/ice protection liquid formulation.

(c) Viscosity - The liquid viscosity and its variation with temperature

affects three areas of operation. First, the liquid viscosity must be low enough to allow the transport of insect remains in a chordwise direction. The viscosity must also be low enough to prevent the formulation of globules on the surface that would also create turbulence. A final concern is the variation of viscosity with temperature. The liquid pressure drop through the surface must be sufficient to ensure uniform spanwise distribution. Over the temperature range from -40°C to +50°C, the viscosity of typical glycols changes by a factor of lO0. These viscosity variations will be reflected in corresponding changes in liquid flow and/or pressure drop across the suction surface. Over the range of ambient tempera- tures in which insects are active it is desirable to maintain a constant flow. This would require a pressure across the suction surface of at least five times the minimum operating pressure. For adequate ice protection the liquid flow required at -40°C is less than that required at O°C so that the pressure increase for ice protection need not offset the full viscosity change.

(d) Volatility - Evaporation rate will also influence the ability of the liquid to coat the surface and to permit insect remains to be trans- ported off the wing. The most critical region that is influenced by volatility is the area of attachment line travel. As discussed in Section 11.6.2.1, dry areas may form if the liquid film is removed by the airstream in less than about twenty seconds.

(e)

Miscellaneous Liquid properties - The liquid must possess properties that are generally required for fluids used on aircraft. For instance, the liquid should be non-corrosive and have a reasonable boiling point.

Two properties that will require special consideration are flammability and toxicity. Since the liquid is applied external to the wing, it must not create a fire hazard. Both a reasonable flash point and low concentration of the vapors in air are required to avoid a fire hazard. Low toxicity is needed to minimize the environmental impact due to the liquid-film system.

(f) Summarizing the foregoing discussion, the physical property require- C? ....

ments of the liquid to be used for contamination-avoidance/ice protection are as follows- I. Freezing point < - 45°C 2. Surface tension < 40 dynes/cm 3. Good wettability 4. Viscosity: The exact limits have not been defined. In general, the viscosity should be low enough to prevent the formation of globules as the liquid runs back along the surface and the ratio of viscosity at cold temperatures to that at maximum temperature should be minimized.

5. Low volatility 6. Low toxicity 7. High flast point 8. Low corrosivity.

Candidate Liquids - A number of candidate liquids were evaluated

(g)

to find a suitable formulation. Alcohols,including glycol,form the basis for the common de-icing agents and were therefore considered first. Five glycols were evaluated: ethylene, diethylene, triethylene, tetraethylene and propylene. A 70 _+ lO percent glycol/ 30 + lO percent water solution is required to satisfy the freezing point requirement. Aqueous propylene glycol has the lowest surface tension: 40 dynes/cm at 70°F compared to about 52 dynes/cm at 77°F for either ethylene or diethylene glycol. However, propylene glycol is too viscous at temperatures below-lO°C.

Some of the glycol ethers offer improvement in both viscosity and surface tension characteristics over the aqueous glycol solutions. Over thirty formulations were screened. Some of the more promising are listed in Table 11-6.

TABLE 11-6 CANDIDATE GLYCOL ETHERS Freezing Viscosity Surface Flash Manufacturer Point at 20oc Tension Point Name oc CP dynes/cm °C (OF) Union Carbide -70 6.4 27 60 (140) Butyl CELLOSOLVE Union Carbide -8_ 3.9 40 147 (188) Methyl CARBITOL Union Carbide -68 6.5 34 I01 (214) Butyl CARBITOL Union Carbide -80 2.8 26 48 (119) PROPASOL Solvent P Union Carbide -95* 1.9 29 33 (91) PROPASOL Solvent M Union Carbide -80* 5.2 33 75 (167) PROPASOL Solvent DM Dow -I00" 1.8 28 43 (109) DOWANOL EE Dow -97* 1.7 28 36 (96) DOWANOL PM Dow -82* 3.4 29 78 (175) DOWANOL DPM *Pour Point Both ethyl and methyl alcohol meet the freezing point, surface tension, and viscosity requirements but have low flash points. Silicate ester based fluids such as COOLANOL (manufactured by Monsanto) and fluoronated hydrocarbons

©

I _ii ¸ also possess good physical properties but are expected to have an unacceptable impact on the environment. Typical properties are listed in Table 11-7.

TABLE 11-7. CANDIDATE ALCOHOLS, FLUOROCARBONS AND HEAT TRANSFER FLUIDS Freezing Viscosity Surface Flash Point at 20oc Tension Point Name Manufacturer (oc) (cp) (dynes/cm) oc (OF) Ethyl Alcohol -- -117 1.2 22.8 Low Methyl Alcohol -- - 98 0.6 22.6 Low "Freon E2" Dupont -123" I.I 12.9 High "Freon E3" Dupont -190" 2.2 14.2 High COOLANOL 15 Monsanto -140" 2.0 21.0 77 (170) COOLANOL 25 Monsanto -120" 5.4 25.0 163 (323) *Pour Point One of the most suitable liquids for insect/ice protection is propylene glycol methyl ether (PGME), the basic constituent of DOWANOL PM and PROPASOL Solvent Mo The major concern in the use of this fluid is its low flash point of 35°C (95°F). The use of an aqueous solution of 60 percent PGME/40 percent water raises the flash point to 52°C (125OF). Considering the extremely low concen- trations that will be present after mixing with the free stream, the flash point of 52°C (125°F) is considered to be acceptable. This liquid formulation is recommended and has been used as the basis of additional investigations.

II .6.4 Liquid-Film System Test and Analysis Summary (a) System Performance - The results of the tests conducted in the Douglas Low Speed Wind Tunnel indicate that complete coverage of the wing leading edge surface can be achieved for a downstream chordwise distance of at least 254 mm (I0 in.) by applying 9.9 cm3/s/m (I0 cu in/min per foot) of span. While applying the liquid, the suction plenums should be pressurized to at least local ambient pressure to prevent the liquid

G

from being forced through the porous surface. To avoid irregular

coverage, the ends of the distribution plenums should be overlapped.

The pressure drop through the surface must be sufficiently greater than the spanwise pressure drop in the panel to obtain uniform distri- bution. During the wind tunnel tests glycol kept the surface moist in excess of twenty seconds after liquid application.

The wind tunnel tests indicated that a liquid with a viscosity as great as that of lO0 percent ethylene glycol will not transport contaminants off the trailing edge of the wing, whereas water did keep the wing clean. This was confirmed by the JetStar tests program (Reference II-26) in which a water spray was effective in preventing adhesion of insects on the wing and transporting the debris off the wing. The recommended solution of 60 percent propylene glycol methyl ether/40 percent water has a viscosity about I/5 that of lO0 percent ethylene glycol.

Tests also indicated that the residue from the ethylene glycol after evaporation did not significantly affect transition of the boundary layer.

Based on the results of Douglas wind tunnel tests and NASA Jetstar flight tests, it is concluded that a liquid-film system meets all of the system performance criteria. One concern that will be evaluated during subsequent flight testing is the limits of operation during low ambient temperature operation, especially considering the increase in viscosity at low temperature.

(b) Compatibility: Tests were conducted to evaluate the compatibility

of the liquid and the porous surface. The major concern was that the liquid would be held in the pores due to capillary action and would clog the porous surface when suction would be required. Test results indicated that both water and glycol substantially reduced the porosity of Dynapore materials. LFC levels of suction failed to restore the porosity in a time compatible with flight operations. Water was ::.

cleared from the pores by a very low positive pressure; however, a positive pressure of one psig was required to clear the propylene glycol methyl ether mixture from the pores. Based on these test results, it appears that a liquid-film insect protection system can be made compatible with a porous surface suction system by the following procedure. Following any application of a liquid to the porous surface (either by operation of the insect/ice protection system or by flying through rain), the suction system will be pressurized to a level of about 7 kPa (I.0 psig) for about 1 minute. The exact values will be determined after the selection of the final suction surface configuration and the liquid formulation.

Design studies also showed that the liquid-film system can be made compatible with the suction system, ice protection system, structure, and space restraints in the leading edge area. The choice of the low toxicity fluid propylene glycol methyl ether should make the system compatible with Environmental Protection Agency (EPA) requirements.

The liquid-film system is not expected to significantly protect the C outer surface against erosion due to sand, dust, hail or rain and normal structural design requirements would apply to these conditions.

Based on the results of studies to date, it is concluded that the liquid-film system can be made compatible with other LFC system requirements. Some concern remains regarding the flammability of the liquid. However, as previously stated, the low concentrations of the fluid in the air make the probability of fire extremely remote. Caution in handling, ventilation design, and isolation from ignition sources should be stressed.

11.6.5 Liquid-Film System Design A central supply panel is envisioned which will include storage, pumping, filtering and pressure/flow control functions. The storage tank would have a capacity of about 0.42 m3 (II0 gallons) and would contain about 390 kg 860 pounds) of usable fluid. This quantity of fluid would provide contamination protection for the entire wing during takeoff and landing as required below 1524 m (5000 feet) altitude for two flights. It would also provide three de-icing applications per flight for the wing and ice protection for the shield.

f- To avoid clogging the porous surface, the filters must be effective down to the l to 2-micron range. A flow control valve will provide constant flow whenever the ambient temperature is above about O°C. Below this ambient temperature, flow will be controlled by the system pressure drop and will decrease with decreasing temperature. This control scheme will provide adequate fluid to protect the wing leading edge from contamination without subjecting the system to excessively high pressures during cold ambient temperature operation.

A breakdown of the weights estimated for the portion of the system within the fuselage of a 300 Passenger aircraft is listed in Table ll-8.

TABLE 11-8. LIQUID SYSTEM WEIGHTS Kg (Lb) Fluid Weight Shield Ice Protection: 82 (180) 82 (180) Attachment Line Region: Hi Flow, Upper Surface: 2 26 ( 500 ) K 9 (Lb) Total Fluid 390 ( 860 ) 20 (44) Tankage Pump, Motor, Filter, Regulator 20 ( 44 ) 430 (948) Total System Weight Fabrication, Operation and Maintenance - The liquid-film system employs concepts that have been used for ice protection of commercial and military aircraft for over 30 years. The only area of concern is the integration of the liquid and suction plenums in the wing leading edge area. Further detail design work is necessary to demonstrate that an easily fabricated design can be devised.

The contamination-avoidance or ice protection system should pose no

operational restrictions. Flight testing is required to determine whether

laminarization can be achieved after operation of the liquid-film system

at low ambient temperatures. Operation of the contamination-avoidance

system can be made completely automatic except for an ON/OFFswitch using

inputs of altitude, static air temperature (SAT), airspeed, and possibly

flap position. Operating procedures for the ice protection system will be

similar to those used on current aircraft.

Maintenance of both liquid-film systems will be similar to that of current

aircraft applications. Care must be exercised in the design phase to reduce

the possibility of leaks. Access in the wing leading edge for inspection and

maintenance is improved with the upper-surface-only LFC concept. In either

case, the plenums could be integral with the surface panel. Excessive leakage

would require panel replacement and overhaul. It may be possible to design

a removable insert which would contain the ice protection plenums. Further

effort is required in this area to develop a practical design.

11.7 SPRAY SYSTEM PRELIMINARY DESIGN AND PERFORMANCE EVALUATION

CJ

An alternative method ofapplying liquid over the surface of the wing leading edge is the use of a spray system. Used in conjunction with a shield, two methods of implementing the spray concept were considered. Figure 11-37 illustrates a "shield spray" in which the spray nozzle is mounted on the aft face of the shield and directed toward the wing leading edge. Figure 11-38 illustrates a "wing spray" in which a pop-out nozzle is mounted in the turbulent lower surface of the wing leading edge and the spray is directed forward and slightly upward. Aerodynamic forces cause the liquid to be distributed across the leading edge. The shield spray is utilized when the shield is deployed while the wing spray can be used when the shield is retracted. A schematic diagram of a liquid supply system suitable for either spray nozzle arrangement is shown in Figure 11-39.

During take-off through a contaminated region or in icing conditions, the shield provides the primary contamination avoidance function. Earlier studies indicated that a supplemental system may be required near the wing tip at minimum angle of attack and for the inner wing if a thick inboard wi_ng section i is used. This augmentation of the shield can be provided by operating the shield spray system as required during take-off and climb or descent. Normally _f L OF _:C'_i_?i '-; .....

S 100 O ETHYLENE GLYCOL --515

f

× PGME--509 [3 ETHYLENE GLYCOL --509 TIME TO 98% POROSITY (SEC) 0 I I 0 20 4 8 12 16 PRESSURE (kPa) 83GEN 23153 FIGURE 11-36. LIQUID CLEARANCE TIME VERSUS PRESSURE FIGURE 11-37. SHIELD SPRAY CONCEPT C _ ORIGINAL PAGE _-3 OF POOR QUALITY POP-OUT NOZZLE FIGURE 11-38. RETRACTABLE NOZZLE CONCEPT AIR PRESSURE l FILL/REI'IEF | FLUID /SUeeLY |TANK ("_ FLUID | _'_ QUANTITY ._.__...J. INDICATOR SUPPLY PUMP REGULATOR/ SHUTOFF VALVE (_ PRESS WARNING SWITCH

-,...°. ,o

-]F,L.. 1

SHIELD J( I SHIELD

OUTBOA"_°o 1

FUSE_NG SPRAY NOZZLE ASSEMBLY FIGURE 11-39. SCHEMATIC DIAGRAM OF LFC LEADING-EDGE CONTAMINATION-AVOIDANCE SYSTEM (SHIELD SPRAY CONCEPT) the shield spray would be shut off and the shield retracted at about 1524m (5,000 feet) altitude. When icing conditions are anticipated between 1524m (5,000 feet) and the cruise altitude, there are two options: (a) To keep the shield extended through the icing condition and use the shield spray to de-ice the wing if required.

(b) To retract the shield and use the wing spray system to de-ice after reaching initial cruise aItitude.

Light icing conditions could be encountered on rare occasions at cruise altitudes. This would be insignificant for a turbulent aircraft but LFC could be lost during and after such an encounter unless the ice cap were removed. Either spray system could be used during these encounters.

Freezing point depressants have been used in numerous ice protection appli- cations generally operating in an anti-icing mode. Liquid sprays have been used to remove ice and snow on the ground, however, the liquid was heated and used at higher flow rates than would be practical for an airborne application, To the best of our knowledge, no data are available to indicate the effective- ness of a freezing point depressant liquid spray for de-icing. Icing tunnel and/or icing flight tests will be required to determine de-icing capability.

11.7.1 Shield Spray System Evaluation The evaluation of the shield spray system feasibility include_: (a) The effect of the high velocity airstream between the shield and wing on the droplet trajectory.

(b) The selection of a nozzle to provide the desired distribution of droplet size and velocity.

(c) Droplet breakup due to high relative air-to-droplet velocities.

11.7.1.1 Liquid Droplet Trajectory - Douglas computer program H9PB, "Water Impingement on Two-Dimensional, Multi-Element Airfoils" has the capability of analyzing the trajectory of a droplet in the flow-field _nduced by an airfoil composed of up to ten elements. This program was originally written as a wing ice protection design tool and considers a small, supercooled droplet starting at about ten chord lengths forward of the wing. The computer program calculates the droplet trajectory and determines the catch on the wing. The program was modified so that an arbitrary starting location, velocity and direction could be input for a single droplet,

The amount of droplet deviation from a straight line would be expected to be a

function of:

(a)

Inertial forces (droplet mass and initial velocity).

(b)

Aerodynamic forces (aircraft altitude, airspeed, angle of attack, flap angle, etc.).

(c)

Trajectory (initial droplet angle and chord length).

Parametric analyses were conducted to obtain a first-order indication of the significance of the various parameters and to obtain preliminary design values. An existing airfoil model which is representative of the wing out- board of 40% span was used for this portion of the study. The range of variation of the parameters is tabulated below: Altitude: Sea level to 4752m (15,000 feet) Airspeed: 118 to 165 m/s (230 to 320 KIAS) 0 ° to 10 o Angle of Attack (a): 0 o Flap Angle: Chord Length (C): 1778mm (70 inches) Initial Droplet Velocity (Vd): 15 to 122m/s (50 to 400 fps) 25 o to 35 o Initial Droplet Angle* (¢): "0 is defined as the angle between the nozzle center line and wing reference plane.

The droplet diameter was varied from 2,000 microns down to the minimum size that would impinge on the wing leading edge.

A large number of trajectories were calculated to find the limiting drop size that is tangent to the wing upper surface for each set of conditions.

Figure 11-40 is a plot of these results for a specific nozzle orientation.

A value of ¢ = 30 o was selected to minimize the loss of liquid resulting from small drops being blown above the wing and large drops impinging too low so that they flow over the wing lower surface, Figure 11-40 shows that the shield spray will be most efficient at : : 0 o and least efficient at m = I0. Outboard portions of the wing may require the spray for insect protection at low angles of attack (while the aircraft is on the runway C_ during take-off and landing). The thick inboard section may require spray at both high and low angles of attack. The center section of the wing probably will not require liquid for insect protection. To obtain a conser- 1,000 SPEED 123 m/s (239 KIAS) 800 • \ ALTITUDE 4,572 m (15,000 FT) -t

t \ c.oRD l_m(70,N.)