CFD Open Series

Patch 2.55

Edited by: Ideen Sadrehaghighi

Pictures provided by (E. N. Tinoco)

ANNAPOLIS, MD
 2

Contents
List of Figures: ..................................................................................................................................................................... 6

1 Aircraft Conceptual Design Practices ................................................................................ 10

1.1 Direct Design ......................................................................................................................................................... 10
1.1.1 Multi-Objective Optimization ............................................................................................................ 10
1.2 Aircraft Conceptual Design and Multidisciplinary Optimization (MDO) as Envisioned
by D. P. Raymer................................................................................................................................................................. 11
1.2.1 An Overview of Multidisciplinary Optimization (MDO) ......................................................... 12
1.2.1.1 Finite Difference.......................................................................................................................... 13
1.2.1.2 Implicit Function Theorem ..................................................................................................... 14
1.2.1.3 Stepping Searches ...................................................................................................................... 14
1.2.1.4 Response Surface ........................................................................................................................ 15
1.2.1.5 Steepest Descent (Non-Gradient Stepping Search) ...................................................... 16
1.2.1.6 Monte Carlo ................................................................................................................................... 17
1.2.1.7 Random Walk ............................................................................................................................... 17
1.2.1.8 Simulated Annealing ................................................................................................................. 17
1.2.1.9 Evolutionary Algorithm ........................................................................................................... 18
1.2.1.10 Genetic Algorithm....................................................................................................................... 19
1.2.1.10.1 Selection (to Breed) ............................................................................................................. 19
1.2.1.10.2 Crossover ................................................................................................................................. 20
1.2.1.10.3 Mutation ................................................................................................................................... 20
1.2.1.11 Decomposition ............................................................................................................................. 21
1.2.1.12 Stopping Criteria ......................................................................................................................... 22
1.2.1.13 Design Variables, Automated Redesign Procedures, and Geometric
Constraints 22
1.2.1.13.1 The Basic Six (or Five) Design Variables .................................................................... 23
1.2.1.13.2 Fuselage Fineness Ratio..................................................................................................... 24
1.2.1.13.3 Design Lift Coefficient (Wing) ......................................................................................... 25
1.2.1.13.4 Geometric Design Constraints......................................................................................... 25
1.2.1.13.5 Fuselage Length and Diameter ....................................................................................... 25
1.2.1.13.6 Wingspan ................................................................................................................................. 26
1.2.1.13.7 Wing Geometry for Pitch up Avoidance ...................................................................... 26
1.2.1.13.8 Net Design Volume............................................................................................................... 26
1.2.1.13.9 Definition of Net Design Volume .................................................................................... 27
1.2.1.13.10 Automated Redesign for Discrete Variables ........................................................... 29
1.2.1.14 References ..................................................................................................................................... 30
1.3 Inverse Design ...................................................................................................................................................... 33
1.4 Geometry Consideration .................................................................................................................................. 34
1.5 Design of the External Geometry .................................................................................................................. 34
1.5.1 What is the Proper Design (Target) Pressure Distribution?................................................. 35
1.5.2 Pressure Distributions Over Aircraft Fuselage........................................................................... 36
1.6 Components Not Required to Generate Aerodynamic Forces.......................................................... 37
1.7 Components of Generating Aerodynamic Forces .................................................................................. 39
1.7.1 Pressure Distributions on Wings ..................................................................................................... 39

2 Direct Design Case Studies for Conceptual Aircraft ...................................................... 41

 3

2.1 Case Study 1 - Aerodynamics of High-Subsonic Blended-Wing-Body (BWB)

Configurations .................................................................................................................................................................. 41
2.1.1 Abstract ....................................................................................................................................................... 41
2.1.2 Introduction .............................................................................................................................................. 41
2.1.3 Tools ............................................................................................................................................................. 42
2.1.3.1 Navier-Stokes Computational Fluid Dynamics (CFD) and Inverse Design ......... 42
2.1.3.2 Multidisciplinary Design Optimization (MDO)............................................................... 43
2.1.4 Approach .................................................................................................................................................... 44
2.1.4.1 Initial CFD-Based Aerodynamic Design ............................................................................ 44
2.1.4.2 Drag Calibration .......................................................................................................................... 47
2.1.4.3 MDO Study ..................................................................................................................................... 48
2.1.4.4 CFD Aerodynamic Refinement .............................................................................................. 49
2.1.5 Results ......................................................................................................................................................... 51
2.1.6 Conclusions ............................................................................................................................................... 51
2.1.7 References.................................................................................................................................................. 52
2.2 Case Study 2 - Changing the Role of CFD in Aircraft Development ................................................ 53
2.2.1 Abstract ....................................................................................................................................................... 53
2.2.2 Introduction .............................................................................................................................................. 53
2.2.3 The Changing World .............................................................................................................................. 55
2.2.4 CFD use in Industry ................................................................................................................................ 56
2.2.5 Getting CFD Into the Product ............................................................................................................. 58
2.2.6 Cruise Wing Design ................................................................................................................................ 59
2.2.7 Iterative/Inverse Design...................................................................................................................... 60
2.2.8 Multipoint Optimization ...................................................................................................................... 61
2.2.9 Engine Airframe/Integration ............................................................................................................. 64
2.2.10 Future Challenges ................................................................................................................................... 66
2.2.11 High Lift ...................................................................................................................................................... 67
2.2.12 Buffet ............................................................................................................................................................ 68
2.2.13 Stability and Control .............................................................................................................................. 69
2.2.14 Conclusions ............................................................................................................................................... 70
2.2.15 Acknowledgment .................................................................................................................................... 70
2.2.16 References.................................................................................................................................................. 70
2.3 Case Study 3 - Fuselage Aerodynamic Drag Prediction....................................................................... 72
2.3.1 Abstract ....................................................................................................................................................... 72
2.3.2 Introduction .............................................................................................................................................. 72
2.3.3 Fuselage Geometries ............................................................................................................................. 73
2.3.3.1 Wing-Fuselage Faring ............................................................................................................... 74
2.3.4 Mesh and Physics Set-Up ..................................................................................................................... 75
2.3.5 Drag Prediction Method ....................................................................................................................... 75
2.3.6 Fuselage Nose Factor............................................................................................................................. 76
2.3.7 Fuselage Cabin Factor ........................................................................................................................... 77
2.3.7.1 Fuselage Tail Cone Factor ....................................................................................................... 78
2.3.7.2 Wing-Fuselage Fairing Effects............................................................................................... 78
2.3.7.3 Method Application ................................................................................................................... 80
2.3.8 Conclusions ............................................................................................................................................... 81
2.3.9 Reference.................................................................................................................................................... 81
2.4 Case 4 - Conceptual Design of A Business Jet Aircraft ......................................................................... 83
2.4.1 Abstract ....................................................................................................................................................... 83
2.4.2 Introduction .............................................................................................................................................. 83
 4

2.4.3 Design Requirements ............................................................................................................................ 84

2.4.4 Initial Layout Analysis .......................................................................................................................... 84
2.4.4.1 Concept Sketch ............................................................................................................................ 84
2.4.4.2 Initial Sizing .................................................................................................................................. 84
2.4.4.3 Sizing from the Conceptual Sketch ...................................................................................... 84
2.4.4.4 Mission Profile ............................................................................................................................. 85
2.4.4.5 Engine Selection:......................................................................................................................... 85
2.4.5 Aerodynamics Selection ....................................................................................................................... 86
2.4.5.1 Aerofoil Selection........................................................................................................................ 86
2.4.5.2 Wing Geometries ........................................................................................................................ 86
2.4.5.3 Flap Selection ............................................................................................................................... 86
2.4.5.4 Stall Characteristics ................................................................................................................... 87
2.4.5.5 Tail Geometries ........................................................................................................................... 87
2.4.6 Landing Gear Arrangement ................................................................................................................ 87
2.4.7 Fuselage Design ....................................................................................................................................... 87
2.4.8 Vertical Stabilizer.................................................................................................................................... 87
2.4.9 Horizontal Stabilizer.............................................................................................................................. 87
2.4.10 Performance Analysis ........................................................................................................................... 87
2.4.11 Trade Study: .............................................................................................................................................. 88
2.4.12 V-n Diagram .............................................................................................................................................. 88
2.4.13 Final Model ................................................................................................................................................ 88
2.4.14 Novel Ideas ................................................................................................................................................ 89
2.4.15 Comparison with Similar Aircraft .................................................................................................... 89
2.4.16 Conclusion.................................................................................................................................................. 90
2.4.17 References.................................................................................................................................................. 91

3 Inverse Design Case Studies for Conceptual Aircraft ................................................... 92

3.1 Case Study 1 - An Inverse Design Method for Airfoils Based on Pressure Gradient
Distribution ........................................................................................................................................................................ 92
3.1.1 Introduction .............................................................................................................................................. 92
3.1.2 Numerical Methods ................................................................................................................................ 94
3.1.2.1 Computational Method............................................................................................................. 94
3.1.2.2 Adjoint Method ............................................................................................................................ 95
3.1.2.3 Optimization Method And Design Variable ..................................................................... 97
3.1.2.4 Inverse Design Based on The Target Pressure Gradient............................................ 97
3.1.2.5 Supercritical Airfoil Design .................................................................................................... 97
3.1.2.6 Supercritical Natural Laminar Flow Airfoil Design ................................................... 101
3.1.2.7 Influence of Pressure Gradient on Supercritical Airfoil Performance ............... 104
3.1.2.8 Target Pressure Gradient of Supercritical Airfoil ...................................................... 104
3.1.2.9 Characteristics of Airfoils with Different Upper Surface Pressure Gradients
105
3.1.2.10 Drag Divergence Performance of Airfoils with Different Pressure Gradients
106
3.1.3 Conclusions ............................................................................................................................................ 108
3.1.4 References............................................................................................................................................... 109
3.2 Case Study 2 - Inverse Aerodynamic Design Method for Aircraft Component....................... 112
3.2.1 Introduction and Background ........................................................................................................ 112
3.2.2 Formulation............................................................................................................................................ 112
3.2.3 Method of Solution .............................................................................................................................. 113
 5

3.2.4 Application and Validation............................................................................................................... 113

3.2.5 Conclusion............................................................................................................................................... 114
3.3 Case Study 3 - A Pseudo-Compressibility Method for Solving Inverse Problems Based
on the 3D Incompressible Euler Equations ....................................................................................................... 116
3.3.1 Introduction & Literature Survey ................................................................................................. 116
3.3.2 Mathematical Model ........................................................................................................................... 118
3.3.3 Dynamics of Materially Moving Boundaries. ............................................................................ 119
3.3.4 Numerical Results for 3D Nozzle ................................................................................................... 121
3.3.5 Inverse Problem on a Stator Blade ............................................................................................... 121
3.3.6 Conclusions ............................................................................................................................................ 123
3.4 Case Study 4 – Inversely Designed Scramjet Flow-Path (Chapter From Hypersonic
Vehicles - Past, Present and Future Developments) .......................................................................................... 124
3.4.1 Introduction ........................................................................................................................................... 124
3.4.2 Inverse Scramjet 2D Centerline Design Approach ................................................................. 125
3.4.3 Scramjet Inverse Design Approach .............................................................................................. 125
3.4.3.1 Aerodynamics of the 2D ‘Forebody’ Configuration ................................................... 127
3.4.3.2 Derivation of the 2D ‘Forebody-Inlet-Isolator’ Configuration .............................. 128
3.4.3.3 Design Point at Station A ...................................................................................................... 128
3.4.3.4 Design Points at Station B .................................................................................................... 128
3.4.3.5 Design Points at Station C .................................................................................................... 128
3.4.3.6 Design Points at Station D .................................................................................................... 130
3.4.4 Computer Aided Design (CAD) Design........................................................................................ 131
3.4.4.1 Overview of the 3D Design Process ................................................................................. 131
3.4.4.2 2D Forebody Construction (Side View).......................................................................... 132
3.4.4.3 2D Inlet Construction (Side View) .................................................................................... 133
3.4.4.4 Streamline Preparation of Flow-Field............................................................................. 133
3.4.4.5 Wedge Geometry Extraction From Flow-Field ........................................................... 133
3.4.5 The Caret Geometry ............................................................................................................................ 134
3.4.5.1 3D Stream Tube Construction using the Wave Rider Approach .......................... 134
3.4.5.2 Transforming Stream Tubes to ‘Star’ Shaped Geometries ..................................... 135
3.4.5.3 Validation Section.................................................................................................................... 135
3.4.5.4 2D Simulations.......................................................................................................................... 135
3.4.5.5 2D Euler Simulations Using AVUS .................................................................................... 136
3.4.5.6 2D Isolator Viscous Simulations ........................................................................................ 136
3.4.5.7 3D Simulations.......................................................................................................................... 137
3.4.5.8 Combustor Diffuser Nozzle Sections ............................................................................... 138
3.4.5.9 Scramjet Flow-Path................................................................................................................. 139
3.4.6 Conclusion............................................................................................................................................... 139

4 Structural Design Perspectives of Aircrafts.................................................................. 142

4.1 Generalities and Load Factor ...................................................................................................................... 142
4.1.1 References............................................................................................................................................... 143
4.2 Composite Material in Aircraft Structure............................................................................................... 143
4.3 Design Phases .................................................................................................................................................... 143
4.3.1 Conceptual Design ............................................................................................................................... 143
4.3.2 Preliminary Design.............................................................................................................................. 144
4.3.3 Detail Design .......................................................................................................................................... 146
4.4 System Integration .......................................................................................................................................... 146
4.5 Design of Composite Structures ................................................................................................................. 146
 6

4.6 Constraints and Allowable ........................................................................................................................... 148

4.7 Non-Destructive Testing ............................................................................................................................... 148
4.7.1 Ultra-sonic Methods ........................................................................................................................... 149
4.8 Optimization Process...................................................................................................................................... 150
4.8.1 Multi-Objective Optimization ......................................................................................................... 150
4.8.2 Multilevel Programming ................................................................................................................... 150
4.8.3 Pareto Optimality................................................................................................................................. 150
4.8.4 Minimizing Weighted Sums ............................................................................................................. 151
4.8.5 Weight and Performance Optimization ...................................................................................... 151
4.8.6 Integrated Cost Optimization.......................................................................................................... 151
4.8.7 Combined Cost/Weight Optimization ......................................................................................... 152
4.8.8 Interdisciplinary Wing Design – Structural Aspects ............................................................. 153
4.9 Flutter Phenomena .......................................................................................................................................... 153
4.9.1 Definition................................................................................................................................................. 153
4.9.2 History ...................................................................................................................................................... 153
4.9.3 Detection ................................................................................................................................................. 154
4.9.4 Aeroelasticity ......................................................................................................................................... 154
4.9.5 Flutter ....................................................................................................................................................... 154

List of Tables:
Table 3.4.1 Definition of geometrical parameters ............................................................................................... 74
Table 3.4.2 Mesh and physics settings...................................................................................................................... 75
Table 3.4.3 Comparison of drag coefficient dimensionless on wing surfaces and on fuselage front
surface. The DATCOM values include skin friction, the contributions of upsweep and windshield and
the base drag. .......................................................................................................................................................................... 80
Table 3.5.1 Summary of different calculations presented with the number of FV cells ............. Error!
Bookmark not defined.
Table 3.6.1 Design Requirements ............................................................................................................................... 84
Table 3.6.2 Historical mission segments weight fractions [1]........................................................................ 85
Table 3.6.3 Trade Study .................................................................................................................................................. 88
Table 3.6.4 Different factors of V-n diagram .......................................................................................................... 89
Table 3.6.5 Comparison with Similar Aircraft ....................................................................................................... 89
Table 4.1.1 Force Coefficient Comparisons Of The RAE2822 Airfoil ........................................................... 94
Table 4.1.2 Aerodynamic coefficients of the airfoils (Ma = 0.73, Re = 6.5 × 106).................................... 99
Table 4.1.3 Aerodynamic Performance of the Airfoils (Ma = 0.73, Re = 6.5 × 106). ............................ 102
Table 4.1.4 Transition computations of the airfoils (Ma = 0.73, Re = 6.5 × 106). ................................. 104
Table 4.1.5 Aerodynamic performance of the airfoils with different pressure gradients (Ma = 0.73,
Re = ........................................................................................................................................................................................... 105

List of Figures:
Figure 1.1.1 The evolution of P1 and P2 towards the Pareto front when using a pseudo OF ............. 11
Figure 1.2.1 Contours of MOM vs. Design Variables (Raymer, 2002).......................................................... 13
Figure 1.2.2 Stepping Search ........................................................................................................................................ 14
Figure 1.2.3 Parametrically-created Measure of Merit Data Points ............................................................. 15
Figure 1.2.4 Response Surface Fit to Data Points ................................................................................................ 15
Figure 1.2.5 Orthogonal Steepest Descent Full-Factorial Stepping Search ............................................... 16
Figure 1.2.6 Random Walk ............................................................................................................................................ 17
Figure 1.2.7 Random Initial Population Generated from Design Variables .............................................. 18
 7

Figure 1.2.8 Aerodynamic Optimization via Evolutionary Strategy51 ....................................................... 18

Figure 1.2.9 Optimization by Decomposition ........................................................................................................ 21
Figure 1.2.10 Optimum Fuselage Fineness Ratios............................................................................................... 24
Figure 1.2.11 Traditional Volumetric Density Graph ......................................................................................... 27
Figure 1.2.12 Fuselage and Wing Volume Estimates.......................................................................................... 28
Figure 1.5.1 Fuselage Pressure Distribution Comparison, Boeing 747. Source: AIAA Paper No 72-
188 ............................................................................................................................................................................................... 36
Figure 1.5.2 Comparisons of Crown Line Pressure Distributions For a Low Wing Transport
Configuration at M∞ = 0.84 and α = 2.8 o , Boeing 747. Source: AIAA Paper No 72-188 .......................... 37
Figure 1.6.1 Mach number distribution ................................................................................................................... 37
Figure 1.6.2 Outer flow velocity vectors. M = 0.80, α = 2.5o. Source: AIAA 83-2060............................. 38
Figure 1.6.3 Boeing 747 cab extension, subsonic area ruling. Source: Aeronautics and Astronautics,
1973 ............................................................................................................................................................................................ 39
Figure 2.1.1 Area Distribution ..................................................................................................................................... 42
Figure 2.1.2 CL and CD results from multiple CFD codes, .................................................................................. 42
Figure 2.1.3 CL variation with angle of attack and CM. ........................................................................................ 43
Figure 2.1.4 Chordwise pressure distributions on an inboard and outboard airfoil, at mid-cruise
and buffet-onset CL. NTF results compared to CFL3D. M=0.85, Re=25M. ..................................................... 43
Figure 2.1.5 Planform area increase with Mach number.................................................................................. 45
Figure 2.1.6 Planform change with increasing Mach number ........................................................................ 45
Figure 2.1.7 Spanwise chord increase with increasing Mach number ........................................................ 46
Figure 2.1.8 Pitching moment curves with increasing Mach number ......................................................... 46
Figure 2.1.9 L/D and ML/D change with Mach number.................................................................................... 47
Figure 2.1.10 Crest critical Mach number as a function of ............................................................................... 47
Figure 2.1.11 Sectional compressibility drag curve ............................................................................................ 47
Figure 2.1.12 Comparison of CFD and WingMOD lift to drag ratios ............................................................. 48
Figure 2.1.13 WingMOD balance analysis ............................................................................................................... 49
Figure 2.1.14 Average cruise Mach number times lift to .................................................................................. 49
Figure 2.1.15 BWB 6-250B configuration ............................................................................................................... 50
Figure 2.1.16 BWB 6-250B pressure contours. CFL3DV6, no N/P or winglet .......................................... 50
Figure 2.1.17 BWB 6-250B Pressure Contours. CFL3DV6, no N/P or Winglet ........................................ 51
Figure 2.1.18 Mach times lift to drag ratio of BWB 6-250B ............................................................................. 51
Figure 2.1.19 Compressibility drag Mach trend ................................................................................................... 52
Figure 2.2.1 Cost and Flowtime Characteristics of Wind .................................................................................. 53
Figure 2.2.2 The Aerodynamic Design Cycle.......................................................................................................... 54
Figure 2.2.3 Total Airplane-Related Operating Cost (TAROC) Breakdown .............................................. 55
Figure 2.2.4 CFD Played a Major Role in the Design of the .............................................................................. 58
Figure 2.2.5 Iterative/Inverse Design Process ...................................................................................................... 60
Figure 2.2.6 Key Features of Optimization Based Aerodynamic Design .................................................... 62
Figure 2.2.7 Mach Profiles of Wing Upper Surface Before ............................................................................... 62
Figure 2.2.8 CFD Prediction of Wing Thickness Effects on Drag ................................................................... 63
Figure 2.2.9 The Installation of a Modern High-Bypass- .................................................................................. 64
Figure 2.2.10 CFD Simulation of Installed Engine Exhaust Effects ............................................................... 65
Figure 2.2.11 Engine/Airframe Installation Drag Increments ....................................................................... 66
Figure 2.2.12 Pressure Distributions for a Multielement ................................................................................. 67
Figure 2.2.13 Surface Streamlines from a 3-D High-Lift Navier-Stokes Solution ................................... 67
Figure 2.2.14 CFD Buffet Prediction - Four-Engine Transport ....................................................................... 68
Figure 2.2.15 Wing Section Lift Coefficient Comparison TLNS3D vs Wind Tunnel ............................... 69
Figure 2.2.16 Effect of Spoiler Deflection on Wing Upper Surface Pressure Distribution .................. 69
 8

Figure 2.3.1 Main fuselage geometrical parameters........................................................................................... 73

Figure 2.3.2 Fineness ratio variations, fuselage nose (left), fuselage cabin (middle), fuselage tail
(right). ........................................................................................................................................................................................ 73
Figure 2.3.3 Wing-fuselage fairing design parameters ...................................................................................... 74
Figure 2.3.4 Computational domain and polyhedral volume mesh ............................................................. 75
Figure 2.3.5 Fuselage Drag Coefficient vs Number of Cells and wall y+ ...................................................... 76
Figure 2.3.6 Nose shape factor as a function of nose fineness ratio, α = 0 deg ........................................ 77
Figure 2.3.7 Cabin shape factor as a function of fineness ratio, α = 0 deg ................................................. 77
Figure 2.3.8 Tail shape factor as a function of tail fineness ratio (FRt), α = 0 deg, and scheme of the
geometric relationship between maximum value of upsweep angle (θ) and FRt....................................... 78
Figure 2.3.10 Difference in drag coefficient from the reference wing-fuselage fairing configuration,
for both cruise (left) and climb (right) conditions. LRFWD = 1.50 ....................................................................... 79
Figure 2.3.9 Difference in drag counts with respect to reference wing-fuselage-fairing for the
configurations analyzed, cruise (left) climb (right). ............................................................................................... 79
Figure 2.3.11 Fuselage geometries used for test cases ...................................................................................... 80
Figure 2.4.1 Block Diagram of Conceptual Design ............................................................................................... 84
Figure 2.4.2 Initial Sketch .............................................................................................................................................. 85
Figure 2.4.3 Mission Profile .......................................................................................................................................... 85
Figure 2.4.4 NACA 65-212 Aerofoil............................................................................................................................ 86
Figure 2.4.5 Low wing with dihedral ........................................................................................................................ 86
Figure 2.4.6 Fowler flap .................................................................................................................................................. 86
Figure 2.4.7 V-n diagram ................................................................................................................................................ 89
Figure 2.4.8 Final Aircraft Design ............................................................................................................................... 90
Figure 3.1.1 Computational grid of the RAE2822 airfoil (257 × 129 points) ........................................... 94
Figure 3.1.2 Pressure coefficient comparisons for the RAE2822 airfoil. (a) Case 6: Ma = 0.725, Re =
6.5 × 106, CL = 0.743; (b) Case 9: Ma = 0.730, Re = 6.5 × 106, CL = 0.803 ........................................................ 95
Figure 3.1.3 Free‐form deformation (FFD) frame of the airfoil optimization.......................................... 97
Figure 3.1.4 Pressure Gradient Target of A Supercritical Type Airfoil ....................................................... 98
Figure 3.1.5 Convergence histories of the optimizations with different weights .................................. 99
Figure 3.1.6 Pressure and pressure gradient distributions of the airfoils. (a) Pressure distribution;
(b).............................................................................................................................................................................................. 100
Figure 3.1.7 Mach number contours of the initial airfoil and the optimized airfoils. (a) Initial airfoil;
(b).............................................................................................................................................................................................. 101
Figure 3.1.8 Optimization histories of the supercritical natural laminar flow (NLF) airfoil .......... 102
Figure 3.1.9 Airfoil, pressure and pressure gradient distributions of the supercritical NLF airfoil
.................................................................................................................................................................................................... 102
Figure 3.1.10 Friction Coefficients of The Initial And Optimized Airfoils ............................................... 103
Figure 3.1.11 Eddy viscosity contours of the airfoils. (a) Initial airfoil; (b) optimized with ω = 0.02;
and ............................................................................................................................................................................................ 103
Figure 3.1.12 Pressure gradient target of supercritical airfoil .................................................................... 104
Figure 3.1.13 Pressure gradients on the upper surfaces of two airfoils (Ma = 0.73, Re = 6.5 × 106, CL
= 0.800)................................................................................................................................................................................... 106
Figure 3.1.14 Geometries and pressure distributions of the airfoils (Ma = 0.73, Re = 6.5 × 106, CL =
0.800). ..................................................................................................................................................................................... 106
Figure 3.1.15 Drag creep and drag divergence of the airfoils (Re = 6.5 × 106, CL = 0.800). (a) Drag
creep; (b) ................................................................................................................................................................................ 107
Figure 3.1.16 Pressure distributions at different Mach numbers (Re = 6.5 × 106, CL = 0.800). (a) Ma
= 0.65; (b) Ma = 0.72; and (c) Ma = 0.74 ................................................................................................................... 107
 9

Figure 3.1.17 Flow fields of airfoils at Ma = 0.72, Re = 6.5 × 106, CL = 0.800. (a) cp,x = 0.40; (b) cp,x
= 0.05. ...................................................................................................................................................................................... 108
Figure 3.2.1 Input, Target, and Airfoil Contours and Surface ...................................................................... 114
Figure 3.2.2 Input, Target, and Computed Nacelle Contours and Surface Pressures for an
Asymmetric Nacelle Design Problem ......................................................................................................................... 115
Figure 3.3.1 Schematic View of the Imposed Boundary Conditions for Inverse Problem on a 3D
Nozzle ...................................................................................................................................................................................... 120
Figure 3.3.2 Inverse Problem on a Stator Blade ................................................................................................ 122
Figure 3.4.1 Pod-Mounted Scramjet Concept ..................................................................................................... 125
Figure 3.4.2 Dual-Mode Scramjet Concept........................................................................................................... 126
Figure 3.4.3 Illustration of the Cross Section of the Scramjet...................................................................... 127
Figure 3.4.4 Conceptual 2D Centerline Cross-Section of the Forebody-Inlet-Isolator Scramjet
Section with Flow Physics Representation .............................................................................................................. 127
Figure 3.4.5 Nonweiller Caret Wing Wave Rider Configuration ................................................................. 131
Figure 3.4.6 Preparation for Extracting Information for 2D Base View .................................................. 132
Figure 3.4.7 Generation of 2D Base View for a Wedge ................................................................................... 133
Figure 3.4.8 Generation of 2D Caret-Shaped Geometry, Base View .......................................................... 134
Figure 3.4.9 Wave Rider Derived Stream Tube ................................................................................................. 135
Figure 3.4.10 AVUS Euler Results Density Contours ....................................................................................... 136
Figure 3.4.11 Forebody-Inlet-Isolator Validation Study with 2D Slices .................................................. 137
Figure 3.4.12 Centerline 2D Pressure Contours ................................................................................................. 138
Figure 3.4.13 Mach Contours at the Isolator Exit.............................................................................................. 138
Figure 3.4.14 2D-3D Geometric Construction from Prescribes Isolator Cross-Section and
Aerodynamic Inputs .......................................................................................................................................................... 139
Figure 3.4.15 Illustration of the Transition-Combustor-Nozzle Element ............................................... 140
Figure 3.4.16 A Dual Mode Ramjet-to-Scramjet Concept – Courtesy of Dhanasar M ....................... 140
Figure 3.4.17 4-Pts Scramjet with Circular Combustor C-Sections ........................................................... 140
Figure 3.4.18 4-Pts Scram Jet with Square Combustor C-Sections ............................................................ 140
Figure 4.3.1 Typical Load of Airliner (Courtesy of Airbus)........................................................................... 144
Figure 4.3.2 Levels in Detailed (Courtesy of H. Assler, Airbus Deutschland GmbH) .......................... 145
Figure 4.3.1 Supply Hierarchy in the Commercial Aerospace Industry ................................................. 146
Figure 4.5.1 Portions of Composite material in Airbus Aircrafts (Courtesy of Airbus Industries)
.................................................................................................................................................................................................... 147
Figure 4.5.2 Specific Strength and Stiffness of Different Metals and Alloys (Courtesy of Kaufmann)
.................................................................................................................................................................................................... 147
Figure 4.7.1 Ultrasonic Test Setup in Through Transmission Mode (Courtesy of Kaufmann et al.)
.................................................................................................................................................................................................... 149
Figure 4.8.1 Design solutions, the corresponding Pareto points and the Pareto frontier ............... 150
Figure 4.8.2 Trade-off Between Acquisition and Operating Costs ............................................................. 152
10
 11

1 Aircraft Conceptual Design Practices

1.1 Direct Design
The optimization algorithms used with the direct design method are mainly the gradient based
methods and the stochastic (random) algorithms. Gradient-based methods rely on derivative
information for all the objectives and all the constraints to determine the optimization search
direction. These methods start with a single design point and use the local gradient of the objective
function with respect to changes in the design variables to determine a search direction by using
methods such as the steepest descent method, conjugate gradient method, quasi-Newton techniques,
or adjoint formulations. These methods are efficient and can find a true optimum as long as the
objective function is differentiable and convex. However, the optimization process can sometimes
lead to a local, not necessarily a global, optimum close to the starting point. Furthermore, such
computations can easily get bogged down when many constraints are considered. Genetic
Algorithms and Evolutionary algorithms are typical stochastic optimization algorithms. These
methods are robust optimization algorithms that can cope with noisy, multimodal functions, but are
also computationally expensive in terms of the necessary number of flow analyses required for
convergence. They start with multiple points sprinkled over the entire design space and search for
true optimums based on the objective function instead of the local gradient information by using
selection, recombination, and mutation operations1.

1.1.1 Multi-Objective Optimization

In practical optimization design, there often exist Multi-Objectives Optimization (MOO). The
classical optimization method usually converts a multi-objective optimization problem into a single
objective problem using penalty functions or weighting coefficients. However, for most cases, these
objectives often are incompatible. It’s difficult to set the appropriate penalty functions or weighting
functions which really depends on the experience and preferences of the designer. Moreover, only
one optimal result is obtained after optimization. The designers have no alternative options to
choose. In fact, in most cases, there is no “best” solution by nature, but an infinite number of feasible
solutions which represent different levels of trade-off between the objectives.
This set of solutions is called Pareto optimal set or Pareto Front. A couple of novel Multi-objective
optimization algorithms based on Pareto optimal concept are proposed and applied into optimal
design. They provide a set of non-inferior solutions rather than one “best” solution, which represents
a more reasonable optimal nature. The practical application of multi-objective optimization
algorithms in engineering design is still far away from success, especially in aerodynamic design.
High computational cost and convergence are two critical problems needed to be investigated.2 But
defining the Pareto front is a time consuming requiring a large number of geometry analyses of which
many will be of no interest [Van den Braembussche]3. An alternative is a combination of the
penalties corresponding to the different objectives into one pseudo-OF. Much less geometry analyses
may be needed to reach the optimum.

1 Zhihui Li, Xinqian Zheng,

“Review of design optimization methods for turbomachinery aerodynamics”, Progress
in Aerospace Sciences, July 2017.
2 Xiaodong Wang, “CFD Simulation of Complex Flows in Turbomachinery and Robust Optimization of Blade

Design”, Submitted to the Department of Mechanical Engineering Doctor of Philosophy at the Vrije Universiteit
Brussel July 2010.
3 R A Van den Braembussche, “Challenges and progress in turbomachinery design systems”, IOP Conf. Series:

Materials Science and Engineering 52 (2013).

 12

OF(G) + w1 P1 (G) + w2 P2 (G)

Eq. 1.1.1
This approach was already used in previous sections and is illustrated here by the optimization of
the cooling system of a HP turbine blade [6]. The optimization aims to lower P1 (increasing with the
amount of cooling mass flow) and P2 (increasing when the required life time is not reached). The
lifetime depends of the equivalent stress,
function of the material parameters, stress
and temperature in each point of the blade
[7]. During the optimization process driven
by a pseudo OF, the optimization follows a
path in the design space towards the point
where the lines of constant pseudo OF
become tangent to the Pareto front.
The main advantage of this approach is that
only 30 geometry analyses are needed to
find this optimum. The disadvantage is that
the pseudo OF approach requires a rather
good idea of the relative weights to be given
to both penalties. Increasing the weight on
P1 emphasizes on minimum cooling mass
flow and hence cycle efficiency. Increasing
the weight on P2 emphasizes on lifetime.
The choice of the relative weights is rather
obvious when one objective must be Figure 1.1.1 The evolution of P1 and P2 towards the
satisfied without compromise. The balance Pareto front when using a pseudo OF
between the different penalties may be less
clear in other cases. However perturbing the design parameters around the optimum geometry
provides information about the interesting part of the Pareto front with minimum extra effort. (
Figure 1.1.1).

1.2 Aircraft Conceptual Design and Multidisciplinary Optimization (MDO) as

Envisioned by D. P. Raymer
Daniel P. Raymer [99], in his doctoral thesis has examined the conceptual approach to MDO, as relates
to aircraft design. Consequently, we extracted the relevant material from his thesis, which is deemed
to be important in this section. But, readers, who are interested in MDO applied to conceptual design
of an aircraft, are encouraged to consult the full thesis.
MDO methods were evaluated in terms of their ability to find the optimal aircraft, as well as total
execution time, convergence history, tendencies to get caught in a local optimum, sensitivity to the
actual problem posed, and overall ease of programming and operation. Several key results were
produced with application to both Aircraft Conceptual Design and Multidisciplinary
Optimization, namely:
➢ MDO techniques truly can improve the weight and cost of an aircraft design concept in the
conceptual design phase. This is accomplished by a relatively small “tweaking” of the key
design variables, and with no additional downstream costs.
➢ For a smaller number of variables (< 6 - 8), a deterministic searching method (here
represented by the full-factorial Orthogonal Steepest Descent) provides a slightly better final
result with about the same number of case evaluations.
 13

➢ For more variables, evolutionary/genetic methods get close to the best final result with far-
fewer case evaluations. The eight variables studied herein probably represent the practical
upper limit on deterministic searching methods with today’s computer speeds.
➢ Of the evolutionary methods studied herein, the Breeder Pool approach (which was devised
during this research and appears to be new) seems to provide convergence in the fewest
number of case evaluations, and yields results very close to the deterministic best result.
However, all of the methods studied produced similar results and any of them is a suitable
candidate for use.
➢ Hybrid methods, with a stochastic initial optimization followed by a deterministic final “fine
tuning”, proved less desirable than anticipated.
➢ Not a single case was observed, in over a hundred case runs totaling over a million parametric
design evaluations, of a method returning a local rather than global optimum. Even the
modified commercial airliner, with poorly selected initial design variables far away from the
global solution, was easily “fixed” by all the MDO methods studied.
➢ The postulated set of automated redesign procedures and geometric constraints provide a
more-realistic final result, preventing attainment of an unrealistic “better” final result.
Especially useful is a new approach defined herein, Net Design Volume, which can prevent
unrealistically high design densities with relatively little setup and computational overhead.
Further work in this area is suggested, especially in the unexplored area of automated
redesign procedures for discrete variables.

1.2.1 An Overview of Multidisciplinary Optimization (MDO)

Multidisciplinary Optimization, or MDO, can be described as a collection of mathematical
techniques for multivariable optimization in which the optimization clearly crosses disciplinary
boundaries. An essential feature of MDO is the presence of design constraints and measures of merit
which are of system-level concern. In a typical aircraft conceptual design application, the measure
of merit (MOM) is either cost or its surrogate, weight, where the aircraft is sized4 to some specified
mission which includes the range and payload requirements. The design constraints are typically the
aircraft’s required performance values such as takeoff distance and climb rates, plus any geometric
or operational constraints such as a wingspan limit.
MDO grew out of prior multivariable optimization methods as a natural consequence of the attempt
to apply optimization to more system-level problems. To the cynic, MDO is just the new “buzzword”
for multivariable optimization, but MDO should be recognized rather as a distinct system-level subset
of multivariable optimization. Furthermore, the fairly recent application of MDO as a serious topic
for research has led to the development of new and distinct optimization methods due precisely to
that system-level focus.
A typical application for MDO in the aircraft design field is the simultaneous aerodynamic and
structural optimization of a wing. The wing is defined in terms of some geometric variables, and the
effects on aerodynamics and structural strength are determined as the geometry is varied. Results
are assessed versus some defined measure of merit, and in the presence of constraints which can be
based on performance, safety, operability, or practicality. Other applications of MDO in aerospace
include such diverse areas such as launch vehicle geometric design, composite materials design,
coupled wing body integrated analysis, advanced structural weights estimation, and aerothermal and

4 “Sizing”is a mathematical iterative process which determines the aircraft takeoff gross weight, empty weight,
and fuel weight required such that a given aircraft concept layout can perform a specified mission (range) at a
specified payload weight. This calculated size is used to redraw the aircraft with a revised wing area, fuselage
length, etc…, appropriate to the determined weight. Note that range is a given (independent variable) whereas
aircraft weight (thus aircraft physical size) is the calculated, dependent variable.
 14

sizing optimizations [31,32,33]. Organizations such as Boeing’s Phantom Works are using MDO
techniques in aircraft conceptual design on a fairly routine basis, and report great success at quickly
“weeding through” numerous design alternatives [34]. For a detailed description see Raymer [11]
chapter three.
A wide variety of approaches are being used for defining and solving multidisciplinary optimization
problems including the following, described below:
➢ Finite Difference
➢ Implicit Function Theorem
➢ Stepping Search Methods
➢ Response Surface
➢ Monte Carlo
➢ Random Walk and Simulated Annealing
➢ Evolutionary Algorithms and Evolution Strategy
➢ Genetic Algorithms
➢ Decomposition
In Figure 1.2.1 the contours of the measure of merit of a two-variable design optimization problem
are shown to illustrate the various optimization methods in the following sections. The sought-after
global optimum point is indicated by a star.

Figure 1.2.1 Contours of MOM vs. Design Variables (Raymer, 2002)

Constraints are not shown in this figure, but would indicate infeasible regions in the design space. In
most aircraft optimization problems, the best usable answer is not the unconstrained global best.
Instead it is usually the point closest to the global best on one or more constraint boundaries, where
one can get no closer without violating a constraint (as defined by the Kuhn-Tucker Theorem
mentioned above).
1.2.1.1 Finite Difference
A widely used technique for multivariable or multidisciplinary optimization uses a Finite Difference
approach. Small parametric changes are made to the system (aircraft) one at a time, and the change
in the measure of merit is used to define a slope (first derivative) which represents the system
response (sensitivity) to a change in that variable. These derivatives are then used to predict the
 15

optimum solution, and iteration is used to drive out the obvious linearization errors. Finite Difference
methods were successfully used in conceptual design optimization of a new business jet as reported
by Gallman et al. [35]. Some have reported that the calculation of the finite differences can be
computationally expensive and may produce inaccurate gradient approximations (Newman, et al
[36]). For each step of the iteration, the number of system analyses required is at least one more
than the number of design variables. Also, the difference between the “real world” and the linear
assumption of the method requires that the solution found for each iteration be kept fairly close to
the values of the parameters used in determining the slopes, which increases the number of steps
required and introduces error.
1.2.1.2 Implicit Function Theorem
The Implicit Function Theorem differentiates the various governing equations to obtain sensitivity
equations. These are used to set up simultaneous linear algebraic equations, which are then solved
for derivatives of the objective function as the design variables are changed. These are used to find a
solution. Unfortunately, the governing equations of a real aircraft design problem are both
complicated and design-specific. In the real world, design analysis is limited by such factors as stall
margin, service ceiling, max-continuous throttle setting, FAR/JAR-restricted climb profiles, and
peculiarities of the selected engine's thrust and SFC curves. These do not lend themselves to an
equational representation, especially considering that there are often discontinuities in the
derivatives of the real-world data.
1.2.1.3 Stepping Searches
There are a variety of methods that can be labeled Stepping Searches, in which the objective function
is evaluated and a decision is made as to what direction to move to find a better value of the objective
function without violating any constraints. This direction of maximum local improvement to the
objective function can be found using derivatives of governing equations or finite difference methods
based on actual calculations of the measure of merit, and is called the direction of Steepest Descent.
After determining the best direction to move, a “step” of some defined distance is made in that
direction to a point which becomes the origin for the next calculation and step. This is shown in
Figure 1.2.2 starting
from the point labeled (1).
Steepest descent is
computationally intensive
especially when
constraints are used. The
“best” direction must be
found in a direction that
doesn’t violate the
constraints, requiring a
large number of
evaluations.
Stepping searches are
prone to finding a local
optimum rather than the Figure 1.2.2 Stepping Search
global “best”, depending
on the location of the
starting point. This can be seen in Figure 1.2.2 starting at the point labeled (2). One way to avoid
this is called Multi-start, in which the optimization is rerun a number of times from different starting
points. Stepping searches are frequently employed due to their robust and deterministic nature.
 16

Herbst and Ross [37] used a stepping search routine coupled to a computerized sizing code for early
design optimization of the F-15. For other examples see Schick et al. [38] or Crawford et al. [39].
1.2.1.4 Response Surface
The Encyclopedia of Optimization states that “the evaluation or approximation of derivatives is a
central part of most nonlinear optimization calculations” [40]. This is typically in many MDO
methods including the Finite Difference method and the Implicit Function Theorem described above.
Unfortunately, there is a critical problem with such methods – the presence of numerical noise, which
results in incorrect gradients and can delay or prevent convergence [41]. In the following methods,
such derivatives are avoided by optimizing directly from evaluations of the objective function
(measure of merit). These methods are called zeroth-order or non-gradient methods, and proceed
by calculating specific
values of the MOM at
various combinations of the
design variables. Non-
gradient methods are
described in detail in Hajela
[42]. The research
described herein is
exclusively restricted to
such methods for the
reasons described in the
Introduction.
In the widely-used Figure 1.2.3 Parametrically-created Measure of Merit Data Points
Response Surface method
(“RS”), the design variables
are repeatedly changed to
create a number of different
designs, using either a
simple parametric scheme
or employing Design of
Experiments (DoE) to
determine the best
combination of variables
for optimization purposes
[41]. The resulting system
(i.e., aircraft) variations are
analyzed as to measure of Figure 1.2.4 Response Surface Fit to Data Points
merit and performance
constraints. This creates a
database of specific combinations of the variables and the resulting measures of merit and
performance values. Sample calculated MOM data are illustrated in Figure 1.2.3 using evenly spaced
parametric combinations of the variables.
This parametrically produced data is then fit by least-squares methods to an approximating
multidimensional surface equation, the Response Surface (Figure 1.2.4). This is mathematically or
numerically solved for an optimum, shown in the figure as a circle. Performance constraints are used
as limitations on the allowable solution space. Note that the optimum on the response surface may
not exactly match the optimum in the real objective function (Figure 1.2.1) – but it should be close,
and the “hill” is flat at the optimum so the result should have almost exactly the same value of the
measure of merit.
 17

Response Surface is one of the leading MDO methods, and has the advantage of bounding the number
of alternative values of each variable that must be made to find an optimum. Essentially, a response
surface can be created from any selected number of parametric variations, and the only question is
this – how well does the surface represent the reality? Here, the number of calculations is traded
against the validity of the result. If, by chance, the “true” optimum is located between and away from
any of the points calculated, then the value obtained from the response surface may not be very close
to the true result.
There is also the problem of “goodness of fit”. The response surface is just a curve fit – it may fit well
at the given data points, but poorly match the actual results between those points. Typically, a
response surface is a multidimensional polynomial of only second degree, because higher degrees
increase the computational workload and increase the chance of obtaining a non-real, “wiggly” fit.
However, use of a second-degree polynomial prevents proper fit to a reflexed region of the design
space such as shown in the curve closest to the Variable 2 axis in Figure 1.2.4.
But, these problems are not severe and many researchers and industry designers report excellent
results with Response Surfaces (Mavris and DeLaurentis [43], Cassidy [34]). In addition to the large
reduction in the number of full design evaluations that must be calculated, RS has a further advantage
of naturally smoothing out numerical noise resulting from the parametric analysis. Sevant et al [44]
employed Response Surfaces to optimize flying wing designs primarily due to this noise-smoothing
characteristic. Also, once the RS is fit it is quite easy to develop numerous design tradeoff graphs [45].
Another benefit of the Response Surface method is that the design points are selected and evaluated
external to, and prior to the optimization. This makes it possible to select design points and have real
engineers working offline do the design and analysis work to calculate the system-level response to
changes in the design variables. In one company they go so far as to have designers prepare initial
layouts of dozens of different aircraft concepts spanning the range of parametric design variables.
These are then analyzed, fit to a response surface, and an optimum is determined [34].
1.2.1.5 Steepest Descent (Non-Gradient Stepping Search)
A mathematically simple method labeled Orthogonal Steepest Descent (“OSD”) uses a full-factorial
stepping searching method. It is similar to the Steepest Descent search described above, but uses
neither derivatives nor finite differences to find the direction of maximum local improvement to the
objective function. Instead, the region around the current best is investigated only along the
variables’ axes, and a step of pre-determined size is made in the best direction found (which is
probably not the best direction
there could be, but a few more
step iterations will reach the
“mountain top”). This is shown in
Figure 1.2.5 starting from the
point labeled (1) and proceeding
to the region of the optimum.
Each variable is parametrically
varied by the selected step size
(plus and minus), and the
resulting aircraft are all analyzed.
The aircraft having the lowest
value of the selected measure of Figure 1.2.5 Orthogonal Steepest Descent Full-Factorial
merit that also meets all Stepping Search
performance requirements
becomes the center point baseline for the next iteration loop. This continues until no better variant
is found, then the stepping distance is shortened and the process repeated until some desired level
of resolution is obtained6. No derivatives or finite differences are required because no attempt is
 18

made to find exactly the best “direction” to move – motion is always along the orthogonal axes of one
or more variables.
Because it is robust and deterministic, always finding the same optimum, this method is employed as
a baseline technique in the research herein. OSD requires a large number of steps and a large number
of calculations for each step, implying a lengthy calculation time. The amount of calculation goes up
by at least 3n, where n represents the number of design variables making OSD a poor choice for
optimization of dozens or hundreds of variables. Also, by its nature OSD may find a “local” optimum
as shown in the starting point labeled (2). This allows the possibility of an unfound better solution
on a “different mountain”.
1.2.1.6 Monte Carlo
In the stochastic Monte Carlo method, a random probability function is used to generate a huge
number of potential designs, and all candidates are defined, analyzed, and compare to the other
designs to find the “best” design. This is defined as the design that meets all performance constraints
and has the best value of the selected measure of merit. There is no proof that the “best” design found
is the best that could exist, or that it is even close to the possible global optimum. However, since the
Monte Carlo method is randomly examining the entire design space, it is unlikely that it will return a
local optimum only, and if the found best design is substantially better than the design obtained
without optimization then the method has served a useful purpose. This method is also used in the
research described herein.
1.2.1.7 Random Walk
A stochastic stepping method
called Random Walk applies a
probability function to define a
random direction away from
the current design variable. If it
leads to a better design, a step is
made in that direction and the
process is repeated. This
resembles the steepest descent
search of Figure 1.2.2 but
rather than step in the best
direction found, a step is made
in the first randomly-chosen Figure 1.2.6 Random Walk
direction that offers an
improved value of the measure
of merit. This is also called Drunkard’s Walk, for obvious reasons as can be seen in Figure 1.2.6. Note
that this too is prone to getting stuck on a local optimum.
1.2.1.8 Simulated Annealing
In Simulated Annealing, the tendency of Random Walk to get trapped in a local optimum is avoided
by adding randomness to the acceptance of a “better direction”. Early in the optimization, a
probability function is applied such that sometimes, a step direction is accepted that actually leads to
a worse value of the objective function. This probability is reduced as the optimization progresses so
that, by the end of the optimization run, only “good” directions are accepted. To further ensure that
the result is not a local optimum, the process can be restarted from this supposed optimum, with a
high acceptance of worse values of the objective function reintroduced, then later driven out again
[46]. Simulated Annealing is said to be analogous to the annealing of metals in which the temperature
during cooling is sometimes increased then reduced again, to allow crystalline structures to settle
into their lowest energy state prior to final solidification.
 19

Simulated Annealing was used by Pant and Fielding [47] to simultaneously optimize aircraft
configuration and flight profile of a commuter/regional transport. The algorithm used employs
random step directions with step size adjusted to keep the number of accepted configurations about
the same as the number rejected.
1.2.1.9 Evolutionary Algorithm
An Evolutionary Algorithm works by applying a heuristic process of survival of the fittest to a
defined population of potential solutions (i.e., aircraft). While Darwin is not normally associated with
aircraft design, the
modeling of aircraft
characteristics as genes of
design variables shows
much promise. The design
variables are coded into
(usually) binary strings
such that a collection of 1s
and 0s defines a particular
aircraft by its design
variables (Crossley[48],
Raymer and Crossley [49]).
Rather than starting with a
single baseline design and Figure 1.2.7 Random Initial Population Generated from Design
trying to improve upon it, Variables
an evolutionary algorithm
starts with a number of
binary strings composed of
random 1s and 0s defining
some initial population of
designs (Figure 1.2.7).
The measure of merit is
evaluated for each of these
designs. The optimum
design is improved
through a process
involving selection and
successive generations of
alternative aircraft
individuals as defined by
the designs’ bit-strings.
One of the earliest
applications of an
evolutionary algorithm
was a demonstration using
a simple hinged model, a
wind tunnel, and a set of
dice. Ingo Rechenberg and
Hans Schwefel [50],
students at Technische
Universität Berlin, put the Figure 1.2.8 Aerodynamic Optimization via Evolutionary Strategy51
model in the tunnel and
 20

started with the non-aerodynamic shape shown at the top of Figure 1.2.8. Then they threw the
dice, and used the random results to adjust the angles between the different hinged segments. The
better of different new “individuals” were kept as the “parents” of the next generation. Results from
the first three generations are shown, and a slight progression towards a better shape can already be
seen. After 200 generations, the obvious minimum drag result – a straight board – was obtained
(almost). Rechenberg calls this the Evolution Strategy (Evolutionstrategies).
A similar method called Evolutionary Programming does not attempt to “mate” parents of one
population to create the next generation. Instead, the next generation is created by mutation applied
to the most-fit of the previous generation. A variation of this method is employed in this research.
1.2.1.10 Genetic Algorithm
Another promising type of evolutionary algorithm is the Genetic Algorithm (GA). Members of a
randomly generated starting population of aircraft are analyzed and evaluated as to fitness, based on
the measure of merit, and the most fit are most likely to be permitted to reproduce. Aircraft variants
are defined parametrically by the values of a chromosome-like genetic bit-string. Reproduction
occurs by “crossing” their genes with those from another selected “parent”. The next generation is
evaluated as to fitness, and the process continues until the population all resemble each other or the
measure of merit is no longer improving. This is presumed to represent an optimum.
The concept of a Genetic Algorithm is attributed to J. Holland in the early 1960’s as described in his
landmark book [52], and has been extensively explored for a variety of optimization problems. GA
falls into the class of stochastic global optimization methods, and by its nature is not prone to falling
into a local optimum. This makes it desirable for classes of problems with non-simple objective
function shapes, and/or with complicated constraint functions. It is also useful for classes of
problems with a large number of design variables, where traditional optimizers may simply be
incapable of finding any usable solution.
An advantage of Genetic and Evolutionary Algorithms is their ability to incorporate noncontinuous
(discrete, or integer) variables such as number of engines in an optimization with continuous
variables such as wing sweep. This is difficult to do with derivative-base optimization methods and
generally has to be “faked” by what amounts to duplicate optimizations, one for each potential value
of every discrete variable.
Genetic Algorithms have been applied to aircraft, rotorcraft, and spacecraft design by a number of
researchers including Blasi et al. [53], Perez et al. 54], Roth et al. [55], Crossley [48], and Mosher [61].
A Genetic Algorithm has been applied to better predict the control inputs that may cause departure
on the X-31 Post-Stall Maneuver aircraft [56]. GA has even been applied to assist in identification of
weakly coupled submodules for a decomposition optimization [57]. Key concepts of the Genetic
Algorithm include Selection, Crossover, and Mutation.
1.2.1.10.1 Selection (to Breed)
can be done in many ways, but is always based on a calculated fitness measure that is related to the
optimization objective function (measure of merit) and the constraint functions. One simple criterion
would be, select the individuals with better values of the measure of merit (such as lower sized
takeoff weight) provided that they meet all performance requirements. As this effectively “kills” all
non-performing individuals, it may be too restrictive in that some individuals may have “good genes”
otherwise, but those genes are lost because those individuals miss one performance value by some
small amount.
Another selection criterion could be the calculated value of the measure of merit, post multiplied by
a factor based on violation of the performance constraints. So, the almost good enough individuals
can still be picked for breeding. Selection of which individuals are allowed to breed can be based on
a global stacking of all individuals or on a one-vs. one tournament in which individuals are randomly
paired to “fight it out”, with the winner being allowed to breed with another winner.
 21

A widely-used strategy, the “Roulette Wheel”, assigns a likelihood of being selected based on the
fitness measure then applies a random number generator to determine which individuals are actually
picked. This is much like spinning a roulette wheel wherein the sizes of the “pie slices” are based on
fitness.
1.2.1.10.2 Crossover
Refers to the actual “mating”, the creation of a new individual in the next generation from (usually)
two selected individuals. It too can be done in many ways. Single-point Crossover is done by
bisecting the chromosome strings of the two parents into two parts at some (usually random)
location. The first half of one parent’s chromosome string is pasted to the second half of the other
parent’s chromosome string. Often, two children are created in this manner from each selected pair
of parents by using the leftover halves for the second child [58].
Another scheme, Uniform Crossover, combines the genes bit-by-bit. Each bit is inspected for both
parents – if they are the same, the child has that value (0 or 1). If they are different, a value is randomly
selected. Another option is to treat the genes defining a particular characteristic, such as wing sweep,
as a unit, and randomly pick them from one parent or the other. Some researchers suggest using a
weighted crossover scheme wherein a more-fit parent contributes a greater percentage of the genetic
information of the child [61].
1.2.1.10.3 Mutation
Involves taking the chromosome strings of the children (i.e., the next generation) and generating a
random number for each bit. If a defined low-probability result is obtained, the bit in question is
“flipped” to the opposing value. This has the effect of creating new information in the population and
serves to avoid premature or local convergence, but also interferes with the convergence to a certain
extent. For example, the Genetic Algorithm (GA) may have finally produced the ideal airplane only
to have it lost when the random mutation operator changes a key design parameter.
An important formal analytical result with implications for the utility of the Genetic Algorithm is the
Schema Theorem [52] (also called the Fundamental Theorem of Genetic Algorithms). This
concerns the overall patterns in the chromosome string itself. Such patterns define “good” values of
related design variables as the algorithm proceeds through many generations, and it is important
that good patterns, or schemata, are not lost. Convergence finally occurs because the best schemata
eventually dominate the population. The Schema Theorem is a mathematical expression of the
number of schemata that will exist in the next generation based on the number in the current
generation, the population size, fitness, crossover probability, mutation probability, length of the
chromosome string, and length and order of the various schemata.
The Schema Theorem has several implications of special concern. When using a bisecting crossover
operator, the position on the chromosome string of the various design variables becomes important.
For example, it is well known that three-engine airliners are optimized with a lower thrust-to-weight
ratio (T/W) than two-engine airliners because the effect of losing one engine is less catastrophic. A
chromosome string with T/W right next to number of engines will likely preserve a good schemata
involving them both, whereas if T/W is far from the number of engines an emerging good schemata
will likely be lost during crossover. Other examples are less obvious. It is unlikely that those
developing an optimization code will be able to anticipate every instance where the definition of the
bitstring itself will affect the quality of the optimization result. Also, long schemata are more likely to
be disrupted than shorter ones by crossover. This may prevent the full development of multi-variable
optimum schemata such as, “this wing loading plus this sweep plus this aspect ratio plus this
taper ratio equals a good airplane.”
In GA routines, there is a dichotomy between good “exploitation” of available schemata and good
“exploration” of the design space. Single-point crossover “exploits” the good schemata of designs that
are selected to reproduce, keeping approximately half of their schemata intact depending upon
 22

where the crossover occurs. Uniform crossover “explores” the design space by blending the schemata
of the two selected designs. This often disrupts existing schemata, but introduces new schemata not
present in either parent.
Probably because of these implications, many researchers believe that Uniform Crossover is
preferable to a bisection crossover. However, Uniform Crossover itself tends to break up schemata
by the random selection of a bit where the parents have different bit values.
Another implication of the Schema Theorem is that excess mutation can continuously destroy
emerging schemata, causing the Genetic Algorithm to become more like a random, Monte Carlo
optimizer. This affects both uniform and bisection-type crossovers.
Guidelines for mutation rate per bit suggest that for single-point crossover, the probability of
mutation per bit should be between (1/N) and (1/NL), where N is the population size and L is the
length of the chromosome bit-string. For uniform crossover, empirical evidence59,60 indicates that
a reasonable mutation rate equals ((1+L)/2NL).
The size of the population is important in the Genetic Algorithm method. If too large, the
computational time is excessive and the number of generations may be limited. If population size is
too small there may be insufficient “genetic information” in the initial population, or the random
recombination’s of genetic information may cause useful information to be lost. Crossley [59]
suggests a population size of 30 or more if the chromosome string is less than 30 bits long, and if
more than 50 bits long, the population should greater than 100. Mosher [61] suggests a population
size four times as large as the number of bits in the chromosome bit-string. Values derived by
Goldberg [62] suggest an optimal population size equal to (1.65) 20.21L.
As the Genetic Algorithm, like Monte Carlo, relies heavily on random probability factors and is hence
not deterministic, there is little assurance of getting a repeatable result or knowing that the
“optimum” discovered is actually the very best possible result. There is not even a mathematical
proof of solution convergence in a Genetic Algorithm – but they do work, and often work well!
There are many variations on this basic GA scheme including the manner in which designs are
selected for reproduction, the manner in which their genes are combined to produce the next
generation, and the use of options such as mutation, replacement, and elitism, as discussed below. It
has been said that there are as many Genetic Algorithms as there are researchers developing them.
Many of these GA variations have been coded and evaluated in the research described herein.
1.2.1.11 Decomposition
Decomposition works
by partitioning a large
engineering design
optimization problem
into a number of smaller,
solvable problems (sub-
modules). During
execution of the
optimizer, top-level
routines pass data
between the
submodules in a
structured manner that
retains their coupling
and accommodates the
defined system Figure 1.2.9 Optimization by Decomposition
constraints (Figure
1.2.9). For example, a wing analysis decomposition may have an aerodynamics module that knows
 23

how to calculate drag and air loads if it knows the wing shape, and a structures module that knows
how to calculate weight and structural deflections if it knows the air loads. Each executes separately,
passing their results to the other until they converge at an optimum for the measure of merit such as
weight or drag, or a blended bit of both. Decomposition should really be understood as a framework
for MDO problem simplification, and once accomplished, the decomposed sub-problems are solved
using one of the other optimization techniques.
The key to successful application of decomposition is to separate the design variables, constraints,
and/or analysis methods into groups that are only weakly interconnected. Then we can perform
separate optimizations within these groups, coordinated and linked such that the entire system is
optimized when the separate optimizations are brought together. The groups may also be broken
into weakly-interconnected subgroups, creating a tree-like structure to the optimization process
with a top level that is the entire system, and sub-levels below it representing groups and subgroups.
In textbooks, decomposition it often illustrated using connected structural members such as a door
frame46. This is readily and obviously decomposed into the separate beams comprising the door
frame. Applying such techniques to the problem of aircraft conceptual design optimization remains
difficult, as there are not such obvious, weakly interconnected groupings to decompose. Sobieski [63]
has developed an analytical method for assessing a multidisciplinary system and determining its
sensitivity to its design variables. This method, called Concurrent Subspace Optimization, uses
global sensitivity equations as a way to decompose an optimization problem analytically rather than
by inspection and intuition. Significant research continues into the use of decomposition for aircraft
design problems, and successful industry utilization has been reported (for example see Hollowell et
al. [64], Batill et al. [65] or Isikveren [66]).
1.2.1.12 Stopping Criteria
An important aspect of any optimization method is deciding when to stop. Poorly defined stopping
criteria can cause the execution to run on far longer than required, or can stop execution before a
better result can be found. Typical stopping rules include:
➢ Fixed number of iterations or populations
➢ Fixed amount of execution time/cost
➢ Objective function unchanged for specified number of generations
➢ Small percent improvement in objective function over last value
➢ Similarity in design variables (genes)
➢ Percent coverage of the possible design space
1.2.1.13 Design Variables, Automated Redesign Procedures, and Geometric Constraints
As described before, an important issue for aircraft conceptual design MDO is the realism problem.
To obtain a realistic revised design from an optimization routine, automated redesign procedures
are required. These should approximate the changes that an experienced designer would make to an
existing layout based on particular parametric revisions to the design variables. So, if some change
to the parametric definition of the fuselage prevents the landing gear from working properly, a
human designer would fix it and so should the computer during MDO evaluations. Such procedures
were including in the huge sizing optimization codes of the major aircraft companies [74], but with
set up times of nearly a month these methods do not seem feasible for conceptual-level MDO. In the
context of an MDO code, automated aircraft redesign is really a set of procedures for revising the
analysis inputs. This could be done by actually modifying, via computer algorithm, the 3-D CAD file
defining the aircraft geometry, then extracting from that revised geometry the inputs needed for
analysis. This would be required if CFD or structural FEM were being used for analysis, but is
probably not needed for the analysis methods typical of today’s conceptual design efforts. Certainly,
 24

the analysis inputs of the RDS5-Professional code as used herein do not require such a laborious
procedure.
Instead, the automated aircraft redesign methods can be applied directly to the analysis input file
data. For example, in the classical wing aerodynamic analysis of RDS the wing input data include
wing reference area, actual exposed wing area, aspect ratio, taper ratio, thickness ratio, sweep, design
lift coefficient, skin roughness parameter, estimated laminar flow, and key airfoil data (CL-max and
leading edge parameter ∆Y). For the fuselage, input data include length, equivalent diameter, wetted
area, aft-end upsweep angle, and frontal areas of any windshield or aft-facing base areas. These input
data can be directly manipulated by automated aircraft redesign procedures of varying degrees of
sophistication. In a simple implementation, wing area and fuselage diameter could be changed with
no regard for mutual interactions. In a better implementation, the change in exposed wing area
resulting from a change in fuselage width could be estimated and the input file revised.
In preliminary work in this area6, this author defined a simple but reasonable set of procedures for
such automated redesign. To improve realism, these automated redesign procedures have been
expanded and enhanced. Even with a better set of such procedures this author believes that an
experienced designer should always make a final layout following analytical optimization.
A related subject – in some cases the parametric variations in design variables may yield an aircraft
that is not acceptable for practical reasons. For large airliners such as the Airbus A-380, the wingspan
should not exceed the available ramp space at major airports, limiting the A-380 to a span of 262 feet.
An MDO-optimized concept with wingspan in excess of this would probably be unacceptable to the
customers, so some means of preventing the selection of such a design must be included in the
optimizer. These geometric design constraints are in addition to the performance-based constraints
previously discussed. Selection of the design variables and the automated aircraft redesign
procedures and geometric constraint methods used for this research are described below.
1.2.1.13.1 The Basic Six (or Five) Design Variables
In prior published work (Raymer [27]), this author identified the six most-important variables for
aircraft conceptual design optimization as:
➢ T/W or P/W (i.e., engine size defined by ratio)
➢ W/S (i.e., wing area defined by ratio)
➢ Aspect Ratio
➢ Taper Ratio
➢ Sweep
➢ Airfoil t/c
These six variables include the performance-driving thrust and wing area, plus the parameters that
define the basic wing geometry. These have at least 50 years of history behind them as key
optimization variables, and in this author’s opinion they should be the foundation of any optimization
method intended for aircraft conceptual design. If designing to an existing (fixed-size) engine, then
engine scaling is not possible so a parametric variation of T/W (or P/W) is not possible, hence only
five key variables remain. In addition to the obvious direct changes to the analysis inputs as these
design variables are changed, the aircraft analysis inputs are further modified as follows:
➢ Thrust and fuel flow vary by T/W or P/W
➢ Wing reference area varies based on W/S
➢ Wing exposed area varies based on W/S, adjusted for fuselage width cutoff
➢ Tail areas vary by the 3/2 power of wing area to hold constant tail volume coefficient
➢ Maximum cross-section area for wave drag calculation varies by wing area, t/c, and by

5 Aircraft design software package (“Raymer’s Design System”)

 25

cos(wing sweep), weighted to baseline percentage of total cross-section area6

➢ Nacelle wetted area varies by T/W
➢ Wing fuel volume varies by 3/2 power of wing area
➢ Airfoil Cl-max varies with t/c using empirical regression of NACA airfoils
➢ Airfoil leading edge sharpness parameter (∆Y) varies with t/c
1.2.1.13.2 Fuselage Fineness Ratio
The key top-level parameter for fuselage design is the fineness ratio (f), the fuselage length divided
by its equivalent diameter (diameter that gives the actual cross-section area). Numerous books such
as the classic Hoerner Fluid Dynamic Drag [75] indicate an optimum fineness ratio around 3-4 for
minimizing drag in subsonic flight. However, this "optimal" fineness ratio assumes a constant frontal
area or diameter - in other words, "what length minimizes total drag given a certain maximum
diameter". A more important question for most aircraft is, "what fineness ratio minimizes drag for a
given total volume enclosed?" In a recent study [76] this author wrote a program to vary fineness
ratio of a streamlined shape and calculate the resulting wetted surface area and the drag "form
factor". This is a term in classical drag analysis that accounts for the pressure drags on the back of a
body as a result of viscous separation. The product of wetted area and form factor, times a flat plate
skin friction coefficient, gives the total drag.

Figure 1.2.10 Optimum Fuselage Fineness Ratios

Results are shown in Figure 1.2.10 as fineness ratio is varied with two different assumptions,
constant frontal area (i.e., diameter), and constant volume. The results are normalized to 1.0 to
illustrate relative merit. As can be seen, the constant diameter assumption gives an optimum of 3-4
just as suggested by the old books and manuals. On the other hand, if volume is held constant then
the optimum is somewhere between 6 and 8 - quite a different result! If an aircraft is volume-tight
and is designed using the old suggested values of 3-4, the fuselage drag will be about 25-50% higher
than possible with a fineness ratio of 6.0, according to this analysis.

6 RDS uses the equivalent Sears-Haack body method to estimate supersonic wave drag, in which total aircraft
length and maximum cross-sectional area are key inputs. This adjustment allows rapidly revising this drag
estimate as wing geometry is changed.
 26

This ignores structural effects on the fuselage, which may push the multidisciplinary optimum
solution towards a lower value. To find the true “best” fuselage fineness ratio, it must be included as
a design variable in a multidisciplinary optimization. This was added to the RDS MDO routines, with
the following automatic redesign procedures employed in addition to the obvious input revisions to
fuselage diameter and length:
➢ To hold fuselage volume constant, diameter varies by cube root of fold/fnew
➢ Tail areas vary inversely with fuselage length to maintain constant tail volume coefficient
➢ Landing gear length is scaled to maintain tail-down angle as fuselage length changes
➢ Maximum cross-section area for wave drag calculation varies by fuselage diameter as
fineness ratio changes, weighted to baseline percentage of total cross-section area.
1.2.1.13.3 Design Lift Coefficient (Wing)
Another wing design parameter with great influence on the resulting aircraft is the wing Design Lift
Coefficient (CL-design). This is used during preliminary design as a target for optimization of twist,
camber, and airfoil shape. Selection of a high design lift coefficient is equivalent to selection or design
of an airfoil with high camber, which provides lots of lift at lower speeds but also lots of drag in
cruising flight. During conceptual design CL-design has often been chosen by the designer or chief
aerodynamicist based on experience with similar aircraft. Aircraft drag during cruise will be
minimized if the aircraft cruises at approximately its wing design lift coefficient, calculated by:

W/S
CL−design =
q
Eq. 1.2.1
Ideally, one could select wing loading W/S to provide the desired CL-design. However, wing loading
must often be set to a lower value (larger wing) to obtain the desired stall speed or takeoff/landing
performance. Also, extreme values of CL-design provide poor values of airfoil lift-to-drag ratio. Since CL-
design has an effect on maximum lift, it and W/S should be determined together. Thus, CL-design was
added as the eighth variable in the RDS MDO routines. In addition to simply changing its value in the
aerodynamic analysis inputs, the following effects were included:
➢ Airfoil leading edge sharpness parameter (∆Y) varies with design CL via camber geometric
approximation
➢ Airfoil CL-max varies with design CL using new empirical regression of data for several NACA
airfoils77 (which also includes variation with t/c)
1.2.1.13.4 Geometric Design Constraints
Geometric design constraints were added to the RDS MDO routines to permit searching for an
optimal design with certain real-world requirements considered. These are treated in the
optimization as additional performance constraints. Violations of them, like missing a takeoff
distance requirement, are handled by multiplication of the calculated value of the measure of merit
by the current value of the scalar penalty factor.
1.2.1.13.5 Fuselage Length and Diameter
Fuselage length and diameter limits can be input by the user at the initialization of the optimization.
The length limit is an upper limit, often required in the design of military aircraft to ensure that the
aircraft will fit in hardened shelters and on aircraft carriers. The fuselage diameter limit serves as a
lower limit. This prevents the optimization from making the fuselage smaller in cross-section than
necessary to hold passengers, payload, or equipment as determined in the baseline configuration
drawing.
 27

1.2.1.13.6 Wingspan
The wingspan limit is an upper limit, based on a value input by the user, and mostly applied to large
commercial transports to ensure usability of existing airport taxiways and gates. For military
aircraft, span is constrained to allow the aircraft to fit in hardened shelters and on aircraft carriers.
During conceptual design for the project that became the F-22, one thing that was known early was
that the wingspan could not exceed that of the F-15, for just that reason.
1.2.1.13.7 Wing Geometry for Pitch up Avoidance
For a tailless aircraft or one with a tail positioned such that its effectiveness may be degraded at high
angle of attack, it is important to avoid certain combinations of high aspect ratio and high sweep.
Otherwise, near the stall the outflow from the high sweep will cause the tips to lose lift first. Due to
the high aspect ratio this lost lift is located behind the center of gravity causing pitch up – an
uncontrollable nose-up divergence leading to stall and spin. A widely used pitch up avoidance
criterion was detailed in NACA 1093 (see Raymer [11]). This gives a chart based on extensive wind
tunnel testing that provides threshold curves of acceptable combinations of aspect ratio and sweep.
Data for maximum allowed aspect ratio (A) were curve-fit for subsonic and transonic flight based on
wing quarter-chord sweep (DQC), resulting in the following constraint equations:

(1.047−0.522tan(∆QC))
Subsonic ∶ Amax = 10
(0.842−0.435tan(∆QC))
Transonic ∶ Amax = 10
Eq. 1.2.2
These equations were added to the RDS MDO routines as an optional geometric constraint option.
During optimization, the appropriate equation is used to calculate the maximum allowable aspect
ratio for the design’s wing sweep. If that value is exceeded, the design is penalized in the same
manner as an airplane missing a performance requirement, as described above.
1.2.1.13.8 Net Design Volume
The final geometric design constraint option added to the RDS MDO routines is intended to ensure
that an optimization that makes the wing substantially smaller does not result in an aircraft that
cannot hold its required fuel and internal equipment. This is done with the aid of a parameter called
Net Design Volume (NDV). Net Design Volume was defined by (Raymer [78]) as the internal volume
of an aircraft less the volume dedicated to fuel, propulsion, and payload (including passengers
and crew). NDV represents the volume available for everything else, including items that are not
precisely known until well into the design process such as structural components, avionics, systems,
equipment, landing gear, routing, and access provisions. Therefore, NDV can be used to assure that a
design layout has a credible geometry such that the design, when finalized, will contain all required
components without requiring excessively tight packaging, which can lead to fabrication and
maintenance difficulties. Furthermore, NDV assessment can be used as a constraint in MDO
optimization to help improve the design realism of the resulting optimized configuration.
In conceptual design, a layout is prepared that shows the overall aerodynamic configuration
including fuselage, wings, tails, and the like, and also shows the major internal items such as the
engine, landing gear, fuel tanks, avionics, and the payload, passenger compartment, and crew station.
This drawing cannot and does not show every item that will be in the final, as-built aircraft. Many
items are simply too small to worry about in the initial layout process, although taken together they
add up to substantial volume. Often, internal components have yet to be designed, and will not be
designed until the later "detail design" phase. Examples include the actual aircraft structure,
equipment items such as actuators and environmental control, and the ducting and wiring that pass
throughout the aircraft.
 28

An experienced aircraft designer knows to provide a generous amount of "un-spoken-for" volume in

the aircraft, spaced properly throughout the aircraft. The correct amount of extra space just "looks
right". Such an intuitive measure of merit is difficult to duplicate or teach, and impossible to program.
For this reason, attempts have been made to provide an analytical evaluation of the "right" amount
of extra volume. Over 30 years ago, Cadell79 looked at a volumetric density evaluation to determine
a reasonable volume allocation.
More recently, O'Brimski [80] reviewed the volumetric density of fighter-type aircraft and
determined a practical limit on total aircraft density to assist in evaluation of aircraft proposed to the
US Navy. His methodology
was simple, based on a graph
similar to Figure 1.2.11 of
aircraft total internal volume
(excluding inlet duct) versus
takeoff weight with full
internal fuel and load. The
actual graph has distribution
restrictions, but overall his
data indicates a maximum
practical density ranging from
about 33 lb/ft3 for small
fighters to 31.5 lb/ft3 for
larger fighters.
This methodology is suitable Figure 1.2.11 Traditional Volumetric Density Graph
for analysis of aircraft
proposals but less useful for
conceptual design optimization purposes, for two reasons. First, it is based on detailed evaluation of
the actual volume plot of the completed design, which is not normally available for each alternative
of a parametric study. Second, it is insensitive to some of the likely optimization variables. For
example, a jet engine with higher bypass ratio occupies more volume, requiring an increase in
fuselage size when the concept is properly drawn. Such an engine burns less fuel, reducing the
volume required. These would not be addressed by this earlier method.
1.2.1.13.9 Definition of Net Design Volume
Net Design Volume is a design metric developed to provide a similar, simple method of ascertaining
design realism, with enough extra detail to account for the differences between different aircraft and
between parametric variations of a design undergoing optimization. Net Design Volume is defined
as the internal volume of an aircraft fuselage, nacelles, and wings, less the volume dedicated to
fuel, propulsion, payload, passengers, and crew. NDV more closely represents this "un-spoken-
for" extra volume that the designer knows will provide enough space for everything else, including
aircraft structure and undefined items such as avionics, systems, equipment, landing gear, routing,
and access provisions. Volume of the tails and pylons are not included because those items rarely
provide usable volume (although some aircraft do put fuel in the tails). Also, tightly packed separate
podded nacelles have no extra volume beyond that used for propulsion, and so separate nacelles are
excluded from the evaluation of NDV.

NDV = {Vfus + Vwing } − Vfuel − Vppc

+ Nengine ⏟
{Vnacelle − Veng − Vduct − Vtail pipe }
Ignore if seperate nacelles
Eq. 1.2.3
 29

Internal weapons bays, not addressed in Figure 1.2.11, are accounted for in NDV by the subtraction
of that internal volume. This makes it easier to compare designs with and without such bays, such as
modern stealth fighters.
From NDV, a density is then calculated based on the weights associated with the NDV areas. Basically,
this Wndv is the empty weight less fuel and engine weight. Since tails, pylons, and separate podded
nacelles are excluded from NDV, their weights are removed. Definition of fuel and payload are not
required because both are excluded from NDV. The proper calculation of NDV would use a detailed
aircraft volume plot. This would be a suitable approach for evaluation of a conceptual design layout
in industry or government design offices. For NDV to be a usable tool in MDO computer programs, a
simpler method of calculating NDV is required, provided that it still provided a reasonable
approximation of the effects of parametric variations in design concept.
A simplified volumetric calculation was defined as follows. Fuselage volume was estimated using
three segments. The nose was assumed to be ellipsoidal in cross-section and plan view, with length
equal to 1.5 times height (2.5 for supersonic designs). A constant-section center section was assumed,
based on a tail length defined as 2.5 times height, or 60% of length for supersonic designs. The tail
was defined as an ellipsoidal shape unless the back is cut off as for a rear-mounted engine, in which
case a straight taper was assumed. This is illustrated in Figure 1.2.12.

Figure 1.2.12 Fuselage and Wing Volume Estimates

Wing volume was estimated by determining the areas of the tip airfoil and the airfoil at the root,
where the wing meets the fuselage (not the theoretical root at the center of the airplane). Airfoil area,
based on geometric data for a number of NACA sections, was approximated as 0.67 times the chord
length and thickness. Volume was determined by integration assuming a straight taper between
these end airfoils.
Propulsion volume for an internal jet engine was estimated from the engine diameter and length, and
the inlet length. The engine front face diameter is typically about 80% of engine maximum diameter,
and is the diameter of the back of the inlet. Thus, the area of the back of the inlet is roughly 0.64 times
engine maximum area. To account for expansion of the inlet from front to back, this factor was
reduced to 0.6 and used to estimate inlet volume.
Volumes of the payload (including internal weapons bays), passenger compartment, and crew
compartment were directly assessed by measurement off design layouts. Fuel volume was estimated
 30

from fuel weight, applying standard fuel density values and an appropriate installation factor (88%
installation factor was assumed).
To evaluate whether this definition of Net Design Volume and the approximate volumetric analysis
as outlined above have validity for actual aircraft, seven representative modern fighter aircraft were
subjected to comparative regression analysis using the author’s least-squares multivariable
nonlinear regression program (see Raymer [78]). The selected aircraft are as follows:
➢ F-15A
➢ F-16A
➢ F-18A
➢ Gripen
➢ Rafale
➢ Typhoon
➢ F-22
These aircraft were modeled using data from Jane's All the World's Aircraft [81] , Aviation Week and
Space Technology weekly magazine (various issues), and miscellaneous other sources, as detailed in
Raymer [78]. The earliest versions of these aircraft were deliberately used under the assumption
that these are closest to the original design layout and hence are most representative of the original
designers' definition of "un-spoken-for" volume.
1.2.1.13.10 Automated Redesign for Discrete Variables
More difficult to obtain are redesign procedures for the “discrete” variables, those that have only
integer values such as number of engines or number of aisles and seats across. It was intended in this
research to define automated procedures for aircraft redesign for use in MDO routines, but it quickly
became apparent that any such procedures would be design-specific and limited to a single class of
aircraft. Any routine capable of modifying a propeller-powered canard-pusher aircraft from single
to twin engines would be useless for a design trade in which a twin-engine F-15-like fighter were to
be studied with a single engine. Similarly, an automated redesign procedure that would modify an
Airbus-like transport from seven seats across to eight seats across would probably be inaccurate if
applied to a corporate jet in which a study of three- vs. two-across seating were to be made.
To incorporate optimization of discrete variables into a general-purpose aircraft design program
such as RDS-Professional would require a large number of alternative routines, one for each class and
type of aircraft, and would still never be trustworthy for use in a type of aircraft not already
considered. This is contrary to the philosophy employed in the rest of the RDS-Professional program
and so was abandoned for now.
Traditionally these discrete variable design optimizations have been done by design trade studies in
which an experienced configuration designer produced a new design layout incorporating the
alternative selection. This will probably remain the best way to treat these discrete variables,
although an organization doing numerous optimizations on similar designs may find it useful to code
specific automated redesign procedures for the studies that they do often.
An alternative means of dealing with discrete variables in MDO is to have numerous quick layout
studies done by experienced designers, then distill the results to numerical relationships via
Response Surfaces and apply them to an MDO routine. Here, the question becomes – how well can
designers do such alternative concepts in the limited amount of time available per concept?
The old approach – actually designing and analyzing alternative design concepts – is really not so
bad. Each alternative concept can be designed and optimized separately, and the best of the best
selected as the optimum design. However, the workload goes up exponentially as more discrete-
variable trades are included.
 31

1.2.1.14 References
[1] Wright, W., Wright, O., and McFarland, M. (ed.), The Papers of Wilbur and Orville Wright, McGraw-
Hill Co., N.Y., N.Y., 2001
[2] Ashley, H., On Making Things the Best-Aeronautical Uses of Optimization, AIAA Paper No. 81-
1738 (Wright Brothers Lecture), August 1981
[3] Sobieszczanski-Sobieski, J., Multidisciplinary Design Optimization: An Emerging New
Engineering Discipline, Advances in Structural Optimization (483-496), Kluwer Academic Publishers,
the Netherlands, 1995
[4] P. E. MacMillin et al, An MDO Investigation of the Impact of Practical Constraints on an HSCT
Configuration , AIAA paper 97-0098
[5] L. Blasi et al., Conceptual Aircraft Design Based on a Multi constraint Genetic Optimizer, AIAA
Journal of Aircraft, Vol 37 No. 2, March-April 2000
[6] Raymer, D., Multivariable Aircraft Optimization on a Personal Computer, SAE/AIAA Paper
9965609, World Aviation Congress, 24 Oct. 1996
[7] Gill, P., Murry, W., and Wright, M., Practical Optimization, Academic Press (Harcourt Brace & Co.),
London, UK, 1981
[8] Nicolai, L., Fundamentals of Aircraft Design, E. P. Domicone Printing Services, Fairborn, OH, 1975.
[9] Roskam, J., Airplane Design, Roskam Aviation & Engineering Corp., Ottawa, Kansas, 1985
[10] Stinton, D. The Design of the Aeroplane, BSP Professional Books, 1983.
[11] Raymer, D., AIRCRAFT DESIGN: A Conceptual Approach, American Institute of Aeronautics and
Astronautics, Washington, D.C., Third Edition 1999
[12] Robinson, D., E-mail Communication, Boeing/Rockwell/North American Aviation (retired).
[13] Sobieszczanski-Sobieski, J., E-mail Communication, NASA-Langley, 3-25-02
[14] Wall, R. and Fulghum, D., Lockheed Martin Strikes Out Boeing, Aviation Week, 10-29- 2001.
[15] Meyer, R., 57th SAWE International Conference Keynote Address Transcript, Weight
Engineering, Society of Allied Weights Engineers, Fall 1998
[16] Hildegrandt, G., and Man-bing, S., An Estimation of USAF Aircraft Operating and Support Cost
Relationships, RAND Note N-3062-ACQ, 1990
[17] Raymer, D., Aircraft Computer-Aided Design and Computer-Aided Manufacturing, RAND Corp.
PM-611-AF, 1996
[18] Jameson, A., Re-Engineering the Design Process through Computation, AIAA Journal of Aircraft,
Vol 36 No. 1, Jan-Feb 1999
[19] Vos, J., Rizzi, A., Darracq, D., and Hirschel, E., Navier-Stokes Solvers in European Aircraft Design,
International Journal of Progress in the Aeronautical Sciences, Pergamon, Oxford, U.K. (publication
expected 2002)
[20] Hoak, D., et al., USAF DATCOM, Air Force Flight Dynamics Lab., Wright-Patterson AFB, OH.
[21] Van der Velden, A., Kelm, R., Kokan, D., and Mertens, J., Application of MDO to Large Subsonic
Transport Aircraft, AIAA Paper 00-0844, 38th AIAA Aerospace Sciences Meeting, Reno, NV, Jan. 2000
[22] Giesing, J., and Barthelemy, J., A Summary of Industry MDO Applications and Needs, AIAA Paper
98-4737, 7th Symposium on Multidisciplinary Analysis and Optimization, St. Louis, MO, Sept. 1998
[23] Corning, G., Airplane Design, (self-published), College Park, MD, 1960
[24] Prandtl, L., Uber Tragflugel des Kleinsten Induzierten Widerstandes, Zeitshrift fur Flugtechnik
und Motorluftschiffahrt, 25 Jg, 1933
[25] Göthert, B., Einfluss von Flachenbelastung, Flugelstreckung und Spanweitenbelastung aufdie
Flugleistungen, Luftfahrtforschung, Vol. 6, 1939
[26] Torenbeek, E., Fundamentals of Conceptual Design Optimization of Subsonic Transport Aircraft,
Delft University of Technology Report LR-292, Delft, the Netherlands, 1980
[27] Raymer, D., Vehicle Scaling Laws For Multidisciplinary Optimization, AIAA Paper 01- 0532, 39th
AIAA Aerospace Sciences Meeting, Reno, NV, Jan. 2001
 32

[28] Anon, Exhibit text, Vasamuseet (Vasa Museum), Stockholm, Sweden

[29] Ball, W., A Short Account of the History of Mathematics (4th Edition, 1908), Transcribed by D.
Wilkins, Trinity College, Dublin
[30] Anon., Encyclopedia of Mathematics, Kluwer Academic Publishers, Dordrecht, the Netherlands,
2001
[31] Anon. (AIAA Technical Committee on Multidisciplinary Design Optimization), Current State of
the Art On Multidisciplinary Design Optimization (MDO), AIAA White Paper, ISBN 1-56347-021-7,
September 1991
[32] Lewis, K., Multidisciplinary Design Optimization, Aerospace America, Dec. 2001
[33] Bartholomew, P., The Role of MDO within Aerospace Design and Progress towards an MDO
Capability, AIAA Paper 98-4705 (Defense Evaluation and Research Agency, UK), 1998
[34] Cassidy, P., Telecon, Boeing Phantom Works, St. Louis, MO, 9 Oct. 2001 35 Gallman, J., Kaul, R.,
Chandrasekharan, R., and Hinson, .M., Optimization of an Advanced Business Jet, AIAA Journal of
Aircraft, Vol 34 No. 3,May-June 1997
[36] Newman, J., Taylor, A., Barnwell, R., Newman, P., and Hou, G., Overview of Sensitivity Analysis
and Shape Optimization for Complex Aerodynamic Configurations, AIAA Journal of Aircraft, Vol 36
No. 1, Jan-Feb 1999
[37] Herbst, W., and Ross, H., Application of Computer Aided Design Programs for the Technical
Management of Complex Fighter Development Projects, AIAA Paper 70-364, AIAA Fighter Aircraft
Conference, St. Louis, Mo., 1970
[38] Schieck, F., Deligiannidis, N., and Gottmann, T., A Flexible, Open-Structured Computer Based
approach for Aircraft Conceptual Design Optimization, AIAA Paper 2002-0593, 40th AIAA Aerospace
Sciences Meeting, Reno, NV, Jan 2002
[39] Crawford, C., and Simm, S., Conceptual Design and Optimization of Modern Combat Aircraft,
RTO MP-35, RTO AVT Symposium on Aerodynamic Design and Optimization of Flight Vehicles in a
Concurrent Multidisciplinary Environment, Ottawa, Canada, 1999
[40] Anon., Encyclopedia of Optimization, Kluwer Academic Publishers, Dordrecht, the Netherlands,
2001
[41] Giunta, A., Aircraft Multidisciplinary Design Optimization using Design of Experiments Theory and
Response Surface Modeling Methods, Ph.D. Thesis, Virginia Polytechnic Institute and State University,
Blacksburg, A, 1997
[42] Hajela, P., Nongradient Methods in Multidisciplinary Design Optimization – Status and Potential,
AIAA Journal of Aircraft, Vol 36 No. 1, Jan-Feb 1999
[43] Mavris, D., and DeLaurentis, D., A Probabilistic Approach for Examining Aircraft Concept
Feasibility and Viability, Aircraft Design (International Journal) vol 3-No. 2, Pergamon, June 2000
[44] Sevant, N., Bloor, M, and Wilson, M, Aerodynamic Design of a Flying Wing Using Response
Surface Methodology, AIAA Journal of Aircraft, Vol 37 No. 4, July-Aug 2000
[45] Knill, D., Giunta, A., Baker, C., Grossmann, B., Mason, W., Haftka, R., and Watson, L., Response
Surface Models Combining Linear and Euler Aerodynamics for Supersonic Transport Design, AIAA
Journal of Aircraft, Vol 36 No. 1, Jan-Feb 1999
[46] Haftka, R., and Gurdal, Z., Elements of Structural Optimization, Third Edition, Kluwer Academic
Publishers, Dordrecht, the Netherlands, 1992
[47] Pant, R., and Fielding, J., Aircraft Configuration and Flight Profile Optimization using Simulated
Annealing, Aircraft Design (International Journal) vol 2-No. 4, Pergamon, Dec 1999
[48] Crossley, W., Martin, E., and Fanjoy, D., A Multi objective Investigation of 50-Seat Commuter
Aircraft Using a Genetic Algorithm, AIAA paper 2001-5247, Oct. 2001
[49] Raymer, D., and Crossley, W., A Comparative Study of Genetic Algorithm and Orthogonal
Steepest Descent for Aircraft Multidisciplinary Optimization, AIAA Paper 2002-0514, 40th AIAA
Aerospace Sciences Meeting, Reno, NV, Jan 2002
 33

[50] Schwefel, H., Numerische Optimierung von Computer-Modellen, Birkhauser, Basel, 1974
[51] Anon., Evolution Strategy, (http://www.bionik.tu-berlin.de), Technische Universität Berlin,
2002
[52] Holland, J., Adaption in Natural and Artificial Systems, MIT Press, Cambridge, Massachusetts,
1975
[53] Blasi, L., Iuspa, L., and Core, G., Conceptual Aircraft Design Based on a Multi constraint Genetic
Optimizer, AIAA Journal of Aircraft, Vol 37 No. 1, Mar-April 2000
[54] Perez, R. E., Chung, J., and Behdinan, K., Aircraft Conceptual Design Using Genetic Algorithms,
AIAA Paper 2000-4938, Sep. 2000
[55] Roth, G., and Crossley, W. , Commercial Transport Aircraft Conceptual Design Using a Genetic
Algorithm Based Approach, AIAA Paper 98-4934, Sep. 1998
[56] Ryan, G., A Genetic Search Technique for Identification of Aircraft Departures, NASA Contractor
Report 4688, contract NAS 2-13445, August 1995
[57] Rogers, J., DeMAID/GA USER’S GUIDE, NASA Technical Memorandum 110241, NASA Langley,
April 1996
[58] Davies, W., et al., Twins (film script), Universal Studios, Hollywood, CA, 1988
[59] Crossley, W., Lecture Notes - Multidisciplinary Optimization, Purdue University, 2001
[60] Williams, E. and Crossley, W. , Empirically-Derived Population Size and Mutation Rate
Guidelines for a Genetic Algorithm with Uniform Crossover, Soft Computing in Engineering Design and
Manufacturing, P. K. Chawdhry, R. Roy and R. K. Pant (editors), Springer-Verlag, 1998
[61] Mosher, T., Conceptual Spacecraft Design Using a Genetic Algorithm Trade Selection Process,
AIAA Journal of Aircraft, Vol 36 No. 1, Jan-Feb 1999
[62] Goldberg, D., Genetic Algorithms in Search, Optimization, and Learning, Addison Wesley
Longman, Inc., Boston, Massachusetts, 1989
[63] Sobieszczanski-Sobieski, J., Optimization by Decomposition: A Step from Hierrachic to Non-
hierrachic Systems, Recent Advances in Multidisciplinary Analysis and Optimization; NASA CP-3031;
Part 1, 1988.
[64] Hollowell, S., Beeman, E., and Hiyama, R., Conceptual Design Optimization Study, NASA
Contractor Report 4298, Contract NAS1-18015, 1990
[65] Batill, S., Stelmack, M, and Yu, X., Multidisciplinary Design Optimization of an Electric powered
Unmanned Air Vehicle, Aircraft Design (International Journal) vol 2-No. 1, Pergamon, March 1999
[66] Isikveren, A., Methodology for Conceptual Design and Optimization of Transport Aircraft,
Licentiate Thesis, Report 98-8, KTH-Royal Institute of Technology, Stockholm, Sweden, 1998
[67] Koch, P., Simpson, T., Allen, J., and Mistree, F., Statistical Approximations for Multidisciplinary
Design Optimization: The Problem of Size, AIAA Journal of Aircraft, Vol 36 No. 1, Jan-Feb 1999
[68] Mavris, D., Bandte, O., and DeLaurentis, D., Robust Design Simulation: A Probabilistic Approach
to Multidisciplinary Design, AIAA Journal of Aircraft, Vol 36 No. 1, Jan-Feb 1999
[69] Raymer, D., RDS-Professional User’s Manual, Conceptual Research Corporation, Sylmar, CA, 2000
[70] Raymer, D., RDS: A PC-Based Aircraft Design, Sizing, and Performance System, AIAA Paper 92-
4226, Aug. 1992
[71] Raymer, D., Aircraft Aerodynamic Analysis on a Personal Computer, SAE Paper 932530,
Aerotech 93, Sept 27, 1993
[72] Raymer, D., & Burnside Clapp, M., Pioneer Rocket plane Conceptual Design Study, AIAA Journal
of Aircraft (summer 2002.)
[73] Raymer, D., Genetic Algorithm Optimization Scheme, Unpublished draft, 11-7-01
[74] Robinson, D., E-mail Communication, Boeing/Rockwell/North American Aviation (retired), 9-
12-2000
[75] Hoerner, S., Fluid Dynamic Drag, Hoerner Fluid Dynamics, Brick Town, N.J., 1965
 34

[76] Raymer, D., Dynamic Lift Airship Design Study (contract report for Ohio Airships Inc.),
Conceptual Research Corporation Report CRC-R-01-9, Sylmar, CA, 2001
[77] Abbott, I., and von Doenhoff, A., Theory of Wing Sections, McGraw-Hill, New York, 1949
[78] Raymer, D., Use Of Net Design Volume To Improve Optimization Realism, AIAA Paper 2001-
5246, 1st AIAA ATIO Forum, Los Angeles, CA, Oct. 2001
[79] Cadell, W., On the Use of Aircraft Density in Preliminary Design, SAWE Paper 813, May 1969 80
O'Brimski, F., The Maximum Practical Density for Tactical Aircraft, Internal Technical Note, NAVAIR
AIR-5223, Washington, DC., 1993
[81] Taylor, J., Jane's All the World Aircraft, Jane's, London, UK, (1976, 1992, 2000).
[82] Raymer,D., and Norris, M., Next Generation Fighters: Roles & Concepts, AIAA Aerospace
Engineering Conference, Los Angeles, 1990
[83] Raymer, D., Next-Generation Attack Fighter Conceptul Design Study, Aircraft Design, Vol. 1 No.
1, Pergamon Press, March 1998
[84] Raymer, D., Next Generation Attack Fighter: Design Tradeoffs and Notional System Concepts, The
RAND Corporation, MR-595-AF, Santa Monica, CA, 1996
[85] Raymer, D., and McCrea, M., Design of Jet STOVL Aircraft using RDS-Professional, International
Powered Lift Conference, Arlington, Virginia, Oct. 2000
[86] Raymer, D., Tactical UAV's: Design Requirements Tradeoffs & Notional System Concepts, RAND
Corp., DRR-1763-AF, 1997
[87] Wang, X., and Damodaran, M., Comparison of Deterministic and Stochastic Optimization
Algorithms for Generic Wing Design Problems, AIAA Journal of Aircraft, Vol 37 No. 5, Sept- Oct 2000
[88] Staton, R., Statistical Weight Estimation Methods for Fighter/Attack Aircraft, Vought Aircraft
Rept. 2-59320/8R-50475, 1968
[89] Staton, R., Cargo/Transport Statistical Weight Estimation Equations, Vought Aircraft Rept. 2-
59320/9R-50549, 1969
[90] Jackson, A., Preliminary Design Weight Estimation Program, Aero Commander Div., Rept. 511-
009, 1971
[91] Hess, R., and Romanoff, H., Aircraft Airframe Cost Estimating Relationships, Rand Corp., Rept. R-
3255-AF, Santa Monica, CA, 1987.
[92] Raymer, D., Developing an Aircraft Configuration Using a Minicomputer, Astronautics and
Aeronautics (now Aerospace America), Nov. 1979
[99] Daniel P. Raymer, “Enhancing Aircraft Conceptual Design Using Multidisciplinary Optimization”,
Doctoral Thesis, Report 2002-2, May 2002, ISBN 91-7283-259-2, Department of Aeronautics, Kungliga
Tekniska Högskolan Royal Institute of Technology, SE-100 44 Stockholm, Sweden.

1.3 Inverse Design

In inverse design, for example, the blade geometry is modified to minimize the difference of profiles
of pressure or velocity between the designed and the specified. This method is widely used in 1980s-
1990s since it provides a cheap way for numerical aerodynamic design, such as 2D and 3D blade
optimization. Whereas, considerable experience of the designer is still needed to give a proper profile
of pressure or velocity. In optimization design, the geometry of the blade is sought to maximize the
overall performance parameters, such as efficiency, total pressures ratio, etc. In some sense, it is not
as efficient as inverse design which has clear direction at the beginning, while it makes the design
process less dependent on designers experience, which gives the potential possibility of better design
than specified by designer. Readers could consult various articles in inverse design methodology
 35

such as [Malone et al - see below], as well as the work by [Yin et al]7, [Tan et al]8, [Daneshkah &
Zangeneh]9 are among many others. The inverse design approach relies on extracting the
configuration of interest from the environment in which it operates, similar to Darwin’s theory
of evolution, where an organism adopts to survive in its environment 10. Following case studies
illustrates the inverse design method for aerodynamic components.

1.4 Geometry Consideration

Be advised that the following material in this chapter are directly duplicated from the book by (Obert,
2009)11, which renders are encouraged to consult. The geometry of the aircraft is defined by four
elements, which are:
1 The internal geometry with respect to:
• Required volume, by taking into account
• The shape, i.e. the usefulness of the volume
• Access to the available volume - very important for quick loading and unloading, i.e.
short turn-around times.
2 The internal geometry with respect to structural (strength and stiffness)
3 The external geometry with respect to aerodynamic considerations (determining the
thickness for the desired super velocities).
4 The external geometry with respect to producibility and cost control
A classic example would be to avoid double curved surfaces when possible. However, this depends
much on the material used: composites are very suitable for double curved shapes, but cannot handle
kinks very well.

1.5 Design of the External Geometry

The external geometry of the aircraft should be designed such that it leads to optimum aircraft
performance and satisfactory flight handling characteristics, whilst fulfilling the requirements on
the internal geometry and producibility. The various flight phases and different airworthiness and
flight handling requirements lead to conflicting design requirements concerning basic design data
such as wing loading, CLmax versus L/D and weight. In fact, aircraft design is always about finding a
compromise between opposing requirements. In this chapter, we shall further discuss the design of
the external geometry as determined by aerodynamic design requirements and objectives. Up to
about 1940, the relation between the geometry of aircraft components and aerodynamic
characteristics was heavily based on experimental data (wind tunnel and flight tests). Well-known
examples are the NACA 4-digit and 5-digit series airfoils and the NACA engine cowlings. The
experience of the designer also played a role here, with some “signature” designs – such as the
vertical tail on the De Havilland Hornet Moth and the Vickers Viscount.
After 1940, (limited) theoretical analysis, with an emphasis on airfoils, was done by those that had
“computing power” – large research institutes and airframe manufacturers. This resulted in, for

7 Junlian Yin K Daneshkah and M Zangeneh, “Parametric design of a Francis turbine runner by means of a three-
dimensional inverse design method”, IOP Conf. Series: Earth and Environmental Science 12, 2010.
8 Lei Tan, Shuliang Cao, Yuming Wang and Baoshan Zhu, “Direct and inverse iterative design method for

centrifugal pump impellers”, Journal of Power and Energy.

9 K Daneshkah and M Zangeneh, “Parametric design of a Francis turbine runner by means of a three-dimensional

inverse design method”, IOP Conf. Series: Earth and Environmental Science, 2010.
10 Mookesh Dhanasar, Frederick Ferguson and Julio Mendez, “Inversely Designed Scramjet Flow-Path”, North

Carolina Agricultural and Technical State University, Greensboro, North Carolina, United States of America,
11 Ed Obert, “Aerodynamic Design of Transport Aircraft”, Published By IOS Press Under The Imprint Delft

University Press, ISBN 978-1-58603-970-7, 2009.

 36

example, the NACA 6-series laminar flow airfoils. With the advent of jet engines and aircraft flying
near the speed of sound, a return to wind tunnel experiments took place, as the computer codes of
that time were not capable of handling transonic (mixed subsonic and supersonic) flow. However, a
limited use of theory was made to determine the next step on the experimental process.
The “peaky” airfoil sections (Pearcey et al.) that have been widely applied in the 1960’s are a good
example of research results in this era. Today, the prime aerodynamic design tools are various design
and analysis computer codes, together indicated as computational fluid dynamics (CFD). For the
design flight condition – the cruise – these methods work quite well. The initial climb out (high-lift
devices extended) can also be treated accurately. However, CFD still does not produce sufficiently
accurate results in case of separated flow. And it is exactly in that area that the designers want to
have accurate results, as this determines the limits of the design. These so-called “off design”
conditions include low-speed stall (important for airfield performance), the buffet boundary and the
flight regime between MMO and MD (which are both important for cruise performance). Consequently,
the crucial question in aerodynamic design is at present (next section):

1.5.1 What is the Proper Design (Target) Pressure Distribution?

Which pressure distribution should one try to obtain in the design flight condition which will fulfil
all design requirements? In order to postpone flow separation as far as possible and to minimize
drag, the general goal in aerodynamic design should be:
➢ For components that do not have to produce resultant forces, local Super velocities should be
minimized.
➢ For components that do need to produce resultant forces (such as the wing or rudder), the
pressure distribution at the relevant flight conditions should be optimized such that the
momentum loss in the boundary layer and behind the shockwave is minimal.
➢ Finally, for components that must tolerate a large variation in local flow direction (such as
tail surfaces and engine intakes), leading-edge shapes and design pressure distributions must
be found, which cope with this variation.
 37

1.5.2 Pressure Distributions Over Aircraft Fuselage

The pressure distribution over an aircraft fuselage shows the same characteristics as explained here
and on the previous page. This is illustrated in Figure 1.5.1. The difference in theoretical and
experimental pressure
distribution at the position of
the tail is caused by the theory
not taking the vertical tail into
account. Notice the increase
in pressure between cockpit
and wing. Because here the
fuselage has a cylindrical
shape the pressure coefficient
increases towards zero, for θ =
0° and 90°. For θ =180° the
pressure coefficient even
becomes positive shortly
before the wing because of the
proximity of the stagnation
area at the wing leading edge.
For θ = 0° the effect of the
rapid changes in the cockpit
contour can be seen. The flow
accelerates, decelerates and
accelerates again rapidly at
these locations. Figure 1.5.2
shows a comparison between
the pressure distribution
from wind tunnel tests and as
obtained with a numerical
method. Apparently, the
method used was not
sufficiently sophisticated to
handle the high local Figure 1.5.1 Fuselage Pressure Distribution Comparison, Boeing
velocities at the cockpit 747. Source: AIAA Paper No 72-188
canopy resulting in large
discrepancies between the data from the tests and the calculations.
 38

Figure 1.5.2 Comparisons of Crown Line Pressure Distributions For a Low Wing Transport
Configuration at M∞ = 0.84 and α = 2.8 o , Boeing 747. Source: AIAA Paper No 72-188

1.6 Components Not Required to Generate Aerodynamic Forces

A number of aircraft components are not intended to
generate aerodynamic forces. On these components
local super velocities should be minimized. Such
components are:
➢ Front Fuselage including cockpit and
canopies
➢ Center and rear fuselage section
➢ Engine Struts and pylons
➢ Fins in cruse flight
➢ Tail plane fin fairings , etc.
Number of various Figures are presented to displayed
pressure distributions on various front and center
fuselages (Obert, 2009)12. In Figure 1.6.1 the
pressure distribution over a cockpit is shown, Figure 1.6.1 Mach number distribution
on fuselage nose, McDonnell-Douglas
expressed as a local Mach number distribution. Due
DC-10, M = 0.85
to the convex shape the area over the cockpit

12Ed Obert, “Aerodynamic Design of Transport Aircraft”, Published By IOS Press Under The Imprint Delft
University Press, ISBN 978-1-58603-970-7, 2009.
 39

windows shows an increase in local Mach number up to velocities close to the supersonic regime.
Using (CFD) calculations, local super velocities can be analyzed and improvements could be made to
minimize these. For example the shape of the cockpit can be altered to reduce the size of areas with
supersonic flow, resulting in less drag and cockpit noise.
Using CFD calculations, more can be done to improve the drag characteristics of the aircraft. In
Figure 1.6.2 the outer flow velocity vectors are shown on the front fuselage of the Boeing 757.
Being able to predict these vectors, designers can design a rain protector (rain guide) above the cabin
door in the direction of the flow. If this guide is not parallel to the local flow direction the flow may
separate and increase drag. When wing and fuselage are considered together, super velocities of the
individual components are added.
Before CFD came in general use this
effect was called an interference effect.
The lift over the wing is increased due to
the presence of the fuselage.
The panel distribution used to calculate
the pressure distribution. Adding the
super velocities of the fuselage to those
of the wing alters the variation of the
local lift coefficient over the wing span.
The lift is increased due to the presence
of the fuselage. Also noticeable is the
difference between experimental data
and calculations, which is due to the
calculations having been performed for Figure 1.6.2 Outer flow velocity vectors. M = 0.80, α =
inviscid flow, whereas the experiments 2.5o. Source: AIAA 83-2060
concern viscous flow. When comparing
data from calculations and experiments, the differences found can be reduced by adding both
thickness as well as viscosity effects. If only one of these two is applied, the result in lift will not be
consistent with results found from experiments. The unfavorable summation of super velocities
causes interference drag. In the design process shapes should therefore be pursued such that a
proper interposition of the various components leads to a favorable summation of super velocities.
This means that if one component has a negative pressure coefficient, the intersecting component
should at that location have a low negative or even positive CP.
This can be done by local shape modifications but in some cases this may not be sufficient. At the tail
plane-fin interaction for example a “waisted” body may be necessary. Studies were performed on
reducing the peak pressure coefficients between the fuselage and center nacelle at the fuselage aft-
end.
The original configuration showed supersonic flow and the associated drag due to shock waves. In
the final configuration, this has been reduced considerably. The unexpectedly high drag that was
found during flight tests. This high drag occurred at cruise Mach numbers and was not found in the
wind tunnel tests. This was due to the unfavorable channel flow between the wing and the nacelles
which in flight showed a stronger supersonic flow and shock waves than in the wind tunnel tests
Wing-pylon-nacelle integration offers a real challenge in minimizing interference drag. The nacelle
must be positioned at a certain distance (in longitudinal and vertical sense) away from the wing. This
relation is based on a history of wind tunnel test data.
Using CFD techniques, it has become possible to move the nacelle closer to the wing, even to
previously thought-to-be unacceptable positions. Small improvements in aerodynamic quality in
order to lower the drag on aircraft in production is an ongoing process with most aircraft
manufacturers. When further development within a programmed takes place a considerable
 40

reduction in drag is sometime realized. When the upper deck of the Boeing 747 was extended the
cross-sectional area distribution according to the transonic area rule was improved. This improved
the drag rise Mach number as is shown in Figure 1.6.3.

Figure 1.6.3 Boeing 747 cab extension, subsonic area ruling. Source: Aeronautics and Astronautics, 1973

1.7 Components of Generating Aerodynamic Forces

Aerodynamic forces are required to be generated by:
➢ Wing
➢ Stabilizer (Horizontal Tail plane)
➢ Fin (Vertical Tail plane)
➢ Control Surfaces (elevator, ailerons, rudder, spoiler panels, speed brakes)
Lift-producing surfaces should have the following characteristics, at an acceptable weight, wetted
area and internal volume:
➢ An as high as possible lift-curve gradient (CL vs α)
➢ An as high as possible maximum lift coefficient
➢ An as low as possible drag
➢ An as high as possible angle-of-attack where flow separation occurs.
For different lifting surfaces the order of importance of these points is not identical (for example: for
a fin the CLmax has the lowest priority but not so for a wing).

1.7.1 Pressure Distributions on Wings

Here, some general design considerations for the pressure distribution on wings designed for
transonic flight conditions (the focus of this book is on this kind of aircraft) are given. In defining the
wing shape the following aerodynamic parameters have to be taken into consideration:
1 In cruising flight:
➢ CL
➢ CD
➢ Drag Creep
2 Around the boundaries of cruise flight conditions:
 41

➢ Buffet Boundaries
➢ Maximum Buffet Penetration
➢ Stability and Control above Buffet onset (Pitch and Roll)
➢ Margined Between MMO and MD
➢ Stability and Control between MO and MD
3 At low speeds:
➢ CLmax for all aircraft configurations
➢ Buffet Boundary
➢ Stalling Characteristics for all Aircraft Configurations Over the complete C.G. range
(both pitch and roll)
➢ Lift/Drag ratio at one engine -out initial climb speed
4 For structural and trim drag reasons:
➢ Section zero-lift pitching moment distribution along the span or for the cruse
condition
➢ Moment distribution for aircraft-less-tail
The prime characteristics defining the wing design are:
• Mdesign and CLdesign (the latter due to wing loading W/S and cruise altitude)
• Aspect ratio (A), sweep angle (Λ) and the basic airfoil section in the outboard wing.
Defining these characteristics is of most importance for the outboard wing, because when minimum
induced drag is pursued, then the spanwise lift distribution (Cl x chord versus wing span) has to be
elliptical. Consequently, on a tapered wing, Clmax is found at 60-70% of the semi-span. This is then
the section with the most severe design requirements.
 42

2 Direct Design Case Studies for Conceptual Aircraft

2.1 Case Study 1 - Aerodynamics of High-Subsonic Blended-Wing-Body (BWB)
Configurations

Authors : Dino Roman, Richard Gilmore and Sean Wakayama

Affiliation : The Boeing Company, Huntington Beach, CA 92647
Title of Paper : Aerodynamics of High-Subsonic Blended-Wing-Body Configurations
Source : AIAA 2003-554; https://doi.org/10.2514/6.2003-554
Citation : Dino Roman, Richard Gilmore and Sean Wakayama ; Aerodynamics of High-Subsonic
Blended-Wing-Body Configurations 41st Aerospace Sciences Meeting and Exhibit. AIAA 2003-554.
January 2003.
Adaptation : Only for Formatting

2.1.1 Abstract
A Mach 0.93 Blended-Wing-Body (BWB) configuration was developed using CFL3DV6, a Navier-
Stokes computational fluid dynamics (CFD) code, in conjunction with the Wing Multidisciplinary
Optimization Design (WingMOD) code, to determine the feasibility of BWB aircraft at high subsonic
speeds. Excluding an assessment of propulsion airframe interference, the results show that a Mach
0.93 BWB is feasible, although it pays a performance penalty relative to Mach 0.85 designs. A Mach
0.90 BWB may be the best solution in terms of offering improved speed with minimal performance
penalty.

2.1.2 Introduction
By integrating the functions of wing and fuselage, the Blended-Wing-Body (BWB) achieves a clean
aerodynamic and efficient structural design that offers tremendous potential for reduced fuel burn,
weight, and cost (Refs. 1-3). With the announcement of the Sonic Cruiser, a 0.95 to 0.98 Mach number
configuration, Boeing expressed a new emphasis on increased speed. While the BWB had previously
been studied as a Mach 0.85 configuration, the new emphasis motivated a study to determine if the
advantages of the BWB could be maintained at higher speeds. The natural area ruling of the BWB
indicated that this might be possible. Area ruling is important when considering the wave drag for a
body, which is governed by the following equation:

1 1 1 ′′ 1
D ≈ q0 [ ∫ ∫ S (x)S′′ (ξ)log dxdξ]
2π 0 0 |x − ξ|
Eq. 2.1.1
This equation shows that wave drag varies with the second derivative of cross-sectional area,
implying that breaks in the area distribution result in high wave drag. Such breaks can be seen in the
area distributions of conventional aircraft, like the MD-11, shown in Figure 2.1.1 along with the
area distributions of the BWB and the theoretical minimum wave drag Sears-Haack body.
The conventional airplane has a very non-smooth area distribution, with sharp breaks where the
wing and empennage meet the fuselage. To solve this increased wave drag problem when going to
higher subsonic speeds, conventional airplanes often use an area ruled, or “coke-bottle,” fuselage.
This modification results in a manufacturing cost penalty associated with changing from a pressure
vessel with constant cross section to one with varying cross section. Unlike a conventional airplane,
the BWB has a smooth area distribution that is similar to the Sears-Haack distribution. Since the BWB
is already area ruled, there is no additional cost penalty for changing the character of the pressure
 43

vessel, suggesting that the BWB may

perform at lower cost than a
conventional airplane when
increasing to higher subsonic
speeds.

2.1.3 Tools
2.1.3.1 Navier-Stokes
Computational Fluid
Dynamics (CFD) and
Inverse Design
Due to the unconventional nature of
the BWB (large inboard chords,
thick airfoils with large trailing edge
closure angles, and extreme
blending), results from standard
drag build-up methods based on flat
plate friction and empirical form Figure 2.1.1 Area Distribution
factors are suspect, especially at
chord Reynolds Numbers as high as
300 million. Navier-Stokes analysis, however, is well suited to represent the three-dimensional
physics involved. To conduct the CFD analysis presented in this study, CFL3D, with the Spalart-
Allmaras one–equation turbulence model, was used. CFL3D (Ref. 4) is a NASA-developed Reynolds-
averaged Navier-Stokes code. It incorporates an upwind differencing scheme, which is better at
capturing shocks and avoiding excessive numerical dissipation than a central differencing scheme.
Figure 2.1.2 shows NTF wind tunnel results for a first generation BWB configuration compared to
results from multiple CFD codes (Ref. 5). CFL3D provides a better drag estimate than other popular
CFD codes at cruise
conditions, matching the
drag within 2 counts at
constant CL. Figure 2.1.3
shows almost perfect
agreement in variation of
lift coefficient with angle of
attack, and only a small
discrepancy in pitching
moment variation with
angle of attack. The
magnitude of the pitching
moment coefficient
discrepancy is about 0.01
at its maximum, but the
pitching moment break
still occurs at nearly the
same CL, an important
consideration for buffet
prediction. Pressure
distributions on an inboard
and outboard airfoil are Figure 2.1.2 CL and CD results from multiple CFD codes,
shown in Figure 2.1.4 for compared to NTF results. M = 0.85, Re = 25M
 44

both the mid-cruise CL and the buffet-

onset CL. Buffet is assumed to occur at
or near the pitching moment break.
Again, the CFL3D results agree well
with the NTF wind tunnel results. The
chordwise shock locations and
magnitudes are captured. Excellent
accuracy in drag, lift, pitching moment
and pressure distributions is obtained
with under one million grid points. The
airfoil stacks for the various wings
designed in this study were extensively
modified using a NASA Langley-
developed constrained inverse design
capability, (Ref. 6). Within CDISC, the
user specifies a pressure distribution
and the code determines the geometry
necessary to achieve those pressures Figure 2.1.3 CL variation with angle of attack and CM.
under user specified geometric and NTF results compared to CFL3D. M = 0.85, Re = 25M.
aerodynamic constraints. Coupled to
CFL3D with specified constraints
on airfoil thickness, leading edge
radius, trailing edge closure
angle, pressure vessel height,
shock strength, pitching
moment, and span load, CDISC
allows for realistic tailoring of
the pressures to achieve a
smooth chordwise and spanwise
distribution with weakened
shocks and less aggressive
trailing edge pressure
recoveries. CFL3D coupled to the
CDISC inverse design capability
proved to be an extremely
valuable tool for BWB clean wing
design.
2.1.3.2 Multidisciplinary
Design Optimization
(MDO)
The Wing Multidisciplinary
Optimization Design (WingMOD)
tool was used to perform the
MDO portion of the current
study. As described in Refs. 7-8,
WingMOD models the BWB with
a simple vortex lattice code and Figure 2.1.4 Chordwise pressure distributions on an inboard and
monocoque beam analysis, outboard airfoil, at mid-cruise and buffet-onset CL. NTF results
coupled to give static aeroelastic compared to CFL3D. M=0.85, Re=25M.
 45

loads. The model is trimmed at several flight conditions to obtain load and induced drag data. Profile
and compressibility drag are evaluated at stations across the span of the wing with empirical
relations using the lift coefficients obtained from the vortex lattice code. The compressibility drag
model is calibrated to CFD results. Structural weight is calculated from the maximum elastic loads
encountered through a range of flight conditions, including maneuver, vertical gust, and lateral gust.
The structure is sized based on bending strength and buckling stability considerations. Maximum lift
is evaluated using a critical section method that declares the wing to be at its maximum useable lift
when any section reaches its maximum lift coefficient, which is calculated from empirical data.
These analysis modules are linked to a non-linear gradient-based optimizer. The optimizer is flexible
and allows the user to designate any analysis input as a design variable and any database variable as
a constraint. In typical wing planform optimizations, as described in Ref. 9, a wide variety of
constraints is applied. Mission constraints such as payload, range, and approach speed are applied as
well as design constraints like maximum running loads and buffet characteristics. These design
constraints are put in place to ensure that the optimizer designs a practical configuration that can be
refined later using higher fidelity methods.

2.1.4 Approach
CFD and MDO were used in conjunction to develop a Mach 0.93 BWB configuration. While CFD can
accurately capture compressibility and other aerodynamic effects, a CFD-based design does not
consider constraints such as balance and structural sizing. Additionally, while a CFD-designed
configuration may be aerodynamically efficient, it is not necessarily low weight. Therefore, MDO was
incorporated in the current study to satisfy non aerodynamic constraints and optimize for minimum
take-off weight. The study consisted of a four-step process:
1. Design an aerodynamically efficient configuration using CFD.
2. Calibrate the MDO tool (WingMOD) to CFD results.
3. Using MDO, optimize a BWB for minimum take-off weight.
4. Verify and refine the resultant configuration aerodynamics in CFD.
2.1.4.1 Initial CFD-Based Aerodynamic Design
Because aerodynamic performance must be evaluated using the lower order, but much faster,
methods in the MDO tool, starting from an efficient aerodynamic design at the desired Mach number
would result in a better overall configuration with acceptable aerodynamic performance. Even
though WingMOD was calibrated to CFD at various Mach numbers, straying too far from an initial
design could lead to overly-optimistic aerodynamic performance. Several design cycle iterations
between WingMOD and CFD revealed the advantages of starting with an aerodynamically efficient
configuration.
Starting from a well-established 0.85 Mach configuration, new wings were developed at 0.90 and
0.93 Mach with the goal of maximizing L/D at each Mach number, with consideration given to various
design constraints. In particular, each wing design was driven by the requirement to enclose the
pressurized passenger and cargo cabin, maintain a reasonable buffet boundary and achieve
acceptable post-buffet characteristics (i.e. avoid severe post-buffet pitch-up). A description of the
additional multi-disciplined interdependent real-world constraints affecting aerodynamic design
particular to a BWB are described in detail in Ref. 10. Although these additional constraints were not
specifically tracked or evaluated during the initial CFD phase of the study, they did play a secondary
role in limiting some of the plan form design choices made. All of these constraints were addressed
by the subsequent WingMOD optimizations.
 46

Figure 2.1.6 Planform change with increasing Mach number

While the span was held fixed, the sweep and chord length of the baseline 0.85 Mach design were
systematically varied (effectively reducing t/c for fixed thickness) to minimize wave drag associated
with wing thickness effects, and to
maintain acceptable buffet margin
and characteristics. The resulting
planforms are shown in Figure 2.1.6,
compared to the baseline 0.85 Mach
design. The sweep increased to
minimize wing thickness effects on
wave drag. Because the inboard wing
thickness was driven by the height of
the pressure vessel, there is little one
can do, apart from increasing sweep
and chord, to reduce transonic
thickness effects in this portion of the
wing. The chord increases are shown
in Figure 2.1.7 as a percentage
increase over the baseline 0.85 Mach
configuration chords.
The inboard chord increases were
mainly driven by thickness Figure 2.1.5 Planform area increase with Mach number
considerations while the outboard
 47

chords were driven by the requirement to maintain a reasonable span load and maintain acceptable
buffet margin and characteristics. Increasing the chord length reduced the section lift coefficient for
a given section loading, ccl, which gave more margin to the critical section buffet cl.
Tailoring the chord
lengths in the spanwise
direction allowed the
designer to locate the
buffet-critical section at a
spanwise location that did
not aggravate post-buffet
pitch-up characteristics.
The increases in chord
length lead to significant
increases in wing area as
Mach number increased,
as shown in Figure 2.1.5.
Note the steepening slope
with Mach number.
Figure 2.1.8 shows the
pitching moment curve
for the three wings at
their respective Mach
numbers. Improvement in
Figure 2.1.7 Spanwise chord increase with increasing Mach number buffet margin and pitch-
up characteristics were
seen as Mach number increased. This is partly due to the spanwise wing chord distribution, but also
due to the natural tendency of the wing center-of-pressure location to move aft with Mach number,
leading to a more stable design. Note that other than the baseline 0.85 Mach configuration, the wings
were not trimmed, an issue later resolved using MDO.
The airfoil stack for each wing was
designed using CDISC coupled to CFL3D
as described earlier. The resulting
aerodynamic performance of the
higher Mach number wings is shown in
Figure 2.1.9 compared to the
baseline 0.85 Mach number wing.
While L/D decreased with Mach
number as expected (due to
compressibility effects), ML/D showed
a significant improvement. It is
important to note, however, that these
are purely aerodynamic, untrimmed
wing-alone results. The many non-
aerodynamic constraints not
addressed by the CFD designs (e.g.
Figure 2.1.8 Pitching moment curves with increasing trim, balance, structures) were later
Mach number addressed by MDO. With proper
 48

calibration, the MDO tool emulated

the CFD results, and there was
confidence that it could capture the
important aerodynamic effects
during planform optimizations.
2.1.4.2 Drag Calibration
In analyzing BWB configurations at
Mach numbers going up to 0.95,
there is concern that the simple
WingMOD models may not capture
significant transonic effects. This
concern was addressed by
calibrating the WingMOD models to
CFL3D Navier-Stokes CFD results
for a number of BWB
Figure 2.1.9 L/D and ML/D change with Mach number
configurations, comparing
calibrated WingMOD and CFD
results, and performing CFD design and analysis on the final WingMOD optimized configuration.
Figure 2.1.10 and Figure 2.1.11 show the WingMOD compressibility drag model. Compressibility
drag is determined on a section-by-section basis. For each section, a thickness to chord ratio and lift
coefficient are evaluated
perpendicular to the effective sweep
line, which is determined from a
source-sink thickness model described
in Ref. 11. These properties are then
input to a function represented in
Figure 2.1.10 to determine the
section crest-critical Mach number
(Mcc). Mcc is described as the
freestream Mach number at which the
local flow at the crest of the airfoil, the
location where the surface is tangent to
the freestream direction, becomes
sonic (Ref. 12). Once Mcc is Figure 2.1.10 Crest critical Mach number as a function of
determined, compressibility drag can t/c and cl.
be derived. For each section,
compressibility drag is related to the
ratio of freestream Mach number to
crest-critical Mach number, as shown
in Figure 2.1.11. The curve shown is
represented by a spline that can be
manipulated by the WingMOD
optimizer during calibration.
To calibrate for the current study, a
Mach 0.85 BWB configuration and two
Mach 0.93 BWB configurations were
analyzed in both WingMOD and CFL3D.
To match WingMOD and CFD
Figure 2.1.11 Sectional compressibility drag curve
representations, WingMOD span loads
 49

were tailored to match CFD, and the configurations were analyzed without nacelles, pylons, or
winglets. The WingMOD compressibility drag model was then adjusted, via the coefficients for the
spline shown in Figure 2.1.11, to minimize the error in compressibility drag over all three
configurations.
Procedures described in Ref. 13 for linking variables were used to enable a simultaneous
optimization over the three
configurations to calibrate the
compressibility drag model. Figure
2.1.12 compares lift to drag ratio
(L/D) for the three configurations
relative to the maximum L/D of the
Mach 0.85 configuration, analyzed
both in CFD and in WingMOD after the
calibration. WingMOD L/D levels are
within 5% of CFD and show similar
trends in CL. This is good agreement
considering the WingMOD
aerodynamic analysis runs in a
fraction of a second. It is also possible
that the less mature Mach 0.93
designs will improve relative to the
Mach 0.85 design with further
aerodynamic refinement in CFD,
resulting in better agreement Figure 2.1.12 Comparison of CFD and WingMOD lift to drag
between WingMOD and CFD. ratios

2.1.4.3 MDO Study

Starting from the Mach 0.93 CFD designed baseline, WingMOD was used to design and analyze a
family of BWB configurations with Mach numbers of 0.85, 0.90, 0.93, and 0.95, and ranges of 7500
nmi and 8900 nmi (study described in detail in Ref. 14). Configurations were optimized for minimum
take-off weight, with 154 design variables and 1,091 constraints specified, of which 134 were critical.
This resulted in a system with 20 unconstrained degrees of freedom. Design variables included
structural gauge thicknesses, structural layout parameters, geometry (chords, thicknesses, twist,
etc.), control commands, control schedules, fuel distribution schedules, etc. WingMOD analyzed 28
different conditions, most of which were subject to trim and balance constraints. The resulting
configurations met the mission requirements (i.e. range and payload) as well as satisfying trim,
balance, performance, stability, maximum lift, buffet, structural sizing, and passenger cabin height
constraints.
Figure 2.1.13 demonstrates some of what WingMOD accomplished by optimizing an
aerodynamically efficient design. Figure 2.1.13.a shows the balance diagram for the CFD baseline
design. The points represent c.g. locations for different conditions. The dashed lines represent the
control limits of the aircraft. Several of the c.g. locations fall outside the limits, indicating the aircraft
is not balanced. Figure 2.1.13b shows how WingMOD was able to balance this airplane using ballast.
The addition of ballast increased operating empty weight (OEW) through increases in structural
weight, in addition to the weight of the ballast itself. As the aircraft balanced at more-aft c.g. locations,
L/D increased, improving fuel burn. Take-off weight (TOW) then increased less than OEW, because
the fuel burn improvement partially offset the empty weight increase.
By reshaping the planform, WingMOD was able to solve the balance problem and reduce the TOW of
the aircraft at the same time. The balance diagram for the optimized 0.93 Mach, 7,500 nmi
configuration is shown in Figure 2.1.13c. Figure 2.1.14 shows the Mach number times lift to drag
 50

Figure 2.1.13 WingMOD balance analysis

ratio (ML/D) for each member of the optimized family, relative to the 7500 nmi Mach 0.85
configuration. Aerodynamically, a Mach 0.90 BWB configuration may be optimal, with the peak in
ML/D occurring at that speed. The increase in ML/D over a Mach 0.85 geometry is a result of speed
increasing faster than the lift to drag ratio (L/D) decreases. With the current study emphasizing
speed, a Mach 0.93 design was
selected for further study.
2.1.4.4 CFD Aerodynamic
Refinement
The Mach 0.93, 7500 nmi
WingMOD-optimized configuration
was the basis for the BWB-6-250B,
shown in Figure 2.1.15. CFL3D
coupled to CDISC inverse design
was used to design the airfoil stack,
as discussed earlier, and verify the
aerodynamic performance of the
wing. For simplicity of this design
study the isolated BWB wing was Figure 2.1.14 Average cruise Mach number times lift to
considered without the added drag ratio trend
complications of modeling the
winglet and nacelle and pylon. Whereas the winglet has a fairly localized effect at the wing tip, the
nacelle and pylon can have a more pronounced effect and would need to be integrated in the design
at whichever speed is deemed most appropriate from this initial study. Similar design techniques and
tools as described earlier would be used to perform the propulsion/airframe and winglet integration.
Without the nacelle and pylon, gridding the wing geometry became a simple task using readily
available tools. Airfoil shape and camber were adjusted by CDISC to achieve a smooth chordwise and
spanwise pressure distribution, limit shock strength, and achieve a center of pressure corresponding
 51

Figure 2.1.15 BWB 6-250B configuration

to the c.g. location determined by WingMOD. Additionally, this inverse design process was subject to
constraints on airfoil thickness, leading edge radius, trailing edge closure angle, pressure vessel
height, and span load.

Figure 2.1.16 BWB 6-250B pressure contours. CFL3DV6, no N/P or winglet

 52

2.1.5 Results
Figure 2.1.16 shows
CFL3D predicted
pressure contours and
chordwise pressure
distributions for the
wing. The pressure
distributions show a
very weak inboard
shock well ahead of the
engine inlet location
with a more
pronounced outboard
shock and a tendency to
double shock near the
wing tip. The double
shock tendency is
typical of sections that
are under loaded. The
same techniques used
thus far could be used
Figure 2.1.17 BWB 6-250B Pressure Contours. CFL3DV6, no N/P or to tailor the span load
Winglet and airfoils to address
this double shock,
though judging from the L/D level, there does not seem to be a significant penalty associated with
this characteristic at the tip. Pitching moment variation with CL is shown in Figure 2.1.17. In the
cruise CL range, pitching moment is close to zero, as it should be for trimmed cruise. Buffet onset, as
defined by the break in the pitching
moment curve, does not occur until
well beyond 1.3g’s (1.3 times cruise
CL), and the pitch break is mild,
indicating that post-buffet pitch-up
characteristics will not be severe.
Figure 2.1.18 shows the ML/D of
the configuration at Mach 0.93, 0.94,
and 0.95 relative to a more refined
Mach 0.85 BWB with a similar
mission. At the design cruise Mach
number of 0.93, the BWB 6-250B is at
92% of the peak ML/D of the Mach
0.85 configuration. As expected, the
ML/D drops off as Mach number is Figure 2.1.18 Mach times lift to drag ratio of BWB 6-250B
increased beyond the cruise Mach versus 5-250. CFL3DV6, no N/P or winglet
number. This drag rise can be seen in
Figure 2.1.19. As indicated, drag
divergence, as defined by a slope of 0.05 of the drag rise curve, happens just beyond Mach 0.93,
indicating that the wing is well designed to cruise at Mach 0.93.

2.1.6 Conclusions
A dual CFD/MDO design study was conducted to develop a Mach 0.93 BWB. CFL3D coupled to CDISC
 53

was used to create an

aerodynamically efficient baseline
design, which was then optimized
in WingMOD for minimum take-off
weight, subject to many non-
aerodynamic constraints not
considered in the baseline design.
CFL3D with CDISC was again used to
refine the WingMOD optimized
design. Analysis of the final design
indicated that it achieved
reasonable L/D and a drag
divergence Mach number just
beyond 0.93. Although additional Figure 2.1.19 Compressibility drag Mach trend
CFD work is needed to quantify
drag stemming from propulsion
airframe interference, the work done so far indicates good potential for creating a BWB that performs
well at Mach 0.93.

Acknowledgement
The authors gratefully acknowledges the contributions of the BWB team, especially the following
individuals who were directly involved with or contributed supporting data for the study described in
this paper: Ron Fox, Antonio Gonzales, Ron Kawai, Roger Lyon, and Jennifer Whitlock.

2.1.7 References
[1] Liebeck, R. H., “Design of the Blended-Wing-Body Subsonic Transport,” 2002 Wright Brothers
Lecture, AIAA Paper 2002-0002, Jan. 2002.
[2] Liebeck, R. H., Page, M. A., Rawdon, B. K., “Blended-Wing-Body Subsonic Commercial Transport,”
AIAA Paper 98-0438, Jan. 1998.
[3] “Blended-Wing-Body Technology Study,” Final Report, NASA Contract NAS1-20275, Boeing
Report CRAD-9405-TR-3780, Oct. 1997.
[4] http://cfl3d.larc.nasa.gov/Cfl3dv6/cfl3dv6.html
[5] Pelkman, R.A., “Key Findings and Conclusions from an NTF Wind Tunnel Test of an Initial Blended-
Wing-Body Concept,” NASA Contract NAS1-20268, Boeing Report No. CRAD-9402-TR-3985, 1998.
[6] Campbell, R. L., “Efficient Viscous Design of Realistic Aircraft Configurations,” AIAA Paper 98-
2539, June 1998.
[7] Wakayama, S., Kroo, I., “Subsonic Wing Planform Design Using Multidisciplinary Optimization,”
Journal of Aircraft, Vol. 32, No. 4, Jul.-Aug. 1995, pp.746-753.
[8] Wakayama, S., Lifting Surface Design Using Multidisciplinary Optimization, Ph.D. Thesis, Stanford
University, Dec. 1994.
[9] Wakayama, S., “Blended-Wing-Body Optimization Problem Setup,” AIAA Paper 2000-4740.
[10] Roman, D., Allan, J.B., Liebeck, R.H., “Aerodynamic Design Challenges of The Blended-Wing-Body
Subsonic Transport,” AIAA Paper 2000-4335, June 2000.
[11] Wakayama, S., “Multidisciplinary Optimization of the Blended-Wing-Body,” AIAA Paper 98-
4938, Sep. 1998.
[12] Shevell, R.S., Fundamentals of Flight, 2nd Ed., Prentice Hall, Englewood Cliffs, New Jersey, 1989.
[13] Willcox, K., Wakayama, S., “Simultaneous Optimization of a Multiple-Aircraft Family,” AIAA
Paper 2002-1423, Apr. 2002.
[14] Gilmore, R., Wakayama, S., Roman, D., “Optimization of High-Subsonic Blended-Wing-Body
Configurations,” AIAA Paper 2002-5666, Sep. 2002.
 54

2.2 Case Study 2 - Changing the Role of CFD in Aircraft Development

Author : Edward N. Tinoco (Technical Fellow, Associate Fellow AIAA)

Title : The Changing Role of Computational Fluid Dynamics in Aircraft Development
Affiliation : Boeing Commercial Airplane Group, Seattle, Washington, USA
Original Appearance : AIAA-98-2512
Citation : Edward Tinoco. "The changing role of computational fluid dynamics in aircraft
development," AIAA 1998-2512. 16th AIAA Applied Aerodynamics Conference. June 1998

2.2.1 Abstract
The application (CFD) to the design of commercial transport aircraft has revolutionized the process
of aerodynamic design [E. N. Tinoco, 1998]13. Today, CFD stands alongside the wind tunnel in terms
of importance. The wind tunnel and CFD each has strengths and limitations, and when used together
in complementary roles they enable the achievement of aerodynamic design objectives that
previously were not achievable. The role of CFD must now change to enable an even greater
revolution, that of significantly reducing the aircraft development cycle. This paper describes today's
role of CFD in airplane design and gives a number of specific examples that illustrate the impact of
CFD. It also summarizes the limitations of CFD and describes the challenges of the future necessary
to achieve significant development cycle time reductions.

2.2.2 Introduction
The application of Computational Fluid Dynamics (CFD) today has revolutionized the process of
aerodynamic design. CFD has joined the wind tunnel and flight test as primary tools of the trade [1-
4]. Each has its strengths and limitations. Due to the tremendous cost involved, modern aircraft
development during flight test
must be minimized, and focus
instead on CFD and the wind
tunnel. Figure 2.2.1 illustrates
today’s characteristics of CFD
and the wind tunnel in terms of
relative cost and flow time. The
optimal usage of the two is
concentrated at opposite
corners of the spectrum.
The wind tunnel has the
advantage of dealing with a
"real" fluid, and can produce
global data over a far greater
range of the flight envelope than
can CFD. It is better suited for
validation and data base
building within acceptable
limits of a development Figure 2.2.1 Cost and Flowtime Characteristics of Wind
Tunnels and CFD
program’s cost and schedule, a
task that is unthinkable with
CFD today. However, the wind tunnel typically does not produce data at flight Reynolds number, is
subject to significant wall and mounting system corrections, and is not well suited to provide flow

13Edward N. Tinoco, “The Changing Role of Computational Fluid Dynamics in Aircraft Development”, AIAA-98-
2512.
 55

details. The strength of CFD is to provide an ability to rapidly and cheaply carry out a small number
simulations leading to understanding necessary for design. CFD, by its very nature, can provide
detailed information everywhere in the flow field, but its accuracy is limited by the assumptions of
the mathematical model, solution algorithm, etc. CFD is also used in an "inverse design" or
optimization direction. Here the necessary geometry shape is calculated to meet certain flow
characteristics or payoff function, making CFD better suited for design. Although CFD has sometimes
been used as a replacement for the wind tunnel, that is not yet its principal value. It is CFD's use in a
complementary fashion with the wind tunnel that has yielded the greatest and most cost-effective
benefits to date.
CFD has had its greatest impact in the aerodynamic design of the high-speed cruise configuration of
a transport aircraft. The respective roles of CFD and the wind tunnel today are shown in Figure
2.2.2. The development of candidate aerodynamic lines for most external parts of an airplane is done
with CFD, as are many of the trades and interplay with other disciplines. The role of the wind tunnel
is for after-the fact design validation, and to measure the aerodynamic loads, handling characteristics,

Figure 2.2.2 The Aerodynamic Design Cycle

 56

and performance characteristics throughout the flight envelope. These roles have evolved in ways
that are complementary and which exploit each other’s respective strengths.
Together, CFD and the wind tunnel have led to design solutions and improved aerodynamic
performance that would otherwise not be achievable within the available time and budget of an
aircraft development project. The impact of some of these contributions are illustrated in this paper.
While CFD has had its impact in the discipline of aerodynamic design, it has yet had much impact in
the overall airplane development process. In the late 1960’s, before CFD became part of the aircraft
design process, it took three years to develop the Boeing 747. In the late 1980’s, with significant
impact of CFD in the aerodynamic design, the development of the Boeing 777 took 5 years! No doubt
that CFD contributed to a more competitive product, one that has dominated its field since being
introduced. Part of this increase in the development time was the lengthening aerodynamic design
cycle. This in spite of having to design and wind tunnel test significantly fewer wings than in previous
projects. Part was also due to the ever increasing complexity of a modern commercial transport,
increased competition, more stringent certification requirements, etc.

2.2.3 The Changing World

The world is changing [5]. The airlines and the manufactures that wish to remain in business have
become "market and customer driven." The "bottom line" is now the prime measure of "goodness."
Technology for its own sake is no longer a driving force. Gone are the days when a man like Juan

Figure 2.2.3 Total Airplane-Related Operating Cost (TAROC) Breakdown

 57

Tripe, then president of Pan American World Airways, could or would launch a new airplane program
like the 747 on the basis that it would be the biggest, fastest, most wonderful thing flying. From the
"bottom line" point of view it almost sunk The Boeing Company in the early 1970's and was a factor
in the eventual demise of Pan American.
The birth of a new transport airplane today is driven not by technology, but by market requirements.
The customers determine overall airplane parameters such as range, speed, and size. Besides the
traditional factors of balancing aerodynamic efficiency against structural weight in determining wing
span, manufacturing cost, customer ground handling requirements, mission versatility, etc., must
also be considered. The market establishes the price of the airplane, not the manufacturer. If the price
is such that the airplane adds value to their operation, the airlines will buy it. If it doesn't, they won't.
The market-base price thus establishes a lid on what the manufacturer's cost to design and build
must be.
Figure 2.2.3 shows a typical breakdown of total airplane-related operation cost (TAROC) for a long
range generic wide body twin. The feature that stands out in this figure is the dominant effect of cost-
of ownership on the total operating cost. This component is more than three times the cost of fuel!
This means that if reducing the fuel burn by 3% results in a 1% increase in the cost of ownership
there is no added value to the customer airline. Hybrid laminar flow and riblet technology can reduce
the fuel burn of an aircraft, but at what price?
Does this mean that reducing fuel burn through improved design is no longer a worthy goal for
engineers to pursue? No! Not at all. The airplane still must make the mission. The value of reducing
the fuel burn may not be in the fuel cost savings, but in making the mission, being able to fly a full
payload year around on the customer's routes. What this does mean is that improved technology
must be viewed in terms of its impact on the total value to the customer. This value can be in terms
of reduced fuel burn through lower drag, or reduced maintenance through a simpler design, or
reduced purchase cost due to lower development costs, etc. or any combination thereof. CFD can, and
must have a role in improving value to the customer or it will not have a place in the aircraft
development process. To have a place in this world CFD must be viewed in terms of what it can add
to the product in "bottom line" terms that will help sell the aircraft to the customer and earn a profit
for the manufacturer.
The competitive market place now dictates that a successful new product must be developed and put
into service in substantially less than the time than it took to develop the Boeing 777. How CFD can
contribute to meeting this development challenge will also be discussed in this paper. CFD is not only
useful for aerodynamic design, but it is a competitive necessity if aircraft development times are to
be substantially reduced without compromising performance and safety.

2.2.4 CFD use in Industry

Solutions for aircraft configurations based on the thin-layer, Reynolds-averaged Navies-Stokes
equations are seeing an increasing use in aircraft design [6-8]. However, just because this level of
complexity in the formulation of the flow physics can now be considered, it does not necessarily
follow that it should be. The solution of these complex flows over realistic geometries has greatly
lengthen solution flow times without necessarily adding any value to the design process. Several
factors must be considered when deciding what level of flow physics formulation should be applied
to a specific problem. These not only include the level of physics necessary to adequately simulate
the relevant flow phenomena, but also the level of accuracy and reliability of the numerical method,
the resources available in terms of time and labor to set up the analysis, the computer resources
required, the experience and skill of the CFD application engineers, and the level of confidence in the
solutions. In a time-constrained aircraft development program, timeliness, confidence, and reliability
become the overwhelming factors.
It is the timeliness, confidence, and reliability factors that distinguish the use of CFD in the aircraft
development process from that in the algorithm research and development process. Timeliness is all
 58

important in the design process. The development of the aerodynamic design of a wing, engine
installation, etc. is never the result of a single computation, but rather of a sequential refinement, a
learning process. A practical aerodynamic design must satisfy a myriad of aerodynamic, structural,
and manufacturing requirements and constraints. To complicate matters, these requirements are
frequently changing up until the time the design is "frozen" for manufacture. The quality of a design
turns out to be largely governed by the amount that can be learned during the time available. All
other things being equal, the best designs will be produced by processes that maximize the rate of
learning! The value of CFD to the aircraft development process will be in terms of how it contributes
to the rate at which the engineering design team learns about what it is trying to accomplish [9]. This
can be expressed as:

Learning Cycle Amount Learned

Rate of Learning = × =
Cycle Time Time Allocated
Quality of Product
=
Time to Market
Eq. 2.2.1
The rate of learning is equal to the product of:
1. the amount learned during a design cycle;
2. the number of cycles that can be executed in a given amount of time. It would seem that if we
could just learn everything necessary during the allotted time, then only one cycle would be
necessary.
This is not possible in the real world of design because the question we are trying to answer keeps
changing as we gain more information. What we need is an adequate amount of information per cycle
times a lot of cycles. To support this development/learning process, the CFD solutions are needed in
hours or days, rather than in weeks or months. These short flow times are for the entire CFD process
, geometry definition, grid generation, solution, and processing of the results. Achieving this
performance requires more than just a good CFD flow solver and a fast computer; it requires an entire
system of software and hardware. Geometry and grid definition tools are a key factor in the software
system. The aerodynamic shapes must first be defined before they can be analyzed. These shapes are
frequently complex, including wing, body, tail, struts, nacelles, flap tracks, etc. Computer Aided
Design (CAD) systems are frequently used to aid in the geometry definition process. However, these
systems were intended for the detail design of the individual parts, and are usually not well suited
for dealing with the entire external shape of an aircraft. Proprietary systems have been developed to
support the geometry definition and grid-generation process that follows [10].
Once the configuration’s surface geometry is defined, grid generation becomes the pacing item in
arriving at a CFD solution. The complexity of the grid generation is highly dependent on the
configuration complexity and the CFD solver requirements. Weeks to months seem to be the norm in
generating grids over rather complete aircraft configurations. However, such time scales are not
consistent with the aircraft development schedule. More ever, the robustness and reliability of the
resulting solutions are still at many times lacking. The user does not have the time to coax and tweak
the CFD code to produce an acceptable solution. In general, the simpler the level of flow physics, the
more robust and reliable the solution process. A key feature of future grid generation technology
must be solution adaptivity, in which the grid adapts to significant flow and geometry features during
the solution process.
Once the CFD solution has been completed, the user still must extract the necessary information and
understanding to aid the aircraft design. Rarely is a single number sought from a CFD solution; rather
tens of thousands of bits of information must be digested and understood. High performance
 59

interactive graphics
workstations are used to display
the vast sums of information that
a single CFD solution will
produce. These workstations are
also heavily used in the geometry
and grid generation process. The
required interaction between the
various pre- and postprocessors,
workstations, and computers,
etc. requires a high level of
systems integration and
communications.
In the Boeing Commercial
Airplane Group over 20,000 CFD
runs a year are made to support
the various product lines. Over (a) Simultaneous Design of 3 Engine
85% of these runs are done by
production engineers outside the Installationsncluding Exhaust Effects
research group. The CFD
methods in use provide timely
results in hours or days not
weeks or months. Sufficient
experience with the methods has
given management confidence in
their results. This means that
solutions are believable without
further comparison of known
results with experiment, that the
CFD methods contain enough of
the right physics and resolve the
important physical and
geometric length scales, that the
numeric of the method are
accurate and reliable, and that
the CFD tools are already in place
for there is no time to develop
and validate new methods. Most
of all that the use of CFD makes
economic sense.

2.2.5 Getting CFD Into the (b) The Impact of CFD on The
Product
Computational Fluid Dynamics, Next Generation 737
CFD, methods have been getting
into the product since the late
1970’s during the development Figure 2.2.4 CFD Played a Major Role in the Design of the
of the Boeing 757 and 767. The Boeing 777 and 737
757 cab, which allowed a
common cockpit and type rating with the 767, was designed entirely with CFD and wind tunnel tested
 60

only after high-speeds lines freeze. The engine/pylon installations and to some extent the wing
design were influenced by CFD in both the 757 and 767. CFD methods were used extensively in the
high-speed configuration design of the new Boeing 777, and the design and development schedule
reduction of the derivative Next Generation 737’s. The advances in computing technology over the
years have allowed CFD to impact problems of greater design relevance to aircraft design as
illustrated in Figure 2.2.4 (a-b). Use of these methods allowed a more thorough aerodynamic
design earlier in the development process, allowing greater concentration on operational and safety
related features.
The 777, being a new design, allowed designers substantial freedom to exploit the advances in CFD
and aerodynamics. High-speed cruise wing design and propulsion/airframe integration consumed
the bulk of the CFD application. These two applications will later be discussed in more detail. Many
other features of the aircraft design did not escape being influenced by CFD applications. CFD was
instrumental in design of the fuselage. Once the body diameter was settled CFD was used to design
the cab. No further changes were necessary as a result of wind tunnel testing.
The need for wind tunnel testing in future cab design was eliminated. CFD augmented wind tunnel
testing for aft body and wing/body fairing shape design. CFD provided insight and guided the design
process through the calculation of pressure distributions and streamlines. In similar fashion, CFD
augmented wind tunnel testing for the design of the flap support fairings The wind tunnel was used
to assess the resulting drag characteristics. CFD was used to identify prime locations for static source,
sideslip ports, and angle-of attack vanes for the air data system. CFD was used for design of the
environmental control system, ECS, inlet and exhaust ports, and to plan an unusual wind tunnel
evaluation of the inlet. The cabin (pressurization) outflow valves were positioned with CFD. Although
still in its infancy in regards to high lift design, CFD did provide insight to high lift concepts and was
used to assess planform effects. The bulk of the high lift design work, however, was done in the wind
tunnel [11]. Another collaboration between the wind tunnel and CFD involved the use of CFD to
determine and refine the corrections applied to the experimental data due to the presence of the
wind tunnel walls and model mounting system.
The Next Generation 737-700/600/800/900, being a derivative of earlier 737's, presented a much
more constrained design problem. Again the bulk of the CFD focused on cruise wing design and
engine/airframe integration. Although a new wing, its design was still constrained by the existing
wing-body intersection, and by the need to maintain manual control of the ailerons in case of a
complete hydraulic failure. Similar to the 777 experience CFD was used in conjunction with the wind
tunnel in the design of the wing-body fairing, modifications to the aft body, design of the flap track
fairings and the high lift system.

2.2.6 Cruise Wing Design

The process of aerodynamic high-speed cruise wing design has been revolutionized by CFD. Into the
1970's and early 80's cruise wing design was a cut-and try process with the "try" taking place in the
wind tunnel or flight. Indeed, the subsonic transport had already been developed to such a high
degree of efficiency through "cut and try" empirical design that further aerodynamic improvements
required a high degree of accuracy in the computational simulations. The principal flow phenomena
of interest involves shock waves and strong viscous flow interactions which was only starting to
become amenable to computation in the late 1970's [12]. Significant progress in dealing with these
more complex physical computational simulations has occurred over the past 25 years, both in terms
of algorithm and computing hardware development.6 Initially CFD provided some insight and could
be used to guide the cut-and-try process. Later CFD was doing some of the "try," and was being used
for screening potential designs. If it looked bad in CFD it would be worse in the wind tunnel. Less
wind tunnel iterations were becoming necessary to arrive at an acceptable design.
 61

2.2.7 Iterative/Inverse Design

The first big breakthrough in 3D cruise wing design came in the mid-1980's with the development of
"inverse design," illustrated in Figure 2.2.5. Initially this method was based on solving the full
potential equation with coupled boundary layer [13]. The designer would start with a preliminary
"seed" wing body design with the required wing thickness envelope and planform, and analyze that
at the desired flight conditions. Undesirable characteristics of the resulting wing pressure

Figure 2.2.5 Iterative/Inverse Design Process

distribution would be identified.

A modified wing pressure distribution would be developed to achieve the necessary span load,
chordwise loading architecture, shock sweep and strength. A CFD design code would then be used
to compute the necessary shape changes. The resulting shape typically would not meet all the
thickness and trailing edge closure constraints necessitating another round of design until the
desired pressure characteristics were achieved along with the thickness and closure constraints. The
design would then be checked at some off-design flight conditions with further refinements being
incorporated. Finally the wing would smoothed and furthered modified to meet any remaining
manufacturing constraints. Two or three wings would be designed with slightly differing design
philosophies and wind tunnel tested to determine the best.
The resulting wings exhibited a significant improvement in performance over the "cut-and-try"
designs. These wings attained a high degree of success in the wind tunnel in achieving the desired
"design" wing pressure distributions. This allowed the wind tunnel to test true design parametric:
fine-tune the pressure distribution architecture, sort out the small differences in drag, evaluate
aero/structural design trades, determine off design performance and handling characteristics, etc.
 62

Wing design could now proceed in a more systematic fashion. CFD was providing clues to the relative
performance of wing designs and some preliminary loads information prior to wind tunnel testing.
On the Boeing 777 program CFD contributed to a high degree of confidence in a three cycle wing
development program.
Significantly fewer wing designs were tested for the 777 than for the earlier 757 and 767 programs.
The resulting final design would have been 21% thinner without the "inverse design" CFD capability.
Such a wing would not have been manufacturable, due to skin gage being too thick for the automatic
riveting machines in the factory, and would have less fuel volume. Or it could meet the skin gage and
fuel volume requirements and have been significantly slower. It would not have achieved customer
satisfaction. The impact of CFD wing design in this case was an airplane that has dominated sales in
its class since being offered to the airlines.
In the mid 90’s, the full potential plus coupled boundary layer design method has been replaced by
Navier-Stokes coupled with the Constrained Direct Iterative Surface Curvature (CDISC) program
[14]. The design process remained pretty much that illustrated in Figure 2.2.5. There are a number
of drawbacks to the iterative design approach, however. Foremost is defining a "good" target
pressure distribution for highly three-dimensional flows. Next is achieving that target pressure
distribution while meeting structural and manufacturing constraints on the geometry. Frequently
the two are not mutually compatible, necessitating compromise. A design might meet the target
distribution with acceptable geometry but have aerodynamic problems at off-design. This rapidly
becomes an open-ended iteration process with no guarantee of finding a design compromise that
best satisfies all the constraints and requirements. However, expert designers can and do produce
excellent wing designs, but a better, faster process is needed.

2.2.8 Multipoint Optimization

A better approach is multipoint optimization. A well-formulated optimization method will allow us
to quickly achieve a good compromise between aerodynamic objectives and the constraints imposed
by structures and manufacturing. It will also allow us to simultaneously consider multiple flight
conditions so that the iterations with off-design considerations can be minimized or eliminated.
Methods based on using finite differences, ADIFOR, or the adjoint formulation with the Navier-Stokes
equations are being investigated by the Douglas Products Division, and the Phantom Works at Boeing
[4,15,16]. Another approach to multipoint optimization capability has been developed for use within
the TRANAIR code [17,18]. TRANAIR is a 3D full potential finite element code with directly coupled
boundary layer. It can handle very general geometry and features automatic geometry and solution
adaptive grid generation. The code has both analysis and design/optimization capability.
Key elements of the TRANAIR optimization package, illustrated in Figure 2.2.6, include: geometry
mode shape specification to define how the geometry may be perturbed; a pay-off function to be
minimized, i.e. pressure distribution least-squares matching, Mach gradient minimization, drag
minimization, etc.; and constraints to bound the resulting solution. A combination of equality and
inequality constraints can be specified on geometry and flow parameters, i.e. specified lift and
moments constraints, minimum thickness, curvature, and smoothness constraints, etc.
These elements are very general and adaptable, a necessity since each optimization problem has its
own differences. For multipoint optimization, TRANAIR calculates the solution sensitivities due to
geometry perturbations at each specified flow condition that are then used to formulate a single
optimization problem subject to the various specified constraints. Each flow condition can be given
a different weight for the optimization problem. The resulting solution represents the best
satisfaction of the pay-off function subject to the specified constraints and the freedom allowed by
the geometry modes. If the resulting design is not acceptable then changes must be made to the
constraints, geometry modes, and/or the pay-off function to improve the design.
The optimization code, in itself, is not the design process. The process must be developed that can
exploit the capabilities of the code to deliver better, faster, cheaper designs. Initial applications of the
 63

Figure 2.2.6 Key Features of Optimization Based Aerodynamic Design

TRANAIR method were limited to single point designs, and used least-squares pressure matching or
gradient reduction as the primary pay-off function. These were essentially learning exercises in
applying this new approach to practical problems. An example is its application to the refinement of
the inboard wing design of a twin-
engine transport configuration.
This configuration had initially
been designed using the
iterative/inverse design method.
The addition of the engine strut
and nacelle resulted in highly 3D
flow with a small region of
excessively
high local Mach and a weak normal
shock on the inboard part of the
wing between the over-wing strut
fairing and the side of body. This
posed a challenging design
problem to improve the flow in
this region while maintaining all
the constraints imposed by
structures and manufacturing. Figure 2.2.7 Mach Profiles of Wing Upper Surface Before
and After Design
 64

The goal of the design was to reduce the drag caused by the shock. At that time the drag
calculation capability of the code was not reliable so the objective function was defined such that the
pressure gradients on the wing upper surface would be minimized. Constraints included a maximum
local Mach number on the wing, lift coefficient, and a very large number of geometric constraints,
including inequality streamwise and spanwise curvature constraints for manufacturing.
Eventually over 13,000 constraints were specified for the problem. Many of these constraints came
about only after several iterations of the design. The geometry mode shape variables were specified
along six chordwise rows between the side of the body and the nacelle strut. The optimizer was able
to eliminate the normal shock while maintaining all the geometric constraints. Local Mach contours
on the wing upper surface before and after optimization are shown in Figure 2.2.7.
A more desirable objective function in cruise wing design is usually drag minimization subject to a
variety of constraints. Least-squares pressure matching, gradient reduction, local Mach limits, etc.
have all been attempts at indirect objective functions to minimize drag. This has been done primarily
because CFD drag calculations were not trusted. Without good drag prediction, particularly drag
increments due to geometry changes, drag optimization is not possible.
Recent advances in drag prediction capability in TRANAIR have led to its use as an objective function
in optimization. Figure 2.2.8 compares experimental drag with results from TRANAIR and TLNS3D,
[19] a Navier-Stokes solver, for two configurations that had the same wing planform and design
philosophy but differed in spanwise thickness distribution.

Figure 2.2.8 CFD Prediction of Wing Thickness Effects on Drag

For clarity only the TRANAIR results are shown on the drag polar. These have been arbitrarily shifted
by the same amount to match the experimental drag level. Also shown on this figure is the
incremental drag difference between the two configurations as predicted by the two codes. The CFD
agreement with experiment is very good.
While the drag predictive capabilities of TLNS3D have been shown to very reliable for transport wing-
body configurations, the same cannot be said when nacelles and struts are added. As will be shown
later in this paper, TRANAIR has excellent predictive drag capability for transport configurations with
underwing nacelles. Drag optimization opens up new approaches to design. It also makes single and
multipoint optimization practical. On a recent project two wings were design in the presence of the
engine strut/nacelles for a long range four-engine aircraft. One wing was designed using the
 65

traditional iterative design method, the other using TRANAIR multipoint optimization at three flight
conditions.
Both wings had identical planforms and thickness distributions. Both were built and tested in the
wind tunnel as four-engine configurations with the same flow-through nacelles. The multipoint
design exhibited better drag polar shape, lower drag level, higher drag rise Mach number, and higher
buffet lift margin compared to the traditional design. The multipoint design also required less
engineering hours and is more amenable to further design flow time reductions. The impact of CFD
design/optimization on future aircraft wing design will be improved performance, but more
importantly reduced design flow time.

2.2.9 Engine Airframe/Integration

An early CFD success was the improvement of the understanding of the interference drag of a pylon-
mounted engine nacelle under the wing. The existence of unwanted interference drag had been
revealed by wind tunnel testing, but the physical mechanism of the interference was still unknown.
To avoid the interference drag, it is common practice to move the engine away from the wing. The
resulting additional weight and drag, due to the longer engine strut, must be weighed against the
potential interference drag if the engine is moved closer to the wing. CFD studies (linear panel
methods), along with specialized wind tunnel testing in the mid-1970's, provided the necessary
insight into the flow mechanism responsible for the interference. This understanding led to the
development of design guidelines that allowed closer coupling of the nacelle to the wing [20]. The
Boeing 757, 767, 777, 737-300/400/500 series, Next Generation 737/600/700/800/900 series and
the KC-135R are all examples of aircraft where very closely coupled nacelle installations were
achieved without incurring a significant drag penalty.
In the case of the 737, this series of aircraft might not have remained in production beyond the mid-
1980's, if not for the installation of a modern, large diameter, high bypass ratio turbofan engine.
Competitive pressures from the McDonnell-Douglas MD-80 demanded a large improvement in
operating efficiency and capability if the 737 family was to remain in production. In addition, new
federal regulations demanded significant noise reduction in the up-coming years. The installation of
a modern turbofan engine could meet the need, if it could be accomplished without incurring an
excessive increase in interference drag, weight, or cost. The conventional wisdom on nacelle
installation for avoiding the interference drag would have required lengthening the landing gear to
provide proper ground clearance. However, increasing the landing gear length would result in extra
weight and excessive costs to modify the aircraft structure. The need, the opportunity, and the
technology were available to provide a solution to this challenge. The resultant solution, which
allowed a much larger diameter engine to fit under the wing without increasing the main landing
gear length is illustrated in Figure 2.2.9. The knowledge provided by CFD made possible the very
successful major derivative to the original 737. The impact of CFD here contributed to the most
successful commercial jet
airplane series in history.
During the aerodynamic
design of the Boeing 777 the
risk of significant
interference drag due to the
engine exhaust was revealed
through CFD analysis.
Neither the earlier linear-
based CFD methods nor
conventional wind tunnel Figure 2.2.9 The Installation of a Modern High-Bypass-
testing techniques, which did Ratio Turbofan Engine on the 737
not simulate the exhaust,
 66

would have detected this potential problem. Only a very expensive powered-nacelle testing
technique could assess these interference effects. Three different manufacturer's engines were being
considered for the new aircraft. Using the powered testing technique to develop the engine
installations would have added considerable expense.
Moreover, such a wind tunnel based development would have required too much time and
unacceptable design flow time. Nonlinear transonic TRANAIR analysis made it practical to address
these installation problems, including the effects of the engine exhaust flows in a timely manner. Had
these problems gone undetected until late in the aircraft's development when the powered testing is
usually done, any fixes would have been extremely expensive to implement.
TRANAIR's ability to provide insight to design changes allowed a close "Working Together"
relationship between the various Boeing engineering disciplines and the engine manufacturers. It is
noteworthy, that the exhaust system of all three engines models is very similar in design, a feature
found only on the 777. Key to the success of this application was the ability to model enough of the
relevant physics and to provide solutions quickly enough to support the development schedule. The
impact of CFD here was to provide information facilitating a closer working relationship between
design groups. This enabled detecting problems early in the development process, when fixing or
avoiding them was least expensive.

Figure 2.2.10 CFD Simulation of Installed Engine Exhaust Effects

One deficiency of CFD for engine/airframe integration on the 777 was its inability to compute drag
with sufficient accuracy. Today that challenge is being assaulted. Figure 2.2.11 presents a
comparison of CFD computed drag with experimental results for a 4-engine transport configuration
with flow through nacelles. The drag values are a combination of semi-empirical estimation for the
profile drag of the body and nacelle struts, and TRANAIR calculations of the wing and fan cowl profile
 67

drag, and induced and wave drag for the entire configuration. Computational data and experimental
results are shown for both a wing body alone configuration and the configuration with the addition
of four pylons and flow-through nacelles.
The computational results are in excellent agreement with the wind tunnel data. Also shown is the
effect of nacelle toe and pitch on installed drag compared with experimental data. Again the
agreement is good. As shown in Figure 2.2.10 the code can predict the effects on nearby pressure
distributions due to thrust changes of the engine. What is needed is the ability to accurately compute
drag for powered nacelles as well as flow-through. As further confidence is gained in the reliability
and accuracy of the CFD method significant reduction of propulsion/airframe integration
development time will be possible.

Figure 2.2.11 Engine/Airframe Installation Drag Increments

2.2.10 Future Challenges

Computational fluid dynamics will continue to see an ever-increasing role in the aircraft development
process as long as it can provide value in the end users' eyes. It has improved the quality of
aerodynamic design in the industrial environment, but has yet had much impact on the overall
airplane development process. CFD is becoming more interdisciplinary, helping provide closer ties
between aerodynamics, structures, propulsion, and flight controls. This will be the key to more
concurrent engineering, in which various disciplines will be able to work more in parallel rather than
in the sequential manner as is today's practice. The savings due to reduced development flow time
can be enormous!
To be able to use CFD in these multidisciplinary roles considerable progress in algorithm and
hardware technology is still necessary. CFD is only, at best, qualitatively representative for flows
characterized by large regions of separated flows. Such conditions are encountered with
maneuvering combat aircraft, and with transports at low speed with deployed high lift devices, at
their structural design loads conditions, or subjected to in-flight upsets that expose them to speed
 68

and/or angle of attack conditions outside the envelope of normal flight conditions to name a few.
Such flows can only be characterized by the Navier-Stokes equations. Routine use of CFD based on
the Navier-Stokes formulations will require improved turbulence models and orders of magnitude
improvement in algorithm and hardware performance. Concurrent improvements in the geometry
and grid-generation to handle such complexity as high lift slats and flaps, deployed spoilers, deflected
control surfaces, etc., will also be necessary. However, improvements in CFD alone will not be enough.
The process of aircraft
development, itself, must change to
take advantage of the new CFD
capabilities.

2.2.11 High Lift

Unlike the routine use of CFD for
the design of the cruise
configuration, extensive use CFD
for low speed design of the high lift
elements of an aircraft does not yet
exist [11,21]. It is also the low-
speed/high-lift configuration that
still undergoes significant “tuning”
during pre-certification flight
testing -- an expensive and time
consuming process. Figure 2.2.12 Pressure Distributions for a Multielement
On the Boeing 777 high lift Airfoil
elements were designed using a
combination of linear inviscid three dimensional codes and more complex viscous two dimensional
codes. These were mainly used in the preliminary design phases of the aircraft development, and to
lend understanding to the wind tunnel interference effects during the detail design testing process.
Much of the actual configuration development was done in the wind tunnel. Some useful capability
does exist now for the analysis of 2D multi-element airfoils [22]. Figure 2.2.12 shows a comparison
of Navier-Stokes predicted pressure distributions with experimental data for a three-element airfoil.
A number of Navier-Stokes solutions have been achieved
to date on some very complex high-lift transport
configurations using overset grid methods [23]. The
numerical prediction for one relatively simple
configuration with leading and trailing edge high-lift
devices including flap and slat track fairings, but without
nacelle/pylons, which has been analyzed by a joint
NASA-Boeing team as part of the Advanced Subsonic
Technology (AST) -Integrated Wing Design (IWD)
program, is shown in Figure 2.2.13. This and other
results from these studies have shown the importance of
modeling detailed geometric features such as the wind
tunnel scale slat brackets when comparing to test data.
However as impressive as these solutions have been,
much work is still needed in reducing the flow time to go
from configuration definition to relevant plotted results
by two orders of magnitude! When this reduction is Figure 2.2.13 Surface Streamlines from
achieved, extensive validation will still be needed, as well a 3-D High-Lift Navier-Stokes Solution
as changing the design process to take advantage of this
 69

new capability. When ready, greater use of CFD in the high lift design process will be an enabler to a
significant reduction of the overall aircraft development cycle time.

2.2.12 Buffet
The buffet boundary is an important factor in aircraft design. Buffet occurs when regions of unsteady
flow separation on the wings grow large enough to cause the aircraft to start shaking. The high-speed
buffet boundary is defined as a flight condition that produces a dynamic vertical acceleration of
amplitude 0.1 g (peak-to-peak) at the pilot’s seat track. This boundary in terms of lift coefficient at
buffet onset as a function of Mach number marks the maximum usable lift. Flight operations must
maintain a comfortable margin below the buffet boundary. This means that wing performance must
be assessed in terms of buffet boundary as well as drag. Currently buffet is evaluated only in the wind
tunnel using semiempirical methods, and not always with the greatest accuracy.

Figure 2.2.14 CFD Buffet Prediction - Four-Engine Transport

Progress is being made in the ability to use CFD to determine buffet boundaries. Figure 2.2.14
shows a comparison of predicted vs. experimentally determined buffet boundaries for two four-
engine configurations. Also shown is the incremental difference in the buffet boundary between the
two configurations A Navier-Stokes code, TLNS3D, was used to make the predictions that are quite
reasonable in these cases. Unfortunately, the process is not yet reliable, other attempts using the
same method to predict buffet with CFD have been less successful. The impact of a reliable and timely
CFD buffet prediction process will be as an enabler to pursuing a more aggressive (higher
performance) wing design while protecting against off-design problems including early buffet onset,
and in reducing the number of design/test cycles in an airplane development program.
Aerodynamic loads prediction is one of the more time consuming items in an aircraft development
program. Aerodynamic loads are needed before the detail structural design of the airframe can
proceed. Critical design loads are generally found at the extremes of the flight envelope. These loads
can be subject to significant Reynolds number effects that the wind tunnel may not catch. Cryogenic
wind tunnels, which might have flight Reynolds number capability, are not a viable option due to
model strength problems, the number of test conditions, and cost. Progress is being made in the
ability to use CFD to determine preliminary loads prior to a wind tunnel entry. Figure 2.2.15 shows
a comparison of wing section lift coefficient at several stations along the wing span.
TLNS3D results are in very good agreement with the test data. The number of flight conditions that
must be evaluated even for preliminary loads demands at least two orders-of-magnitude reduction
in CFD cost and flow time before it can be used effectively. The impact of CFD in loads prediction will
 70

be as an enabler to more concurrent engineering reducing development cycle time, and better
accounting of Reynolds number effects to reduce risk of an airplane development program.

Figure 2.2.15 Wing Section Lift Coefficient Comparison TLNS3D vs Wind Tunnel

2.2.13 Stability and Control

Stability and control considerations can eliminate an aerodynamic design that otherwise meets its
mission requirements if they compromise safety, lead to unacceptable handling characteristics, or
are not certifiable. Refining handling characteristics is area of significant “tuning” during pre-
certification flight test. As in the case of loads data, stability and control data in the wind tunnel is
usually acquired after the aerodynamic design is finished which adds to the development time.
Figure 2.2.16 shows a pressure distribution comparison of experimental and CFD results [24]. The
data were taken at a high Mach number at the 70% span station on a transport wing with and without
a deflected spoiler. The low angle of attack case shows the expected effect of decreasing lift due to

Figure 2.2.16 Effect of Spoiler Deflection on Wing Upper Surface Pressure Distribution

the spoiler deflection. The high angle of attack case shows just the opposite, lift increasing with
spoiler defection. The spoiler effectiveness has reversed!
CFD was able to match the experimental results in this case. Like the case for aerodynamic loads, the
 71

number of flight conditions that must be evaluated for determining preliminary stability and control,
and handling characteristics demand at least two orders-of-magnitude reduction in CFD cost and
flow time before it can be used effectively. The impact of CFD on stability and control prediction will
be as an enabler to reducing the number of design/test cycles, and risk in an airplane development
program.

2.2.14 Conclusions
CFD will not be displacing the wind tunnel in the near future. The wind tunnel will continue to
provide validation of CFD-based design and will still be the most cost-effective provider of global
performance numbers, i.e., lift, drag, and airplane stability characteristics. A well-conducted wind
tunnel test can be an extremely efficient and effective way to obtain data beyond today’s reach of
computational methods alone. In a typical wind tunnel test, where data are acquired over a broad
range of flight conditions, the cost per condition (one Mach number, one angle-of-attack) is less
expensive than with CFD. Today, the strength of CFD is not in providing data, but of providing
understanding so necessary for improved design.
The design/optimization role of CFD has no counterpart in the wind tunnel. This has allowed us to
achieve design solutions that are otherwise unobtainable within the allowable time frame of a
development program. CFD is being used to provide corrections to the wind tunnel for effects of
model mounting system, wind tunnel walls, Reynolds number corrections, etc. Further penetration
into the design process and to significantly reduce the aircraft development flow time requires
additional CFD development.
As aircraft development processes change to take further advantage of new CFD capabilities there
will be significant reductions in the development flowtime. Significant advances are also occurring in
wind tunnel model design and manufacturing techniques,25 instrumentation, data acquisition, and
testing productivity. Together, CFD and the wind tunnel provide a better and more cost-effective
solution than either alone. The future of CFD is bright provided it can continue to deliver "bottom
line" value to an airplane development program. Advanced computers and algorithms coupled with
increased design experience and the changing role of CFD will allow further improvements and flow
time reductions in the aircraft development process.
The ever increasing demands for reducing development flow time, simplifying manufacturing design,
reducing cost of ownership, all without loss of performance demand increased use of CFD. If these
necessities are achieved at a loss of performance, you might find that "If you can't make the mission
- You don't have a market." The increased use of CFD in aircraft development is not being driven by
technology for its own sake, but by the competitive marketplace. CFD is now an integral part of the
aerodynamic design process in partnership with the wind tunnel and flight. CFD will be a major
enabler in significant reductions of the aircraft development cycle. There are no alternatives if we
wish to remain competitive!

2.2.15 Acknowledgment
The author wishes to acknowledge the contributions of W.H. Huffman, R.A. Rozendaal, N.J. Yu, M.M.
Curtin, D.P. Witkowski, and members of the Boeing Phantom Works in Long Beach, California. Work
presented here were was supported by various independent sources of the Boeing Commercial Airplane
Group, and by NASA Contract NAS1-20267.

2.2.16 References
1. Rubbert, P. E and Tinoco, E. N., "Impact of Computational Methods on Aircraft Design," AIAA-83-
2060, August 1983.
2. Bengelink, R. L., and Rubbert, P. E., "The impact of CFD on the Airplane Design Process: Today and
Tomorrow," SAE Aerospace Engineering, March 1992
 72

3. Bengelink, R. L., and Purcell, T. W., "The Value of a Computational/Experimental Partnership in

Aerodynamic Design," ICAS 92-4.3.2, 1992.
4. Jameson, Antony, "Re-Engineering the Design Process through Computation," AIAA-97-0641.
5. Rubbert, P. E., "AIAA Wright Brothers Lecture: CFD and the Changing World of Airplane Design,"
ICAS-94-0.2, September 1994.
6. Jameson, Antony, "The Present Status, Challenges, and Future Developments in Computational Fluid
Dynamics," Technical Report, 77th AGARD Fluid Dynamics Panel Symposium, Seville, 1995.
7. Jameson, Antony, “Essential Elements of Computational Algorithms for Aerodynamic Analysis and
Design,” NASA/CR-97-206268, December 1997.
8. Yu, N. J. . Su, T. Y., and Wilkinson, W. M., "Multiblock Grid Generation Process of Complex
Configuration Analysis Using Navier-Stokes Solvers," AIAA-96-1995, June 1996.
9. Rubbert, P. E., “The Use of CFD in Airplane Design,” CFD 97 Fifth Annual Conference of the CFD
Society of Canada, Victoria, B.C., May 1997.
10. Capron, W. K., and Smit, K. L., "Advanced Aerodynamic Applications of an Interactive Geometry
and Visualization System," AIAA-91-0800, January 1991.
11. Nield, B. N., "An Overview of the Boeing 777 High Lift Aerodynamic Design," Aeronautical Journal,
November 1995, pp 361-371.
12. da Costa, A. L., "Application of Computational Aerodynamics Methods to the Design and Analysis
of Transport Aircraft,' ICAS-78.B2-01, September 1978.
13. Rubbert, P. E., and Goldhammer, M. I., "CFD in Design: An Airframe Perspective," AIAA-89-0092,
January 1989.
14. Campbell, R. L., "An Approach to Constrained Aerodynamic Design with Applications to Airfoils,"
NASA TP-3260, November 1992.
15. Agrawal, S.. “Non-Linear Aerodynamic Shape Optimization at Boeing Long Beach,” AIAA-98-
2534-CP, June 1998.
16. Herling, W., LeDoux, S., Ratclift, R., Treiber, D., and Warfield, M., “3DOPT-An Integrated System
for Aerodynamic Design Optimization,” AIAA-2514-CP, June 1998.
17. Jou, W. H., Huffman, W. P., Young, D. P., Melvin, R. G., Bieterman, M. B., Hilmes, C. L., and Johnson,
F. T., "Practical Considerations in Aerodynamic Design Optimization," AIAA-95-1730-CP, June 1995.
18. Young, D. P., Huffman, W. P., Johnson, F. T., Melvin, R. G., and Hilmes, C. L., “Non-Linear
Aerodynamic Shape Optimization at Boeing,” AIAA-98-2535-CP, June 1998.
19. Vatsa, V. N., Sanetrik, M. D., and Parlette, E. B., "Development of a Flexible and Efficient Multigrid-
Based MultiBlock Flow Solver," AIAA-93-0677, January 1993.
20. Tinoco, E. N. and Chen, A. W., "CFD Applications to Engine/Airframe Integration," Numerical
Methods for Engine-Airframe Integration, Vol. 102 of Progress in Astronautics and Aeronautics,
AIAA, 1986.
21. Lynch, F. T., Potter, R. C., and Spaid, F. W., "Requirements for Effective High Lift CFD," ICAS-96-
2.7.1, September 1996.
22. Kusunose, K., and Cao, H. V., "Prediction of Transition Location for a 2-D Navier-Stokes Solver for
Multi-Element Airfoil Configurations," AIAA-94-2376, June 1994.
23. Rogers, S. E., Cao, H. V. and Su, T. Y., "Grid Generation for Complex High-Lift Configurations," AIAA
Paper 98-3011, June 1998
24. Wilkinson, W. M., Lines, T. R., and Yu, N. J., "Navier-Stokes Calculations For Massively Separated
Flows," AIAA-96-2383, June 1996.
25. Gionet, C. and Yandrasits, F., “Developments in Model Design and Manufacturing Techniques,”
The Royal Aeronautical Society, 1997 European Forum, Wind Tunnels and Wind tunnel Test
Techniques, April 1997.
 73

2.3 Case Study 3 - Fuselage Aerodynamic Drag Prediction

Article Information
Authors : Fabrizio Nicolosi, Pierluigi Della Vecchia, Danilo Ciliberti, Vincenzo Cusati
Affiliation : University of Naples Federico II, Via Claudio 21, 80125, Naples,
Title of Paper : Fuselage Aerodynamic Drag Prediction Method By CFD
Citation : (Nicolosi, Vecchia, Ciliberti, & Cusati, 2015)
Bibliography : Nicolosi, F., Vecchia, P. D., Ciliberti, D., & Cusati, V. (2015). Fuselage Aerodynamic
Drag Prediction Method By CFD. CEAS 2015 paper no. 37.

2.3.1 Abstract
The aim of this work is the development of a new methodology to predict fuselage aerodynamic drag
through CFD aerodynamic calculations. The investigation has been focused on typical large
turboprop fuselage geometry. The geometry has been divided into three main components: nose,
cabin, and fuselage tail. Fuselage fineness ratio, windshield angle (Ψ), and upsweep angle (θ), have
been used as independent (geometric) variables to derive the drag prediction methodology. These
parameters have been varied one by one, keeping the others constant. Several fuselage geometries
have been generated and then analyzed with Star-CCM+ in viscous, compressible flow regime. The
effect of a high-wing-fuselage fairing has been also evaluated in terms of fuselage drag, varying the
length and the fairing height. Results present a simple method to estimate the isolated fuselage drag
coefficient and to take into account for a high-wing fairing geometry, typical for a turboprop aircraft.

2.3.2 Introduction
This paper presents new preliminary design methodologies to estimate the drag aerodynamic
coefficient of transport large turbo propeller aircraft fuselage. Method has been developed by
numerical aerodynamic analyses performed with STAR-CCM+®[1] and it has been focused on the
estimation of aerodynamic drag coefficients. Similar numerical approach to develop a methodology
to be applied in preliminary design phase has been already carried out by the authors, which have
deeply investigated the aerodynamics of the vertical tail plane and the aerodynamic interference
among airplane components caused by rudder deflection [2]. The result of these studies is a
methodology which effectiveness is not enclosed only for the turboprop air transport category, but
it has also been exploited for the preliminary design of a new general aviation commuter aircraft [4].
The aerodynamic design of the fuselage of a transport aircraft is a crucial item in airplane preliminary
design. About 30% of zero lift drag is due to the fuselage [8]. Aircraft cruise performance, such as
maximum flight speed or fuel consumption, are mainly dependent from the zero lift drag coefficient
and they could be improved with a more accurate aerodynamic design. Moreover aircraft
longitudinal and directional stability characteristics are strictly related to the fuselage contribution,
thus an accurate estimation of the latter could lead to a better tail plane design and aircraft stability
characteristics.
Aircraft preliminary design usually relies on semi-empirical methodologies, based on heritage
aircraft geometries and wind tunnel tests conducted mainly by NACA [9]. Semi-empirical methods
consider the drag coefficient as the sum of different contributions that can be evaluated by relations
obtained from wind tunnel test data, most of which are collected in the USAF DATCOM database [14].
The total drag coefficient of an aircraft can be expressed as the sum of the zero lift drag coefficient
and the drag-due-to-lift coefficient. This assumption is valid when the approximation of a parabolic
drag polar is used in order to estimate the drag coefficient for low incidence such as cruise and climb,
that is until the lift coefficient becomes greater than 1. The zero lift drag coefficient is also known as
 74

parasite drag coefficient and it includes skin friction (function of wetted area), windshield effect angle
ψ, upsweep effect angle θ, and base drag contributions [14].
Researchers at University of Naples have been working on the development of design techniques for
light and general aviation aircraft since 1996 [19]. Application of such developed or matured
methodologies for aerodynamic design have been previously shown in [20]. The matured experience
have been also applied through deep investigation on the aerodynamic effects of wing tip and
boundary layer control on aircraft performance [21]. The matured know-how in aircraft
aerodynamic design have been also linked to some validation obtained through dedicated flight test
activity [23].

Figure 2.3.1 Main fuselage geometrical parameters

2.3.3 Fuselage Geometries

A modular model of a 80-seats fuselage of a generic regional turboprop aircraft has been considered
as reference layout, which leads to a fuselage length of about 30 m and a fuselage diameter of 3.4 m.
This geometry has a fineness ratio of about 9 and it has been divided into three main components:
nose, cabin, and tail cone (see Figure 2.3.1). For each component, main geometrical parameters
have been defined as shown in Figure 2.3.1, Figure 2.3.2 and summarized in Error! Reference
source not found.. The ratio of the fuselage length and diameter is the fineness ratio FR, whereas
the ratio between the nose length and diameter and tail cone length and the diameter are respectively
the fineness ratio of the nose FRn and of the tail FRt. In order to define the windshield (ψ) and
upsweep (θ) angles, the hw and hu parameters have been introduced. The first locates the height of
the intersection point between the horizontal line and the tangent to nose contour. The latter locates

Figure 2.3.2 Fineness ratio variations, fuselage nose (left), fuselage cabin (middle), fuselage tail (right).
 75

the height of the intersection point between the horizontal line and the tangent to tail contour. Both
are defined in Table 3.4.1.

Table 2.3.1 Definition of geometrical parameters

2.3.3.1 Wing-Fuselage Faring

In order to estimate the wing-fuselage fairing effect on fuselage drag coefficient, a simple straight-
tapered high wing geometry has been considered, joined to the fuselage with a fairing geometry as
suggested in [8] by the authors. This geometry can be ideally divided into two main parts: the forward
part which has a length of 8% of Lf, and an aft part which has a length of about 6% Lf. Starting from
a reference fairing geometry shown in Figure 2.3.3, the latter has been modified by an isomorphism
in length and height (see Figure 2.3.3). This figure also shows the parameters used to define the
stretching ratio of the wing-fuselage fairing. For both the forward and the aft parts, the stretching
factor is defined as the ratio of the length of the modified geometry to the length of the baseline
geometry (LRFWD, LRAFT). The modifications in height are expressed with respect to the reference
height which is equal to 2% Lf measured from the fuselage constant cabin.

Figure 2.3.3 Wing-fuselage fairing design parameters

 76

2.3.4 Mesh and Physics Set-Up

The models investigated in this work are the
fuselage of a large regional transport
turboprop aircraft and the wing-fuselage
with a fairing geometry. The numerical
domain is a typical parallelepiped domain
with model located on the longitudinal
plane of symmetry, at one third of the block
length from the inlet face (Figure 2.3.4).
Polyhedral mesh (built into STAR-CCM+®
[1] has been used with 20 prismatic layers
to better predict the boundary layer
phenomena. Viscous compressible RANS Figure 2.3.4 Computational domain and polyhedral
equations have been solved with Spalart- volume mesh
Allmaras turbulent model. Mesh
independency and solution convergence have been monitored by looking equations residuals
(around to 10 - 7), aerodynamic coefficients stabilization and y+ ≈ o 1 value varying the number of
cells (see Figure 2.3.5). Isolated
fuselage aerodynamic analyses have
been performed with a mesh of about
2.3 M14 of cells for fuselage semi-
model and 6.4 M of cells for the wing-
body-fairing configuration.
All the aerodynamic analyses have
been performed at typical cruise flight
condition with a Mach number equal
to 0.52 and a Reynolds number (based
on the fuselage length) of 2.0 x108 at
zero degrees of angle of attack (alpha
= 0∘, M = 0.52, Re = 2.0 x 108). In order
to evaluate the wing fuselage fairing
effect in a wide range of angle of
attack, a typical climb condition has
been also investigated (alpha = 8∘, M =
0.23, Re = 1.5 x 108). All the mesh and
physics set-up data are summarized in Table 2.3.2 Mesh and physics settings
Table 3.4.2.

2.3.5 Drag Prediction Method

The method allows computing the fuselage drag coefficient as the sum of the contributions of each
component (nose, cabin, and tail cone). This approach does not allow evaluating some sources of drag
as leakage, wiper, surface roughness, and excrescences. The hypothesis of the super-positioning of
the effects has been verified, since the geometry modifications of one part of the fuselage affect only
the drag coefficient of that part [18].
Swet nose Swet cabin Swet tail Swet
CD fus = K n + Kc + Kt CDfp
Swet Swet Swet Sfront
Eq. 2.3.1

14 M=million
 77

where definitions of the main parameters present in the Eq. 2.3.1 are:
• CDfus is the drag coefficient of the fuselage referred to Sfront.
• Kn is the nose shape factor. It depends on windshield angle, ψ, and on the FRn .
• Kc is the cabin shape factor. It depends on the FR.
• Kt is the tail cone shape factor. It depends on upsweep angle, θ, and on the FRt .
• CDfp is the drag coefficient of the equivalent flat plate and it coincides with the skin friction
coefficient, which can be computed from the following :

0.455
CD fp =
(LogRe)2.58 (1 + 0.144M 2 )0.58
Eq. 2.3.2

Figure 2.3.5 Fuselage Drag Coefficient vs Number of Cells and wall y+

2.3.6 Fuselage Nose Factor

The nose shape factor Kn represents the contribution of the nose to the global drag coefficient and it
takes into account the effect of the nose fineness ratio and of the windshield geometric angle ψ. It is
defined in Eq. 2.3.2Error! Reference source not found.. In Error! Reference source not found.
the curves of Kn are drawn as a function of FRn and parameterized with ψ (windshield angle).

CDn Sfront
Kn =
CDfp Swet nose
Eq. 2.3.3
These curves have been obtained from numerical results. For each numerical simulation, the value
of CDn (which is the value of the drag coefficient estimated only on the nose surface and it is referred
to Sfront) has been calculated, along the value of the equivalent skin friction coefficient C Dfp by Eq.
2.3.2, and the wet surfaces for each component been accounted for, then the Kn curves have been
generated with Eq. 2.3.3. The end-user does not have to use the equation to calculate the Kn factor,
but he only has to refer to the chart. As it can be seen in Figure 2.3.6, the Kn factor decreases as FRn
increases and ψ angle decreases, meaning that the fuselage drag coefficient decreases.
 78

Figure 2.3.6 Nose shape factor as a function of nose fineness ratio, α = 0 deg

2.3.7 Fuselage Cabin Factor

The fuselage shape factor Kc represents the contribution of the cabin to the global drag coefficient
and it takes into account the effect of the cabin length (or cabin fineness ratio). It is important to
notice that fuselage FR is primarily dependent from cabin FR. It is defined in Eq. 2.3.4. In Figure
2.3.7 the curve of Kc is drawn as a function of FR. As for the nose shape factor, this curve must be
used to get the value of the shape factor in order to apply the method. Again, the end-user of the
method has to refer to the chart only, whereas the equation states how the Kc curves have been
generated from the results of numerical analyses. Figure 2.3.7 shows that Kc factor decreases with
the fuselage fineness ratio up to FR ≈ 10 and it tends to be constant for higher FR values. This fact
implies that it is not useful to increase the fuselage cabin (and so FR) over FR = 9-10 because of an
increment in wetted surface and hence an increment in parasite drag coefficient (see also Eq. 2.3.1).

Figure 2.3.7 Cabin shape factor as a function of fineness ratio, α = 0 deg

 79

CDc Sfront
Kc =
CDfp Swet cabine
Eq. 2.3.4
2.3.7.1 Fuselage Tail Cone Factor
The tail shape factor Kt represents the contribution of the tail to the global drag coefficient and it
takes into account the effect of the upsweep angle θ. It has been estimated from the CFD value of the
coefficient CDt (value of the drag coefficient estimated only on the tail surface) which is referred to
Sfront. In the same chart, the curve of θmax is traced. This curve is the locus of the maximum possible
value of the upsweep angle for a fixed fineness ratio. The following equation states how the Kt curves
have been calculated.
CDt Sfront
Kt =
CDfp Swet tail
Eq. 2.3.5
In Figure 2.3.8 the curves of Kt are drawn as a function of FRt and parameterized in θ. As for the
other shape factors, these curves must be used to get the value of the tail cone shape factor in order
to apply the method. For a given upsweep angle (see Figure 2.3.8), the longer is the tail, the bigger
is the drag coefficient. This is due to the increased wetted area. Conversely, for a given tail cone
slenderness, the higher is the upsweep angle, the bigger is drag coefficient. In this last case, what is
saved in skin friction (wetted area) is lost in pressure drag.

Figure 2.3.8 Tail shape factor as a function of tail fineness ratio (FR t), α = 0 deg, and scheme of the
geometric relationship between maximum value of upsweep angle (θ) and FR t.

2.3.7.2 Wing-Fuselage Fairing Effects

Wing-fuselage geometry has been joined with a typical high-wing-fuselage fairing as suggested in
Ref. [8] and shown in Figure 2.3.3. The reference wing-fuselage-fairing has been lengthened and
shortened both forward and aft. The drag coefficient results due to fairing length variation are shown
in Figure 2.3.9 and for cruise and climb conditions respectively. These charts depict the difference
in terms of drag coefficient of the reference geometry and the modified ones, where the abscissa
 80

Figure 2.3.10 Difference in drag counts with respect to reference wing-fuselage-fairing for the
configurations analyzed, cruise (left) climb (right).

represents the stretching of the rear part and the different curves represent the stretching of the
forward part. As it can be seen, lengthening the rear part leads to a drag reduction 4 times (4 drag
counts) higher than to the stretching of the forward part (about 1 drag count). This is due to the fact
that, on the forward part the flow accelerates with a favorable pressure gradient (after a first
deceleration in correspondence of the fuselage-fairing junction). Conversely on the aft part of the
fairing-fuselage geometry the flow proceeds with an adverse pressure gradient, slowing down until
separation occurs. In climb condition (see Figure 2.3.9 right) the effect of lengthening of the fairing
rear part is similar to the cruise condition, while the slightly shorter aft panel leads to a higher drag
reduction than the reference one (due to the angle of attack). However the drag reduction effect of
stretching the aft part is better than the stretching of the forward part (4 times higher). Both cruise

Figure 2.3.9 Difference in drag coefficient from the reference wing-fuselage fairing configuration, for
both cruise (left) and climb (right) conditions. LRFWD = 1.50
 81

and climb conditions of Figure 2.3.9 show that for a given fairing forward length, a stretching value
LRAFT ≈2 leads to a minimum value for drag coefficient.
Figure 2.3.10 shows the effect on drag coefficient of lowering wing-fuselage fairing geometry, for a
given forward length ratio LRFWD =1.5 as function of LRAFT. For a given fairing rear part, a 8%, 16% of
geometry lowering leads to a drag reduction of about 3, 6 drag counts respectively. This behavior is
similar both in cruise and
climb condition.
Assuming a wing-fuselage
geometry Figure 2.3.9 and
Figure 2.3.10 can be useful
in order to preliminary
design of a high wing-
fuselage faring geometry. On
a typical turboprop aircraft a
stretching of the aft part is
more suitable than the
stretching of the forward
part. A good compromise
could be a length ratio of the
forward part LRFWD = 1.25
(10% of Lf) and a length ratio
of the rear part LRFWD = 1.9
(11% of Lf). The fairing
Figure 2.3.11 Fuselage geometries used for test cases
height should be the lower
possible according to
configuration.
2.3.7.3 Method Application
In order to verify the methodology proposed, fuselage geometries shown in Figure 2.3.11 have
been analyzed with STAR-CCM+®[1] and compared with the new proposed method. Results are
summarized in Table 3.4.3, compared to typical semi-empirical methodology. Results show a very
good agreement between the CFD analyses and the proposed method, especially on test cases 1, 2,

Table 2.3.3 Comparison of drag coefficient dimensionless on wing surfaces and on fuselage front surface.
The DATCOM values include skin friction, the contributions of upsweep and windshield and the base drag.

and 3, with a difference of about 1 drag count. Test 4 and Test 5 show a drag coefficient difference of
6 and 4 drag counts respectively. This is probably due to windshield and tail cone shapes. It has to
 82

be noted that typical semiempirical method (based on skin friction contribution, upsweep,
windshield, and base drag) usually overpredicts the drag coefficient respect to the CFD value.

2.3.8 Conclusions
CFD-based fuselage aerodynamic drag prediction method has been addressed. The proposed method
is fast to apply and reliable on slender fuselage geometries, also highlighting the effects of
geometrical parameters on drag coefficient. It is the authors’ opinion that these new methods can be
useful in the preliminary design phase of an aircraft. Some applications have shown the effectiveness
of the new approach on typical turboprop transport airplane geometries. The effect of a high-wing-
fuselage fairing variation on fuselage drag coefficient is presented, useful in the design of this
component.

2.3.9 Reference
[1] Star-CCM+, Software Package, Ver. 8.04, CD-adapco, Melville, NY, 2013.
[2] Nicolosi, F., Della Vecchia, P., Ciliberti, D., “An investigation on vertical tail plane contribution to
aircraft side force,” Aerospace Science and Technology, Vol. 28, N. 1, July 2013, pp. 401–416. DOI:
10.1016/j.ast.2012.12.006
[3] Nicolosi, F., Della Vecchia, P., Ciliberti, D., “Aerodynamic interference issues in aircraft directional
control,” ASCE's Journal of Aerospace Engineering, Vol. 28, N. 1, January 2015, ISSN 0893-1321. DOI:
10.1061/(ASCE)AS.1943-5525.0000379.
[4] Nicolosi, F., Della Vecchia, P., Corcione, S., “Design and Aerodynamic Analysis of a Twin-engine
Commuter Aircraft,” Aerospace Science and Technology, Volume 40, January 2015, Pages 1–16.
doi: 10.1016/j.ast.2014.10.008.
[5] Nicolosi F., Della Vecchia P., Corcione S., “Aerodynamic Analysis and Design of a Twin Engine
Commuter Aircraft,” ICAS 2012 Proceedings, Optimage Ltd., UK, September 2012, Vol. 1 pp. 321-332.
Proceedings ISBN 978-0-9565333-1-9.
[6] Nicolosi F., Della Vecchia P., Corcione S., “Commuter Aircraft Aerodynamic Design: Wind-Tunnel
tests and CFD Analysis”, ICAS 2014 Proceedings [CD-ROM], Optimage Ltd., UK, 2014. Proceedings
ISBN 3-932182-80-4.
[7] Nicolosi, F., Corcione, S. and Della Vecchia P., “Commuter Aircraft Aerodynamic Characteristics
through Wind Tunnel Tests”, Aircraft Engineering and Aerospace Technology, ISSN: 0002-2667. doi:
10.1108/AEAT-01-2015-0008.R1.
[8] Della Vecchia, P., and Nicolosi, F., “Aerodynamic guidelines in the design and optimization of new
regional turboprop aircraft,” Aerospace Science and Technology, Vol. 38, October 2014, pp. 88-104.
doi: 10.1016/j.ast.2014.07.018
[9] Munk, M. M., The Aerodynamic forces on airship hulls, NACA TR-184, 1924.
[10] Abbott, I. H., Fuselage drag tests in the variable density wind tunnel: streamline bodies of
revolution, fineness ratio of 5, NACA TN-614, 1939.
[11] Gilruth, R. R. and White, M. D., Analysis and prediction of longitudinal stability of airplanes, NACA
TR-711, 1941.
[12] Multhopp, H., Aerodynamics of the fuselage, NACA TM-1036, 1942.
[13] Draley, E. C., High speed drag test of several fuselage shapes in combination with a wing, NACA
WR-L-542, 1947.
[14] Finck, R. D., “USAF stability and control DATCOM,” McDonnell Douglas Corporation, Rept.
AFWAL-TR-83-3048, Wright-Patterson Air Force Base, OH, 1978.
[15] Roskam, J., Airplane Design Part VI: Preliminary Calculation of Aerodynamic, Thrust and Power
Characteristics, DAR Corporation, Lawrence, KS, 2000.
[16] Torenbeek, E., Synthesis of Subsonic Airplane Design, Delft University Press, Delft, The
Netherlands, 1982.
[17] Perkins, C. D., and Hage, R. E., Airplane Performance Stability and Control, Wiley, NY, 1949.
 83

[18] Nicolosi, F., Della Vecchia, P., Ciliberti, D., and Cusati, V., “Development of new preliminary design
methodologies for regional turboprop aircraft by CFD analyses,” ICAS 2014 Proceedings [CD-ROM],
Optimage Ltd., UK, 2014. Proceedings ISBN 3-932182-80-4
[19] Coiro, D.P., Nicolosi, F. “Design of Low-Speed Aircraft by Numerical and Experimental
Techniques Developed at DPA,” Aircraft Design Journal, Vol. 4, N. 1, Pag1-18, March 2001. ISSN 1369-
8869, doi: 10.1016/S1369-8869(00)00020-3. CEAS 2015 paper no. 37
[20] Coiro, D.P., Nicolosi, F. Grasso, F. "Design and Testing of Multi-Element Airfoil for Short-Takeoff-
and-Landing Ultralight Aircraft,” Journal of Aircraft, Vol. 46, N. 5, Sept.-Oct. 2009, pp. 1795-1807. ISSN
0021-8669. doi: 10.2514/1.43429.
[21] Coiro, D. P., Bellobuono, E.F., Nicolosi, F., Donelli, R., "Improving aircraft Endurance through
turbulent separation control by pulsed blowing, ” Journal of Aircraft, Vol. 45, N. 3, May-June 2008, pp.
990-1001. ISSN 0021-8669. doi: 10.2514/1.33268.
[22] Coiro, D.P., Nicolosi, F., Scherillo, F., Maisto, U. "Improving Hang-Glider Maneuverability using
multiple winglets: a numerical and experimental investigation, ” Journal of Aircraft, Vol. 45, N. 3, May-
June 2008, pp. 981-989. ISSN 0021-8669. doi: 10.2514/1.33265
[23] Nicolosi, F., De Marco, A., Della Vecchia, P., “Flight Tests, Performances and Flight Certification
of a Twin-Engine Light Aircraft,” Journal of Aircraft, Vol. 48, N. 1, January-February 2011, pp. 177-
192. ISSN 0021-8669. doi: 10.2514/1.C031056
[24] Nicolosi, F., De Marco, A., Della Vecchia, P. “Stability, Flying Qualities and Longitudinal Parameter
Estimation of a Twin-Engine CS23 Certified Light Aircraft, ” Aerospace Science and Technology
(Elsevier), Vol. 24, N. 1, January–February 2013, pp. 226-240. ISSN 1270-9638.
doi: 10.1016/j.ast.2011.11.011.
 84

2.4 Case 4 - Conceptual Design of A Business Jet Aircraft

Authors : Jannatun Nawar 1, Nafisa Nawal Probha 2, Adnan Shariar 3, Abdul Wahid 4, Saifur Rahman
Bakaul 5
Affiliations : 1,2,3,4 Student, Department of Aeronautical Engineering, Military Institute of Science &
Technology, Dhaka-1216,
5 Associate Professor , Department of Aeronautical Engineering, Military Institute of

Science & Technology, Dhaka-1216,

Original Appearance : ICMIEE-PI-1403531- International Conference on Mechanical, Industrial and
Energy Engineering 2014 25-26 December, 2014, Khulna, BANGLADESH

2.4.1 Abstract
The modern jet transport is considered as one of the finest integration of technologies. Its economic
success depends on performance, low maintenance costs and high passenger appeal and design plays
a vital role in summing up all these factors. Conceptual design is the first step to design of an aircraft.
In this paper a business jet aircraft is designed to carry 8 passengers and to cover a range of 2000
NM with maximum Mach No of 0.7 and with maximum ceiling of 29,000 ft. The conceptual design
consisted of initial sizing, aerodynamics and performance analysis. Through trade studies and
comparison with other business jet aircrafts a final model of the aircraft was built to achieve the
requirements.
Key Words: Business jet, Conceptual design, Initial Sizing, Aerodynamics, Aircraft performance,
Trade study.

2.4.2 Introduction
Airplane design is an art with scientifically approach. It requires both the intellectual engineering
and sensible assumptions. Aircraft design is actually done to meet certain specifications and
requirements established by potential users or pioneer innovative, new ideas and technology. Now-
a-days business jet aircraft is one of the most popular forms of transport aircraft. A business jet is a
jet aircraft designed to transport passenger and goods. It is also known as biz jet, executive jet or
private jet. Biz jet is used by public bodies, government bodies and armed forces and used to parcel
deliveries, transporting people and evacuation in case of casualties. This paper will illustrate about
the conceptual design of a business jet.
Aircraft design has three distinct phases that are carried out in sequence. In chronological order,
conceptual design, preliminary design and detail design. Conceptual design is the first step of
designing aircraft. The main focus of this paper is to design a business jet transport aircraft
conceptually. In conceptual design the configuration arrangement, size and weight and performance
parameters will be calculated. The first thing required in a conceptual design is the design
requirements which guide and evaluate the development of the overall aircraft configuration
arrangement. In this paper we have tried to design gross weight, wing and tail geometry and some
performance parameters with trade study.
Conceptual design is a series of activities which includes some basic question and answers. What
requirements drive the design, what should it look like, what tradeoffs should be considered, what
technologies should be used and do these requirements sale a viable and salable plane all these sums
up conceptual design in a nutshell. This paper covers every aspects of conceptual design. (see Figure
2.4.1).
 85

Design Technology Concept

Requirements Availibility Sketch

Initial First Guess

Initial Layout
Analysis Sizing

Sizing &
Performance
Optomization
Figure 2.4.1 Block Diagram of Conceptual Design

2.4.3 Design Requirements

Mach Number 0.7
The design requirements are tabulated in Table 3.6.1.
Range 2000 NM
2.4.4 Initial Layout Analysis Maximum Ceiling 29000 ft
The actual design effort usually started with a Payload 8 pax
conceptual sketch. This is a back of napkin drawing Endurance 30 minutes
which gives a rough indication of what the design may Max g +2.5 , -1.2
look like. It first starts with a conceptual sketch.
Table 2.4.1 Design Requirements
2.4.4.1 Concept Sketch
A conceptual sketch is a rough sketch to show how the
future design will look like. In this design a rough handmade sketch was first drawn as an initial
sketch (See Figure 2.4.2).
2.4.4.2 Initial Sizing
The conceptual design is used to estimate aerodynamics and weight fractions by comparison to
previous design. In this paper first we estimated the required total weight and fuel weight to perform
the design mission by sizing process .
2.4.4.3 Sizing from the Conceptual Sketch
Sizing is the most important calculation in aircraft design. It determines size of the aircraft, specially
the take off-weight. Design take off gross weight can be calculated using the following equations, In
this paper a jet for 8 passengers each of approximate 200 lbs, 2 crew members and 40 lbs of luggage
for each passenger is considered. Fuel fraction is calculated based on the mission to be using
approximations of the fuel consumption and aerodynamics.
 86

Wcrew + Wpayload
Wo =
W W
1− e− f
Wo Wo
Eq. 2.4.1

2.4.4.4 Mission Profile

In this paper a simple cruise
mission profile is
considered as most
transport and general
aviation design use this
profile. So there are five
legs including takeoff,
climb, cruise, loiter and
landing. For warm up and
take off, climb and landing
we have taken weight
fractions from the historical
data. (see Figure 2.4.3).
For calculating weight
fractions for cruise and
endurance a Lift-to-Drag
ratio which is a measure of
The design overall
aerodynamic efficacies is Figure 2.4.2 Initial Sketch
needed. Using aspect ratio
of 10.67 and from analyzing
different business jet we
have taken maximum Lift-
to-Drag ratio as 23. Thus
fuel fraction for cruise is
0.995 and for endurance is
0.991is found. An overall
weight fraction of 0.8055is
found and from this, takeoff
gross weight is calculated
Figure 2.4.3 Mission Profile
11035.624 lb. So it is a mid-
sized business jet. Here the
Mission Segment Weight Fraction
empty weight is 6439.16 lb
Warm up and takeoff 0.970
and 2275.2146 lb of fuel is
Climb 0.985
needed. (see Table 3.6.2).
Landing 0.995
2.4.4.5 Engine Selection:
Design Mach no is 0.7 and Table 2.4.2 Historical mission segments weight fractions [1]
so analyzing both historical
data and different jet
aircraft High Bypass Turbofan engine. High BPR turbofan engine is used for high thrust and good fuel
efficiency. Rolls-Royce/MAN Turbo RB193 .
 87

2.4.5 Aerodynamics
Selection
Before designing layout, a
number of parameters
including aerofoil(s), the
Figure 2.4.4 NACA 65-212 Aerofoil
wing and tail geometries,
wing loading as it affects
the cruise speed, takeoff
speed and landing
distances, stall speed,
handling qualities during
all flight phases,
2.4.5.1 Aerofoil
Selection
Camber aerofoil allows
the airflow to remain
attached have to be
calculated. Thus it
Figure 2.4.5 Low wing with dihedral
increases lift and reduces
drag and also it produces
lift at zero angle of attack.
A five digit aerofoil
analyzing different types
of wing used so we used
65-2XX series is selected
where last two digits
define position of
maximum camber. From
calculation we have got Figure 2.4.6 Fowler flap
the position of mean
camber at .12 of mean
chord i.e. 12%. Thus our aerofoil is NACA 65-212. (Figure 2.4.4).
2.4.5.2 Wing Geometries
1. From different business jet and other transport aircraft historical trend the wing loading is
120 is found but considering different aerodynamic facts a wing loading of 80 is taken and so
our wing area becomes 137.94 square feet. The relevant aspect ratio is 10.67 and so wing
span is 38.35 ft.
2. The maximum co-efficient of lift is 1.5. As wing sweep increases lateral stability, a leading
edge sweep of 20 degree is taken. A twist angle of 3 degree to prevent tip stall is taken. Wing
incidence angle is given 2 degree to minimize drag during cruise. As it is a subsonic aircraft
so a positive dihedral of 5 is taken.
3. Virtually all high-speed commercial transport aircraft are low wing so we have chosen a low
wing aircraft as it gives us advantage for landing gear stowage.
4. We have chosen winglet as it increases lift-to drag ratio by 20%. The following figure explains
why winglet is chosen over the conventional wingtip. (Figure 2.4.5).
2.4.5.3 Flap Selection
In this design double slotted fowler flap was used. From calculation, leading edge sweep angle is 20
 88

degree. For flap maximum co-efficient of lift was calculated to be 2.82. (Figure 2.4.6).
2.4.5.4 Stall Characteristics
For Mach no 0.7, aerofoil thickness ratio of 14% is selected. So it is a moderately thick aerofoil where
stall starts from leading edge.
2.4.5.5 Tail Geometries
1. Tails are little wings and it provides for trim, stability and control. Here T-tail is considered
as our tail arrangement not because of it is stylish but it allows the use of engines mounted in
pods on the aft fuselage.
2. Leading edge sweep of the horizontal tail is usually set 5 degree more than the wing sweep
so It is 25 degree. For T-tail no aspect ratio and taper ratio needed for horizontal tail and for
vertical tail taper ratio is 1 and aspect ratio is 0.8.

2.4.6 Landing Gear Arrangement

Today’s most commonly used landing gear arrangement is tricycle gear. So it is used with two main
wheels aft of center of gravity, ahead of the main wheels. As the aircraft is under 50,000 lb (11,036
lb) so two main wheels per strut will be used so we will get advantage in case of flat tire. The tires
carries 90% the total weight of the aircraft so it is an important parameter. The diameter of the main
wheels as 14.52 in and width is 9.12 in is calculated and the recommended tire pressure is 80 psi.

2.4.7 Fuselage Design

For a jet transport, fuselage length is 36.7 ft. Internal cabin system, aerodynamics related calculations
etc are required to estimate Fuselage diameter. For rough estimation, taking circular fuselage,
fineness ratio is 8 and diameter is 4.58 ft.

2.4.8 Vertical Stabilizer

From the historical data taper ratio for vertical stabilizer is 0.6-1.0[2] and the taper ratio considered
is 1.0 in this design. T-tail reduces end plate effect by 5 percent. So tail volume coefficient is calculated
0.0855. From historical data, tail arm for aft mounted engine is about 45-50 % [3] . For a better design
a tail arm should be as big as possible 50 % of fuselage length was taken and it was calculated to be
18.35 ft. Vertical stabilizer surface area was calculated to be 24.66 ft. square feet. Typical aspect ratio
is 0.7-1.2. In this design a aspect ratio of 0.8; a higher aspect ratio was not chosen as increase in
aspect ratio increases span which needs a stronger structure. Thus the weight will be increased so a
lower aspect ratio was chosen.

2.4.9 Horizontal Stabilizer

Leading edge sweep for horizontal stabilizer is 5 degree greater than the sweep angle for wing. So it
was taken as 25 degree. From statistical data tail volume coefficient for horizontal stabilizer is 1.00
but for clean air it was reduced 5% and taken as 0.95. Horizontal stabilizer surface area was
calculates 27.25 square feet.

2.4.10 Performance Analysis

Thrust-to-weight ratio directly associates the performance of the aircraft. It is not constant and it
varies during flight as fuel burns. By analyzing historical data we estimated it for takeoff is 0.35 and
for cruise is .0434[4] and it was calculated using the following equation:

T 1
=
W (L)
D cruse
Eq. 2.4.2
 89

From Oswald efficiency as .8[5] and using aspect ratio we calculated our co-efficient of drag is 0.1344.
For a jet propelled aircraft to maximize range it has to fly at CL0.5/CD and the associated velocity can
be calculated as 655 ft/sec and maximum range is 1975 NM which was calculated. Maximum climb
angle calculated is 17 degree using the following equation:

T
Sinθmax = − √4kCD0
W
Eq. 2.4.3
And for maximizing this angle velocity needed is 310 ft/sec. The maximum rate of climb is,

0.5

2 k W
VR =( √ ( )) = 60 ft/sec
( )
C max ρa 3CD0 S
Eq. 2.4.4
Its associated velocity is 867 ft/sec. Maximum distance covered in a gliding path is 70 NM and
associated velocity is 310 ft/sec. Maximum endurance is calculated to be 1.11 hrs and the velocity
associated is 500 ft/sec. Maximum velocity is 1600 ft/sec which is much more than our required
velocity. Stalling velocity is,

2W ft
Vstall = √ = 212
ρSCL sec
Eq. 2.4.5
Takeoff distance was calculated to be 5100 ft and landing distance is 3445 ft.

2.4.11 Trade Study:

Trade study is an important part of
conceptual design. It helps the
designer to choose the best
convenient design parameters.
From the above trade study it can
be seen that the best thrust to
weight ratio will be 0.40 and wing
loading will be 80. (Table 3.6.3).

2.4.12 V-n Diagram

V-n diagram is a very important
Table 2.4.3 Trade Study
diagram for both the designers and
the pilots. It is actually a graph
showing the limiting factors of design and flying. It shows stall region, corner velocity, maximum
velocity, maximum and minimum load factor etc. (Table 3.6.4 and Figure 2.4.7).

2.4.13 Final Model

After all the calculations and the estimations done a final aircraft model was designed which is shown
in the next page with the help of Solid Works software. (Figure 2.4.8).
 90

2.4.14 Novel Ideas

During the design process some
novel ideas were developed.
These are explained in details
below:
1. T-tail was considered not
because it is stylish but it
allows the use of engines
to be mounted on the aft
fuselage in pods. It’s not
only stylish but it also Table 2.4.4 Different factors of V-n diagram
reduces tail area due to
endplate effect. Though it
is heavier than the
conventional wing it
reduces buffet on the
horizontal tail.
2. Winglet was chosen as the
wing tip as it offers lower
drag. It is both cambered
and twisted so the rotating
vortex flow at the wing tip
creates a forward lift
component which reduces
total wing drag.
3. As it is a small aircraft so
aft-engine arrangement
was used in this design to
Figure 2.4.7 V-n diagram
maintain adequate wing
nacelle and nacelle-
ground clearances. As
there is no wing-pylon
interference so less drag is
created. Less asymmetric
yaw after engine failure
with engine close to
fuselage and lower
fuselage height are also
the advantages of using
aft- engine arrangement.

2.4.15 Comparison with Similar

Aircraft
A comparison is done between the
designed aircraft and HondaHA-
420 Jet aircraft. Both the aircraft
has almost same wing span and
same passenger number. But with Table 2.4.5 Comparison with Similar Aircraft
almost double range the weight
 91

increased only 2000 lb. It is a remarkable success in design. (Table 3.6.5).

Isometric View Front View

Top View Side View

Figure 2.4.8 Final Aircraft Design

2.4.16 Conclusion
In comparison with other business jet aircraft the total weight of the aircraft that was calculated is a
compatible one. Maximum range calculated is almost close to the requirement assumed but the
maximum velocity calculated exceeds the required velocity. Comparing with statistical data the
takeoff distance and the landing distance calculated is a good one. The aero foil chosen is a compatible
one with the design requirement. Some during the design process some requirements were fulfilled
but some were not. But still the calculation of total weight was remarkable one with a higher range
thus this design is totally fuel efficient and economical.
Nomenclature
Wc : Total gross Weight, lb CD0 : Zero Lift Drag coefficient
Wf : Fuel weight, lb V : Velocity, m/s
We : Empty weight, lb R/C : Rate of climb, m/s
T : Thrust, lb ῤ : Density, kg/m3
L : Lift, lb W/S : Wing loading
D : Drag, lb CL : Lift coefficient
ϴ : Climb angle, degree n : Load factor
 92

Vne : Velocity Never Exceed

2.4.17 References
[1] Daniel P. Raymer, Aircraft Design: A Conceptual Approach, Fourth Edition, AIAA, page 21
[2] Daniel P. Raymer, Aircraft Design: A Conceptual Approach, Fourth Edition, AIAA, page 84
[3] Daniel P. Raymer, Aircraft Design: A Conceptual Approach, Fourth Edition, AIAA, page 85
[4] Daniel P. Raymer, Aircraft Design: A Conceptual Approach, Fourth Edition, AIAA, page 99
[5] John D. Anderson Jr. Aircraft Performance and Design, Tata-McGraw-Hill edition
 93

3 Inverse Design Case Studies for Conceptual Aircraft

3.1 Case Study 1 - An Inverse Design Method for Airfoils Based on Pressure
Gradient Distribution
Authors : Yufei Zhang , Chongyang Yan and Haixin Chen
Title : An Inverse Design Method for Airfoils Based on Pressure Gradient Distribution
Affiliation : School of Aerospace Engineering, Tsinghua University, Beijing 100084, China
Original Appearance: Energies 2020, MDPI
Source: Energies 2020, 13, 3400; doi:10.3390/en13133400
Citation : Zhang, Y., Yan, C., & Chen, H. (2020). An Inverse Design Method for Airfoils Based on Pressure
Gradient Distribution. Energies, 13(13), 3400. doi:10.3390/en13133400
An airfoil inverse design method is proposed by using the pressure gradient distribution as the
design target. The adjoint method is used to compute the derivatives of the design target. A
combination of the weighted drag coefficient and the target dimensionless pressure gradient is
applied as the optimization objective, while the lift coefficient is considered as a constraint. The
advantage of this method is that the designer can sketch a rough expectation of the pressure
distribution pattern rather than a precise pressure coefficient under a certain lift coefficient and
Mach number, which can greatly reduce the design iteration in the initial stage of the design process.
Multiple solutions can be obtained under different objective weights. The feasibility of the method is
validated by a supercritical airfoil and a supercritical natural laminar flow airfoil, which are designed
based on the target pressure gradients on the airfoils. Eight supercritical airfoils are designed under
different upper surface pressure gradients. The drag creep and drag divergence characteristics of the
airfoils are numerically tested. The shock free airfoil demonstrates poor performance because of a
high suction peak and the double‐shock phenomenon. The adverse pressure gradient on the upper
surface before the shockwave needs to be less than 0.2 to maintain both good drag creep and drag
divergence characteristics.

3.1.1 Introduction
An airfoil is a basic element of wing, wind turbine, and turbomachinery. The characteristics of the
airfoil strongly affect the performance of the aerodynamic parts, which influence on the energy cost
of the machine. There are many types of airfoils, such as low‐speed airfoils [1], supercritical airfoils
[2], and natural laminar flow airfoils [3,4]. Different types of airfoils have different aerodynamic
characteristics, which are closely related to the pressure distribution. The typical patterns of an
airfoil can be described by pressure distribution features. More specifically, the patterns can be
numerically expressed by pressure gradient values. For instance, a supercritical airfoil may have a
suction plateau with a weak adverse pressure gradient on the upper surface to suppress the
shockwave strength under cruise conditions [5]; a natural laminar flow airfoil should have a
favorable pressure gradient on both the upper and lower surfaces to suppress the increase in the
Tollmien–Schlichting wave and delay the transition from laminar to turbulent flow [6]. The
dimensionless adverse pressure gradient dcp/d(x/c) (c is the chord length) near the trailing edge of
an airfoil should be less than 3.0 to avoid flow separation [7]. Consequently, the pressure gradient is
more convenient to describe the airfoil characteristics than the absolute value of the pressure
coefficient.
In the airfoil design process, a direct problem computes the aerodynamic coefficients such as the lift,
drag, and pressure coefficients, by the computational fluid dynamics (CFD) method; by contrast, an
inverse design method obtains the airfoil geometry by giving a design expectation. The inviscid
velocity distribution [8,9] or pressure distribution [10,11] on the airfoil are usually used as the design
 94

target. Inverse methods based on the potential equation [12] or Navier–Stokes equation [13] have
been studied by many researchers. Numerical optimization methods are commonly used on inverse
design problems. Newly developed genetic algorithms [14], evolutionary algorithms [15], proper
orthogonal decomposition [16], deep learning methods [17], and deep convolutional neural
networks [18] are applied for airfoil inverse design. However, the methods have low design efficiency
because of the low search speed of evolutionary‐type algorithms and the training process of machine
learning type methods. The adjoint method [19], which was developed by Jameson [20] thirty years
ago, solves the derivatives of the design variable.
This method has good design efficiency and fast convergence characteristics for the airfoil inverse
design problem [21,22]. However, it is complicated to introduce real‐life constraints and the design
experience into the optimization process. One of the most important obstacles for the engineering
applicability of the adjoint method is the numerical expression of the design expectations into the
optimization process and the provision of different candidate designs for decision‐making. Another
obstacle is whether the adjoint equation is relevant to the target function or constraint. Each target
function or design constraint has a corresponding adjoint equation and the computational code has
to be re‐programmed. Moreover, in the airfoil design of a modern aircraft, for example, the design
expectation is implicitly involved in the pressure distribution patterns. However, even an
experienced designer cannot set a proper pressure distribution under a certain design lift coefficient
and Mach number. Moreover, sometimes the target pressure distribution applied by a designer is
unfeasible, i.e., there is no airfoil geometry that can coincide with the target pressure distribution.
Consequently, the design target of pressure distribution might be an excessive demand for airfoil
design. This is a drawback of the traditional inverse design method. Zhao et al [5] applied pressure
gradient constraints to design a supercritical airfoil. Zhang et al [6] adopted pressure gradient
constraints to design a supercritical natural laminar flow airfoil. They expressed the experience of
the designer through a series of pressure gradient constraints and applied in the optimization
process. Li et al [23] proposed the pressure distribution patterns of a typical supercritical airfoil,
which were mostly described by pressure gradient features.
In this paper, a new inverse design method based on the target pressure gradient is developed and
applied to airfoil design. The method was inspired by the pressure gradient constrained method [5,6],
which focused on a rough design expectation of aerodynamic designers. The patterns of different
types of airfoils are characterized by several pressure gradients rather than the absolute value of the
pressure coefficient. This method relieves the excessive requirement of the target pressure
distribution, which is hard to apply in the initial stage of airfoil design. For example, the pressure
coefficient values are quite different for airfoils optimized under different Mach numbers.
However, the classical patterns of a supercritical airfoil are the same, including a suction plateau,
frontal loading, aft loading, etc. These patterns can be expressed by a similar dimensionless pressure
gradient distribution, although the absolute values of the pressure coefficients are different at
different Mach numbers or different lift coefficients. The discrete adjoint method [21] is used to
compute the derivatives of the target function. To the authors’ knowledge, this is the first use of the
pressure gradient in the inverse design based on the adjoint method. This method can provide
multiple candidate designs under different weights of pressure gradient distribution and drag
coefficient. The method shows good feasibility on the test cases of a supercritical‐type airfoil and a
natural laminar flow‐type airfoil. Several supercritical airfoils with different upper surface pressure
gradients are designed and numerically tested. The results illustrate that the upper surface pressure
gradient has a strong influence on the drag creep and drag divergence performance of an airfoil. An
appropriate pressure gradient before the shockwave of the upper surface is proposed for a
supercritical airfoil design.
 95

3.1.2 Numerical Methods

3.1.2.1 Computational Method
The present computation was based on a structural CFD (computational fluid dynamics) code ADflow
[24]. It applied the scalar Jameson–Schmidt–Turkel (JST) scheme [25] for spatial discretization. A
diagonalized‐diagonally dominant alternating direction implicit (D3ADI) algorithm, an approximate
Newton–Krylov (ANK) algorithm,
and a full Newton–Krylov (NK)
algorithm [24] were all integrated
in the code. To increase the
convergence speed, the code first
applied D3ADI to obtain an
approximate converged flow field,
then it used ANK and NK solvers to
accelerate the convergence. The
original version of the Spalart–
Allmaras (SA) turbulence model
[26] was used for the closure of
the Reynolds‐averaged Navier–
Stokes equation. The turbulence
Figure 3.1.1 Computational grid of the RAE2822 airfoil (257 ×
production term of the SA model
129 points)
was based on the strain rate of the
flow. The turbulence viscosity of
the free‐stream boundary condition was 0.009 times the molecular viscosity.

Table 3.1.1 Force Coefficient Comparisons Of The RAE2822 Airfoil

In the present computation, the airfoil was computed based on an O‐type grid. The grid was
automatically generated by a hyperbolic mesh matching method [22]. Figure 3.1.1 shows the
medium grid of the RAE2822 airfoil, which has 257 grid points in the circumferential direction and
129 points in the wall normal direction. The far‐field was located 80 c (c is the chord length) away
from the wall.
The first off‐wall layer height was 3 × 10-6 c to ensure that the first layer Δy+ was less than 1.0. A fine
grid and a coarse grid were generated to test the grid convergence. The fine grid had 369 × 185 points
in the circumferential direction and wall normal direction, respectively, while the coarse grid had
185× 97 points. The RAE2822 supercritical airfoil [27] was used as the test case of the code. The
Mach numbers were 0.725 and 0.730 for case 6 and case 9, respectively [27]. The Reynolds number
was 6.5 × 106. Table 4.1.1 shows the aerodynamic force coefficients of the airfoil. In the CFD
computation, the angle of attack (AOA) was adjusted to achieve the same lift coefficient as that in the
 96

experiment. The computed AOA had a deviation compared with the experiment, which is commonly
seen in CFD computation. The
drag coefficient errors of the fine
grid and experiment were 4.4%
and 4.0% for case 6 and case 9, a
respectively. The drag
coefficients of the coarse,
medium, and fine grids had a
clear convergence tendency. The
drag error of the medium and fine
grid was less than 1%. Figure
3.1.2 shows the pressure
coefficient comparisons of the
two cases. The pressure
distributions matched well with
the experiment. A minor flaw was
that the shockwave location of
case 6 was slightly upstream of b
the experiment. The results of the
drag coefficient and pressure
distribution demonstrate that the
present computation method was
satisfactory. The medium grid
was chosen as the baseline grid in
the following design process.
3.1.2.2 Adjoint Method
The inverse design method was
based on a discrete adjoint Figure 3.1.2 Pressure coefficient comparisons for the RAE2822
method [24]. In the framework of airfoil. (a) Case 6: Ma = 0.725, Re = 6.5 × 106, CL = 0.743; (b) Case
9: Ma = 0.730, Re = 6.5 × 106, CL = 0.803
the discrete adjoint method, the
Reynolds‐averaged Navier–
Stokes equation can be expressed as Eq. 3.1.1. 𝒘 is the state vector, which contains 𝑛 elements
including all the flow quantities and turbulence quantities at every computational grid cell. 𝒙 is the
design variable vector. This vector has 𝑛฀ variables for a given optimization problem. 𝑹 is the
residual vector of the equations, which has 𝑛฀ elements relevant to 𝒘. The optimization target 𝑓 is a
relation of the design variable and state vector, as Eq. 3.1.2.

𝐑(𝐱, 𝐰) = 0
Eq. 3.1.1
f = f(𝐱, 𝐰)
Eq. 3.1.2
The adjoint method is a gradient method that solves the derivatives of the target function df/dx. The
derivatives can also be calculated by the finite difference method by disturbing the elements of 𝒙 one
by one with an approximate relation df/dx ≈ ∆f/∆x. However, this is very time consuming if the
design variable number is large. The adjoint method uses an additional adjoint equation to compute
the total derivative of the target function df/d𝒙. The derivative of the residual vector 𝑹 is shown
 97

d𝐑 ∂𝐑 ∂𝐑 d𝐰
= + =0
d𝐱 ∂𝐱 ∂𝐰 d𝐱
Eq. 3.1.3
Then,
d𝐰 ∂𝐑−𝟏 ∂𝐑
=
d𝐱 ∂𝐰 ∂𝐱
Eq. 3.1.4
We constructed an adjoint vector
𝐓
∂𝐑−𝟏 ∂𝐟
𝛙 =
∂𝐰 ∂𝐰
Eq. 3.1.5
The “T” in the superscript means the transposition of the vector. Then, the total derivative of the
target function ∂f/∂𝒙 can be expressed as :

df ∂f ∂f d𝐰
= +
d𝐱 ∂𝐱 ∂𝐰 d𝐱

∂f ∂f ∂𝐑−𝟏 d𝐑
= +
∂𝐱 ∂𝐰 ∂𝐰 d𝐱

∂f ∂𝐑
= + 𝛙𝑇
∂𝐱 ∂𝐱
Eq. 3.1.6
The adjoint vector 𝝍T is controlled by the adjoint equations, as shown in :

∂𝐑𝐓 ∂f T
𝛙=
∂𝐰 ∂𝐰
Eq. 3.1.7
If the optimization problem had multiple target or constraint functions, each target or constraint had
an adjoint equation set and was solved separately. The partial derivatives in the code were computed
by the algorithmic differentiation (AD) method [28]. Reverse‐mode AD was applied for the target
derivative computation in the present computation.
In this paper, the dimensionless pressure gradient on the airfoil was applied as a target function of
the inverse design. The coordinates were normalized by the airfoil chord length, and the pressure
was normalized as the pressure coefficient. In the following text, the pressure gradient refers the
dimensionless pressure gradient. The pressure gradient along the streamwise direction is
approximately computed by
Cp2 − Cp1
Cp,x = x x
2
− 1
c c
Eq. 3.1.8
Cp1 and Cp2 are pressure coefficients at two adjacent cell centers on the airfoil surface, i.e., each
pressure gradient was decided by two grid cells on the airfoil surface. At some locations, the absolute
values of the pressure gradient were very large, for example, the leading edge or the shockwave
 98

location. The definition of Cp,x might be not applicable if the x coordinates of the two cell centers were
the same. However, we only needed to specify the most important pressure gradient patterns at some
points on the airfoil, and the other locations could be freely changed. Consequently, the leading edge
cells with the same x‐coordinates can be excluded in the target function. In the design process, the
derivative of the pressure gradient
dCp,x
d𝐱
Eq. 3.1.9
must be computed. The derivative code was derived with the help of the Tapenade automatic
differentiation tool [28].
3.1.2.3 Optimization Method And Design Variable
After the gradient of the target function was calculated by the adjoint method, the sequential
quadratic programming algorithm [29] was used to search the optimal solution. The design variable
of the present paper was based on
the free‐form deformation (FFD)
[30,31] method. Figure 3.1.3
shows the FFD frame of the present
computation. Twenty FFD points
were used to control the
deformation of the airfoil, as shown
by the red symbols in Figure3. Only
the y‐coordinates of the FFD control
points were used as the design Figure 3.1.3 Free‐form deformation (FFD) frame of the airfoil
optimization
variables. In the optimization, the
leading edge and trailing edge of the
airfoil were fixed. This was done by constraining the FFD control points to move oppositely with the
same step size on the upper and lower sides at x/c = 0.0 and x/c = 1.0. Consequently, the design
variable number was 18 for the airfoil. The y‐coordinates of the FFD control points could be moved
in the range of [–0.1, 0.1], which was large enough for generating different types of airfoils.
3.1.2.4 Inverse Design Based on The Target Pressure Gradient
In this section, airfoil design was carried out based on the target pressure gradient. Two types of
airfoils, including a supercritical‐type airfoil [5,32] and a supercritical natural laminar‐type airfoil
[6,32,33], were designed based on the adjoint method of the target pressure gradient. The summation
of the weighted drag coefficient and the deviation in the pressure gradient target was used as the
design objective.
3.1.2.5 Supercritical Airfoil Design
A supercritical type pressure distribution was used as the design target for the first test case. A
pressure distribution was manually drawn by the tool Javafoil [34], which is a well‐known low‐speed
airfoil design tool. It does not have the ability to design a supercritical airfoil. However, it is a useful
tool to sketch a typical supercritical type pressure distribution to represent the design expectation,
as shown by the black dashed line in Figure 3.1.4.
If we directly used the pressure coefficient as the target for inverse design, we would not know
whether the pressure distribution was feasible for a smooth airfoil; even if it was feasible, we could
not make sure it had good aerodynamic performance. Consequently, the pressure gradients at twenty
locations were used as the design target, as shown by the red diamonds in Figure 3.1.4, including
ten points on the upper surface and ten points on the lower surface. These pressure gradients roughly
characterize a supercritical type pressure distribution. The values of the pressure gradients, Cp,x, at
 99

the twenty locations are also labeled in the figure. Positive values indicate an adverse pressure
gradient, while negative values represent a favorable pressure gradient. The red segments in the
figure indicate the slopes of the pressure with the labeled pressure gradient. The pressure gradient
near the trailing edge was less than 3.0 to avoid flow separation. The pressure gradient on the front
half of the upper surface had a weak adverse gradient to weaken the shockwave. The pressure
gradient near x/c = 0.5 was not specified because we do not know which location was the best
shockwave location for minimizing the drag coefficient. Consequently, the shockwave could be freely
moved at 0.41 < x/c < 0.59 to obtain a minimum drag coefficient design.

Figure 3.1.4 Pressure Gradient Target of A Supercritical Type Airfoil

It was obvious that the optimization was an ill‐posed problem if the objective only had a pressure
gradient because we could obtain countless solutions that satisfied the pressure gradient
expectation. In the present computation, the lift and drag coefficients were also included in the
inverse problem. This ensured that the candidate design not only satisfied the pressure gradient
expectation but also optimized aerodynamic performance with a lift coefficient constraint. The flow
conditions in this case were Ma = 0.73 and Re = 6.5 × 106. The lift coefficient CL was constrained in
the range of 0.80 ± 0.01, and the airfoil internal volume was no less than 0.064 to prevent the airfoil
from becoming too thin. The thickness distribution was allowed to change slightly because the
geometries of different types of airfoils are quite different. The thickness of the airfoil in the present
study was not strictly constrained. The relative thickness at x/c = 0.40 of the airfoil was constrained
to be no less than 0.98 times of the initial airfoil. The optimization problem of this case is defined as
The Cp,x, target is the target pressure distribution of the twenty prescribed locations. ω is a parameter
that adjusts the weight between the drag coefficient and the target pressure gradient.
20
2
min CD + ω ∑(Cp,x − Cp,x,target )
i
i=1
s. t. CL = 0.8 ± 0.01 , Volume ≥ 0.064
Eq. 3.1.10
 100

A NACA0012 airfoil was used as the initial design. Two cases with ω = 0.001 and 0.002 were
calculated for this case. The computation obtained different designs with different weights of the
objective. Figure 3.1.5 shows the convergence histories of the two optimizations. More than 300
iterations were computed for both
cases. The optimizations converge
after 150 iterations. Two optimal
airfoils were obtained from the
computation.
Table 4.1.2 shows the
aerodynamic coefficients of the
airfoils. It is noteworthy that the
angle of attack α was treated as a
design variable, which is similar to
the geometry variables. The
derivatives of α were computed
during the optimization. For a
certain airfoil shape, the angle of
Figure 3.1.5 Convergence histories of the optimizations with
attack was not changed to satisfy different weights
the design lift coefficient. The initial
angle of attack was 2.0 degrees for
the NACA0012 airfoil, so the initial lift coefficient was 0.359. The pressure gradient deviations

20
2
∑(Cp,x − Cp,x,target )
i
i=1
Eq. 3.1.11

Table 3.1.2 Aerodynamic coefficients of the airfoils (Ma = 0.73, Re = 6.5 × 106).

for both cases all converge to less than 2.0. The lift coefficients and volumes all fulfill the design
constraints. The drag coefficients were greatly reduced, and the lift‐to‐drag ratios were increased
considerably. Because the weights were different, the drag coefficients of the two optimized airfoils
were slightly different.
 101

Figure 3.1.6 Pressure and pressure gradient distributions of the airfoils. (a) Pressure distribution; (b)
pressure gradient of the upper surface; and (c) pressure gradient of the lower surface.

Figure 3.1.6 shows the pressure and pressure gradient distributions of the initial and optimized
airfoils. The target pressure gradient in Figure 3.1.6 a is demonstrated by red segments with
diamonds. It can be seen that both optimized airfoils have satisfactory pressure gradient tendencies
compared with the design target. Figure 3.1.6 b-c further presents the pressure gradient
distributions of the upper and lower surfaces. It can be seen that the original airfoil had a quite
different pressure gradient distribution. Because the leading edge had a large pressure gradient, the
optimal designs and the target deviated slightly at the leading edge. In addition to the leading edge,
the two optimized airfoils satisfied the target pressure gradient well. This demonstrates that the
present inverse design method for the pressure gradient was effective.
In the present computation, the region at 0.41 < x/c < 0.59 on the upper surface had no design target
because this region was where the shockwave is supposed to reside. The shockwave of a supercritical
airfoil is very important for its transonic performance because the shockwave greatly influences the
drag divergence characteristics. Aerodynamic designers usually need to design different airfoils or
wings with different shockwave patterns and strengths to balance the trade‐off between cruise
efficiency and drag divergence performance [32].
Figure 3.1.7 shows the Mach number contours of the initial airfoil and the optimized airfoils. The
initial airfoil had a strong shockwave and a large drag coefficient. The airfoil optimized with ω = 0.002
showed a weak shockwave at x/c = 0.50, and the drag coefficient was slightly higher than ω = 0.001
because the drag coefficient had a relatively lower weight in the optimization, while the ω = 0.001
case obtained a shockless airfoil, and its drag coefficient was the lowest. The present method was
able to provide different airfoils with satisfactory pressure gradient patterns, while the shockwave
 102

characteristics were different for decision‐making. The airfoil performance with different
shockwaves is discussed.

Figure 3.1.7 Mach number contours of the initial airfoil and the optimized airfoils. (a) Initial airfoil; (b)
optimized airfoil with ω = 0.002; and (c) optimized airfoil with ω = 0.001

3.1.2.6 Supercritical Natural Laminar Flow Airfoil Design

In this subsection, a supercritical natural laminar flow airfoil (NLF) was optimized by the inverse
method based on the pressure gradient target. Both the upper and lower surfaces of the airfoil had a
favorable pressure gradient for suppressing the Tollmien–Schlichting wave and delaying the
transition from laminar to turbulent flow [6].
The flow conditions were Ma = 0.73 and Re = 6.5 × 106, and the lift coefficient was constrained in the
range of 0.6 ± 0.005. The initial airfoil was the RAE2822 supercritical airfoil, and the initial AOA was
1.5 degrees. Because the pressure gradient at different locations had quite different values, the
relative error of the actual pressure gradient Cp,x and target pressure gradient Cp,x,target was set as the
design objective, as shown in Error! Reference source not found.. The pressure gradients at 27
locations were set as the design targets, including 18 on the upper surface and 9 on the lower surface.

27
Cp,x
min CD + ω ∑ | − 1|
Cp,x,target
i=1 𝑖
s. t. CL = 0.6 ± 0.005 , Volume ≥ 0.064
t t
|x ≥ |x
c =0.4 c =0.4 initial
c c
Eq. 3.1.12
 103

Two cases with different weights,

ω = 0.02 and 0.04, were calculated.
Figure 3.1.12 shows the histories
of the optimizations. The objective
descended very quickly in the first
100 iterations and then gradually
converges. Two optimal designs
were obtained through the
optimization. The aerodynamic
coefficients are shown in Table
4.1.3. It should be noted that the
computation was carried out with
the full turbulence SA model.
Therefore, the drag coefficient was Figure 3.1.8 Optimization histories of the supercritical natural
slightly increased to satisfy the laminar flow (NLF) airfoil
pressure gradient requirement.
The results of the transition computation are presented in the following.
Figure 3.1.9 shows the airfoil geometry, pressure, and pressure gradient distributions of the initial
and optimized airfoils. It is clear in Error! Reference source not found.a that the maximum
thickness locations of the optimized airfoils were moved backward to obtain a favorable pressure
gradient. The design expectation of the pressure gradient at the different locations is marked by the
red segments with diamonds in Figure 3.1.9. On the upper surface, the target pressure gradient
was Cp,x = ‐ 0.50 at 0.15 ≤ x/c ≤ 0.52, and the pressure gradient when x/c < 0.15 was even smaller than
‐0.50. The design target was changed from a favorable pressure gradient at x/c = 0.52 to an adverse

Table 3.1.3 Aerodynamic Performance of the Airfoils (Ma = 0.73, Re = 6.5 × 106).

Figure 3.1.9 Airfoil, pressure and pressure gradient distributions of the supercritical NLF airfoil
optimization. (a) Airfoil geometry; (b) pressure distribution; (c) pressure gradient of the upper
surface; and (d) Pressure gradient of the lower surface
 104

pressure gradient at
x/c = 0.63. The region
of 0.52 < x/c < 0.63 was
supposed to place the
weak shockwave of the
supercritical NLF
airfoil. On the lower
surface, the target
pressure gradient was
Cp,x = ‐0.50 at 0.20 ≤ x/c
≤ 0.56 and was smaller
than ‐0.50 when x/c <
0.20. The design target
was changed from a
favorable pressure
gradient at x/c = 0.56 to Figure 3.1.10 Friction Coefficients of The Initial And Optimized Airfoils
an adverse pressure
gradient at x/c = 0.80.
The pressure recovery region of 0.56 ≤ x/c ≤ 0.80 was not constrained to explore more airfoil shapes.
The pressure distributions of the two optimized airfoils were very similar. The airfoil optimized with
ω = 0.02 had a slightly stronger shock wave and a higher drag coefficient. It can also be seen from
Figure 3.1.9 c ,d that the target pressure gradients ooth the upper and lower surfaces were achieved
after the optimization, except there was minor deviation near the leading edge of the upper surface.
Once again, the capability of the new inverse method was demonstrated to be effective by this
supercritical NLF airfoil optimization case.

Figure 3.1.11 Eddy viscosity contours of the airfoils. (a) Initial airfoil; (b) optimized with ω = 0.02; and
(c) optimized with ω = 0.04
 105

The airfoils were computed by the SST k‐ω‐γ‐Reθ transition model [35] to validate the efficiency of
the NLF design. The freestream turbulence intensity was 0.01%. The aerodynamic coefficients are
listed in Table
4.1.4. It is clear
that the drag
coefficients were
reduced by the
transition
computation.
Figure 3.1.10
shows the friction
coefficient Table 3.1.4 Transition computations of the airfoils (Ma = 0.73, Re = 6.5 × 106).
distributions of
the initial and
optimized airfoils. The transition locations of the optimized airfoils were delayed by approximately
10–15% by the optimization of both the upper and lower surfaces. Figure 3.1.11 shows the eddy
viscosity 𝜇t/𝜇∞ contours of the three airfoils. The eddy viscosity below 0.1 was blanked, and that of
the y‐coordinate was amplified. It can be seen that the location of the high eddy viscosity moved back.
Consequently, the inverse design method with the favorable pressure gradient target had the
capability of designing a supercritical natural laminar flow airfoil.
3.1.2.7 Influence of Pressure Gradient on Supercritical Airfoil Performance
In this section, supercritical airfoils with different upper surface pressure gradients were designed
based on the inverse method. Then, the airfoils were numerically tested to study the drag creep and
drag divergence behaviors.
3.1.2.8 Target Pressure Gradient of Supercritical Airfoil
The design targets of different supercritical airfoils were based on the pressure gradient distribution.
The flow conditions were
Ma = 0.73 and Re = 6.5 ×
106, and the lift
coefficient CL was 0.8 ±
0.005. Pressure gradients
at 28 points on the airfoil
were used as the design
target, including 18
points on the upper
surface and 10 points on
the lower surface. The
RAE2822 airfoil was used
as the initial design. The
pressure gradient target
was modified from the
RAE2822 airfoil. As Figure 3.1.12 Pressure gradient target of supercritical airfoil
shown in Figure 3.1.12,
the pressure gradient on
the lower surface was the same as the RAE2822 airfoil, and the upper surface was adjusted.
The pressure gradient at the front half of the upper surface was changed to an adverse pressure
gradient or a very small favorable pressure gradient to suppress the shockwave strength. The
adverse pressure gradient of the second half was also slightly strengthened to provide enough lift
 106

coefficient. The pressure gradient in the region 0.10 ≤ x/c ≤ 0.45 was adjustable in the optimization
to design different airfoils. Optimizations with different pressure gradients were carried out. The
design problem is defined as Figure 3.1.13. The weight is ω = 0.001.

28
Cp,x
min CD + ω ∑ | − 1|
Cp,x,target
i=1 𝑖
s. t. CL = 0.8 ± 0.005 , Volume ≥ 0.064
t t
| ≥ |x
c x=0.4 c =0.4 initial
c c
Eq. 3.1.13
3.1.2.9 Characteristics of Airfoils with Different Upper Surface Pressure Gradients
Eight airfoils with different pressure gradients on the upper surface were designed by the method.
Table 4.1.5 presents the aerodynamic coefficients of the airfoils. Two more airfoils are also listed in
the table. One is the initial RAE2822 airfoil, and the other is an airfoil optimized by minimizing the
drag coefficient without a pressure gradient target. The drag coefficient of the initial airfoil was the
largest, and the airfoil optimized by minimizing the drag coefficient had the lowest drag. The drag
coefficient monotonically decreased with increasing pressure gradient at 0.10 ≤ x/c ≤ 0.45.

Table 3.1.5 Aerodynamic performance of the airfoils with different pressure gradients (Ma = 0.73, Re =
6.5 × 106)

Figure 3.1.14 shows the geometries and pressure distributions of the four airfoils designed with
different pressure gradients, as well as the initial airfoil and the minimum drag airfoil. Because the
target pressure gradients cp,x of the ten points on the lower surface were not modified, the shapes
and pressure distributions of the lower surfaces of the inversely designed airfoils were almost the
same as those of the RAE2822. However, the shape and pressure coefficient on the lower surface of
the minimum drag airfoil were considerably changed with strong aft‐loading emerging and
frontloading disappearing. On the upper surface, the variations in the airfoil shapes were quite small,
which confirms the highly nonlinear characteristics of the transonic flow. A small change in geometry
may have had a strong effect on the shockwave and drag coefficient. The initial airfoil had a favorable
pressure gradient on the upper surface before the shockwave, which increased the shock strength
and deteriorated the drag coefficient. With the inverse design method, the expected pressure
gradient can be easily achieved. The shockwave strength weakens with increasing adverse pressure
gradient on the upper surface. The airfoil of the cp,x = 1.0 case was almost shock free.
 107

Figure 3.1.14 Geometries and pressure distributions of the airfoils (Ma = 0.73, Re = 6.5 × 106, CL =
0.800).
(a) Airfoil geometries; (b) pressure distributions.
Figure 3.1.13 shows the pressure gradients on the upper surfaces of the airfoils of Cp,x = 0.05 and
0.40. The front halves (x/c < 0.5) of the airfoils had different Cp,x targets, and the designs satisfied the
targets. The second halves (x/c > 0.5) had the same Cp,x target. One can see that the pressure gradients
near the shockwave location (0.5 < x/c < 0.6) had some deviations compared with that of the targets.
However, the present inverse
method was expressed as an
optimization problem of the
combination of the drag coefficient
and pressure gradient target,
which inherently includes a trade‐
off between the drag and the
pressure distribution.
3.1.2.10 Drag Divergence
Performance of Airfoils
with Different Pressure
Gradients
The upper surface pressure
gradient had a strong effect on the Figure 3.1.13 Pressure gradients on the upper surfaces of two
shockwave strength, which is airfoils (Ma = 0.73, Re = 6.5 × 106, CL = 0.800).
important for the drag creep and
drag divergence characteristics of a supercritical airfoil. The drag creep phenomenon is where the
drag slowly increases from the subcritical Mach number to the supercritical Mach number, and the
drag divergence is the fast increase after the design Mach number
[36]. Poor drag creep performance may decrease the efficiency of the second‐segment climb phase
of an aircraft and increase the thrust requirement of the engine. In contrast, the drag divergence
performance influences the cruise efficiency of a civil aircraft when cruising at a higher speed or
altitude. Both good drag creep performance and drag divergence performance are demanded for
aircraft design.
Figure 3.1.15 a shows the drag creep characteristics of the airfoils. The computations were carried
out with the same lift coefficient CL = 0.800 and Reynolds number 6.5 × 106. Under subcritical
conditions, the drag coefficient slowly increased with increasing Mach number. The initial airfoil had
 108

Figure 3.1.15 Drag creep and drag divergence of the airfoils (Re = 6.5 × 106, CL = 0.800). (a) Drag creep;
(b)
drag divergence.
the best drag creep performance, with the drag increasing the least from Ma = 0.45 to 0.65. In
contrast, the minimum drag airfoil had the worst drag creep performance in that the drag increased
dramatically. The second‐segment climb phase of a civil aircraft is quite time consuming, especially
for a short or medium‐range flight mission. Consequently, an airfoil optimized by minimizing the drag
coefficient only may not be practical. Figure 3.1.16 further shows the pressure distributions of the
airfoils. The critical pressure coefficient Cp∗ is also marked in the figure, which is computed by :

γ
2 2 γ − 1 2 γ−1
Cp∗ = {[ (1 + Ma∞ )] − 1}
γMa2∞ γ + 1 2
Eq. 3.1.14
Figure 3.1.15 a is at Ma = 0.65. It is clear that the suction peak of the minimum drag airfoil was
higher than those of the others, so the shock wave was stronger. At this Mach number, the drag
coefficient monotonically increased with the suction peak. Moreover, for the airfoil inversely
designed by the upper surface pressure gradient, the suction peak, and the drag coefficient at Ma =
0.65, also monotonically increased with the adverse pressure gradient.
Figure 3.1.15 - b shows the drag divergence characteristics of the airfoils. The initial airfoil had the
worst drag divergence performance as the drag increased quickly when Ma > 0.72. The airfoil with
the minimum drag had a “spoon” type drag curve because it had a high drag coefficient at a small

Figure 3.1.16 Pressure distributions at different Mach numbers (Re = 6.5 × 106, CL = 0.800). (a) Ma =
0.65; (b) Ma = 0.72; and (c) Ma = 0.74
 109

Mach number and low drag at the design condition (Ma = 0.73). The curve of Cp,x = 1.0 was similar to
that of the minimum drag airfoil because this airfoil was similar to a shock free airfoil, as shown in
Figure 3.1.14 b. The drag vs. Mach curves of Cp,x = 0.05 and 0.2 were quite flat when Ma < 0.72, and
they did not show obvious “spoon” type drag curves.
Figure 3.1.16 b,c shows the pressure distributions of Ma = 0.72 and 0.74, which were slightly lower
and higher than the design condition (Ma = 0.73). When Ma = 0.72, the initial airfoil presented a
strong shockwave, while the other airfoils presented a reacceleration area after the first shock.
Figure 3.1.17 a shows the Mach number contour of the Cp,x = 0.40 airfoil at Ma = 0.72. It is clear that
the reacceleration area formed a second shock. The drag coefficient of the airfoil with double‐shock
compression was slightly lower than that with single‐shock compression because the entropy
production of a double shock was lower than that of a single shock when the pressure increase was
the same [37]. However, the double shock made the drag curve not monotonically increase with the
Mach number.

Figure 3.1.17 Flow fields of airfoils at Ma = 0.72, Re = 6.5 × 106, CL = 0.800. (a) cp,x = 0.40; (b) cp,x =
0.05.

Figure 3.1.17 b also shows the flow field of Cp,x = 0.05. The reacceleration area was quite weak, and
no obvious double shock phenomena appeared. When the Mach number was higher than the design
Mach number of 0.73, the pressure distributions demonstrated strong shockwaves, as shown in
Figure 3.1.17-c. The drag coefficient was determined by the shockwave strength, i.e., the pressure
difference before and after the shockwave.
Considering the lift‐to‐drag ratio at the design condition, the drag creep and the drag divergence
characteristics, the airfoil optimized with Cp,x = 0.20 had a stable performance under different Mach
numbers and a moderate lift‐to‐drag ratio at the design condition. Consequently, the pressure
gradient of Cp,x = 0.20 on the front half of the upper surface was a satisfactory value for supercritical
airfoil design. The airfoil of Cp,x = 0.05 was also satisfactory with a stable performance; however, the
lift‐to‐drag ratio was slightly lower than that of Cp,x = 0.20.

3.1.3 Conclusions
This paper proposes an inverse design method based on the pressure gradient target. The adjoint
method is adopted to compute the derivatives of the pressure gradient, and the sequential quadratic
programming algorithm is used to find the optimal solution. The free form deformation method is
applied for geometric deformation. The advantage of this method is that only a rough expectation of
the pressure distribution pattern is required in the design process rather than a precise pressure
coefficient under a certain lift coefficient and Mach number. The method will be further studied on a
three‐dimensional wing. The work of this paper can be summarized as:
 110

➢ The computational method is validated well by the experimental data of the RAE2822 airfoil.
Two test cases of inverse design, including a supercritical airfoil and a supercritical NLF
airfoil, are used to test the effectiveness of the method. The results show that the method can
design airfoils with good performance by target pressure gradients of less than 20 points on
the upper surface and less than 10 points on the lower surface. The airfoil designed by the
method has both low aerodynamic drag and an appropriate pressure distribution pattern.
The supercritical natural laminar airfoil is also validated by the transition model and
demonstrates good laminar performance.
➢ Several airfoils with different upper surface gradients are designed and compared. When the
upper surface gradient before x/c = 0.5 is larger than 0.6, the airfoil tends to be shock free.
The drag creep and drag divergence characteristics of the airfoils are tested. The shock free
airfoil has deteriorated drag creep behavior, and the drag vs. Mach number curve becomes a
spoon shape. The high leading edge suction peak is the reason for the poor drag creep
performance. A double‐shock phenomenon is observed on airfoils with large adverse
pressure gradients. A pressure gradient Cp,x = 0.2 for the front half of the upper surface is
proposed for a supercritical airfoil to obtain both good drag creep and drag divergence
performances.
Author Contributions: Conceptualization, Y.Z.; Data curation, Y.Z.; Formal analysis, H.C.;
Investigation, C.Y.; Methodology, C.Y. and H.C. All authors have read and agreed to the published
version of the manuscript.
Funding: This work was supported by the National Natural Science Foundation of China (91852108
and 11872230).
Acknowledgments: The authors would like to thank the Multidisciplinary Design Optimization
Laboratory of the University of Michigan for providing the ADflow code under the GNU Lesser General
Public License (LGPL).
Conflict of interest: The authors declare that they have no conflicts of interest with regard to this
work.

3.1.4 References
1. Zhang, Y.; Yin, Y. Study on Riblet Drag Reduction Considering the Effect of Sweep Angle_ Energies
2019, 12, 3386.
2. Xu, Z.; Saleh, J.H.; Yang, V. Optimization of Supercritical Airfoil Design with Buffet Effect_ AIAA J.
2019, 57, 4343–4353.
3. Driver, J.; Zingg, D.W. Optimized Natural‐Laminar‐Flow Airfoils_ In Proceedings of the 44th AIAA
Aerospace Sciences Meeting and Exhibit, Reno, NV, USA, January 2006; 247.
4. Raheem, M.A.; Edi, P.; Pasha, A.A.; Rahman, M.M.; Juhany, K.A. Numerical Study of Variable Camber
Continuous Trailing Edge Flap at Off‐Design Conditions_ Energies 2019, 12, 3185
5. Zhao, T.; Zhang, Y.; Chen, H.; Chen, Y.; Zhang, M. Supercritical Wing Design Based on Airfoil
Optimization and 2.75D Transformation_ Aerosp. Sci. Technol. 2016, 56, 168–182.
6. Zhang, Y.; Fang, X.; Chen, H.; Fu, S.; Duan, Z.; Zhang, Y. Supercritical Natural Laminar Flow Airfoil
Optimization for Regional Aircraft Wing Design_ Aerosp. Sci. Technol. 2015, 43, 152–164.
7. Edi, P.; Fielding, J.P. Civil‐Transport Wing Design Concept Exploiting New Technologies_ J. Aircr.
2006, 43, 932–940.
8. Gopalarathnam, A.; Selig, M.S. Hybrid Inverse Airfoil Design Method for Complex Three‐
Dimensional Lifting Surfaces_ J. Aircr. 2002, 39, 409–417.
9. Selig, M.S.; Maughmert, M.D. Multipoint Inverse Airfoil Design Method Based on Conformal
Mapping. AIAA J. 1992, 30, 1162–1170.
10. Malone, J.B.; Narramoret, J.C.; Sankar, L.N. Airfoil Design Method Using the Navier‐Stokes
Equations. J. Aircr. 1991, 28, 216–224.
 111

11. Dadone, A.; Grossman, B. Fast Convergence of Viscous Airfoil Design Problems. AIAA J. 2002, 40,
1997–2005.
12. Gopalarathnam, A.; Selig, M.S. Low‐Speed Natural‐Laminar‐Flow Airfoils: Case Study in Inverse
Airfoil Design. J. Aircr. 2001, 38, 57–63.
13. Eyi, S.; Lee, K.D. Inverse Airfoil Design Using the Navier‐Stokes Equations. In Proceedings of the
7th Applied Aerodynamics Conference, Irvine, CA, USA, February 1993; 0972.
14. Gardner, B.A.; Selig, M.S. Airfoil Design Using a Genetic Algorithm and an Inverse Method. In
Proceedings of the 41st Aerospace Sciences Meeting and Exhibit, Reno, NV, USA, January 2003; 0043.
15. Hacioglu, A. Fast Evolutionary Algorithm for Airfoil Design via Neural Network. AIAA J. 2007, 45,
2196–2203.
16. Bui‐Thanh, T.; Damodaran, M.; Willcox, K. Aerodynamic Data Reconstruction and Inverse Design
Using Proper Orthogonal Decomposition. AIAA J. 2004, 42, 1505–1516.
17. Ylmaz, E.; German, B.J. A Deep Learning Approach to an Airfoil Inverse Design Problem. In
Proceedings of the 2018 Multidisciplinary Analysis and Optimization Conference, Atlanta, GA, USA,
June 2018; 3420.
18. Sekar, V.; Zhang, M.; Shu, C.; Khoo, B.C. Inverse Design of Airfoil Using a Deep Convolutional Neural
Network. AIAA J. 2019, 57, 993–1003.
19. Tang, H.; Lei, Y.; Li, X.; Gao, K.; Li, Y. Aerodynamic Shape Optimization of a Wavy Airfoil for Ultra‐
Low Reynolds Number Regime in Gliding Flight. Energies 2020, 13, 467.
20. Jameson, A. Aerodynamic Design via Control Theory. J. Sci. Comput. 1988, 3, 233–260.
21. Mader, C.A.; Martins, J.R.R.A.; Alonso, J.J.; Weide, E. ADjoint: An Approach for the Rapid
Development of Discrete Adjoint Solvers. AIAA J. 2008, 46, 863–873.
22. He, P.; Mader, C.A.; Martins, J.R.R.; Maki, J. DAFoam: An Open‐Source Adjoint Framework for
Multidisciplinary Design Optimization with OpenFOAM. AIAA J. 2019, 58, 1304–1319.
23. Li, R.; Deng, K.; Zhang, Y.; Chen, H. Pressure Distribution Guided Supercritical Wing Optimization.
Aeronaut. J. 2018, 31, 1842–1854. Energies 2020, 13, 3400 18 of 18
24. Kenway, G.K.W.; Mader, C.A.; He, P.; Martins, J.R.R.A. Effective Adjoint Approaches for
Computational Fuid Dynamics. Prog. Aerosp. Sci. 2019, 110, 100542.
25. Jameson, A.; Schmidt, W.; Turkel, E. Numerical Solution of the Euler Equations by Finite Volume
Methods Using Runge–Kutta Time Stepping Schemes. In Proceedings of the 14th Fluid and Plasma
Dynamics Conference, Palo Alto, CA, USA, June 1981; 1259.
26. Spalart, P.; Allmaras, S. A One‐Equation Turbulence Model for Aerodynamic Flows. In Proceedings
of the 30th Aerospace Sciences Meeting and Exhibit, Reno, NV, USA, January 1992; 0439.
27. Cook, P.H.; McDonald, M.A.; Firmin, M.C.P. Experimental Data Base for Computer Program
Assessment; Aerofoil RAE 2822—Pressure Distributions, and Boundary Layer and Wake
Measurements; AGARD Advisory Report No. 138; 1979.
28. Hascoet, L.; Pascual, V. The Tapenade Automatic Differentiation Tool: Principles, Model, and
Specification. ACM Trans. Math. Softw. 2012, 7957.
29. Gill, P.E.; Murray, W.; Saunders, M.A. SNOPT: An SQP Algorithm for Large‐Scale Constrained
Optimization. Siam. Rev. 2005, 47, 99–131.
30. Zhang, T.; Wang, Z.; Huang, W.; Yan, L. A Review of Parametric Approaches Specific to
Aerodynamic Design Process. Acta. Astronaut. 2018, 145, 319–331.
31. Zhao, K.; Gao, Z.; Huang, J.; Li, Q. Aerodynamic Optimization of Rotor Airfoil Based on Multi‐Layer
Hierarchical Constraint Method. Chin. J. Aeronaut. 2016, 29, 1541–1552.
32. Li, J.; Gao, Z.; Huang, J.; Zhao, K. Robust Design of NLF Airfoils. Chin. J. Aeronaut. 2013, 26, 309–
318.
33. Han, Z.H.; Chen, J.; Zhang, K.S.; Xu, Z.M.; Zhu, Z.; Song, W.P. Aerodynamic Shape Optimization of
Natural‐Laminar‐Flow Wing Using Surrogate‐Based Approach. AIAA J. 2018, 56, 2579–2593.
34. Hepperle, M. JavaFoil—Analysis of Airfoils. Available online: http://www.mhaerotools.
 112

de/airfoils/javafoil.htm (accessed on 2 July 2020).

35. Menter, F.R.; Langtry, R.B.; Likki, S.R.; Suzen, Y.B.; Huang, P.G.; Völker, S. A Correlation‐Based
Transition Model Using Local Variables‐Part I: Model Formulation. J. Turbomach. 2006, 128, 413–422.
36. Harris, C.D. NASA Supercritical Airfoils: A Matrix of Family‐Related Airfoils; NASA‐TP‐2969; NASA
Langley Research Center: Hampton, VA, USA, 1990.
37. Liepmann, H.W.; Roshko, A. Elements of gas dynamics, In Waves in Supersonic Flow; Dover
publications: New York, NY, USA, 2001; Chapter 4, pp. 84–123. © 2020 by the authors. Licensee MDPI,
Basel, Switzerland. This article is an open access article distributed under the terms and conditions
of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0
 113

3.2 Case Study 2 - Inverse Aerodynamic Design Method for Aircraft Component

Authors : J.B. Malone1, J. Vadyak2 and L.N. Sankar3

Affiliations : 1-2Lockheed-Georgia Company, Marietta, Georgia,
3Georgia Institute of Technology, Atlanta, Georgia

Title of Paper : Inverse Aerodynamic Design Method for Aircraft Component

Citation : Invalid source specified.
Bibliography : Nicolosi, F., Vecchia, P. D., Ciliberti, D., & Cusati, V. (2015). FUSELAGE AERODYNAMIC
DRAG PREDICTION METHOD BY CFD. CEAS 2015 paper no. 37.

The motivation for using automated, inverse aerodynamic design methods is to reduce the overall
effort required to develop aircraft geometries possessing favorable aerodynamic performance or
aerodynamic interference characteristics, as investigated Invalid source specified.. Invalid source
specified.15 described an iterative aerodynamic design procedure suitable for automated wing
design (referred to here as the GM method). They demonstrated their method by incorporating it
into a three dimensional, full-potential, transonic wing, aerodynamic analysis code. In the GM design
method, an auxiliary partial differential equation governing the spatial location of wing surface
ordinates was solved iteratively in the computational plane, together with the fluid flow equation, to
achieve given target surface-pressure distributions. The technique, was recommended for use over
only a limited portion of the wing geometry. From the present authors' experience, the original GM
method of updating wing ordinates in a normal, or nearly normal, direction to the surface can lead to
irregularities in the final design pressures near the leading edge of the design geometry. This is due
primarily to the non-uniform stretching that can occur near the leading edge where the surface
normal directions are parallel to the longitudinal axis of the component geometry.

3.2.1 Introduction and Background

The present inverse design procedure is formulated in a manner similar to that of the original
[Garabedian-McFadden] scheme, but the auxiliary equation is solved directly in the physical domain,
rather than in the computational domain. The new design method, which will be referred to in the
following paragraphs as the [Modified Garabedian-McFadden (MGM)]16 method, was developed to
improve airfoil or wing designs where the control of leading-edge pressures is desirable and to
extend the method to handle different types of configurations, including axisymmetric or asymmetric
body geometries. The surface perturbations generated with the MGM procedure are interpreted as
changes in the coordinate direction perpendicular to the longitudinal axis of the geometry. This
interpretation for the movement of the surface coordinates leads to smoother leading-edge geometry
designs.

3.2.2 Formulation
A two-dimensional airfoil design problem is used here to illustrate the salient features of the MGM
procedure. For this two-dimensional application, the original GM auxiliary equation is rewritten as

∂z ∂2 z ∂3 z
β0 + β1 + β2 2 = ∆Q2 = Q2 − q2
∂t ∂x ∂t ∂x ∂t
Eq. 3.2.1
Where the Q's are user specified target pressures expressed as flow velocities, the q computed flow

15 Garabedian, P. and McFadden, G., "Design of Supercritical Swept Wings," AIAA Journal, Vol. 20, March 1982.
16 Malone, J.B., Vadyak, J., and Sankar, L.N., "A Technique for the Inverse Aerodynamic Design of Nacelles and Wing

Configurations," AIAA Paper 85-4096, Oct. 1985.

 114

velocities predicted by the given fluid flow solution procedure, and the 3 user-defined constants that
improve the convergence of the algorithm. The time coordinate in Eq. 3.2.1 is actually a pseudo time
variable representing different iterations in the solution process. As Q approaches q, the right-hand
side of Eq. 3.2.1 vanishes and the surface coordinates z(x , t) stop varying with pseudo time. Partial
derivatives with respect to the time coordinate are interpreted as a change in the surface coordinate
Az between any two design iterations. In order to apply Eq. 3.2.1 correctly to both upper and lower
surfaces, the value of Az must have opposite signs on each surface for equal values of the quantity,
Q2 - q2.

3.2.3 Method of Solution

Eq. 3.2.1 is normally evaluated along the complete surface of the design geometry. Target pressures
are supplied at discrete points around the geometric contour and then interpolated at locations
corresponding to computational mesh points on the aerodynamic surface. These target pressures
are then converted to target values of surface velocity. The computed surface velocities required to
evaluate the right-hand side of Eq. 3.2.1 can then be obtained from the fluid flow solution procedure
without the need for further interpolation. Next, finite difference expressions are written for each
term of Eq. 3.2.1. Assuming that there are a total of N computational points on the airfoil surface,
Eq. 3.2.1 is written for each of these points i, where 1 < I < N. A typical equation evaluated at the i-
th point of the surface is

Ai ∆zi−1 + Bi ∆zi + Ci ∆zi+1 = ∆Q2i

−β1 −2β2
Ai = +
(xi+1 − xi ) (xi+1 − xi )(xi+1 − xi−1 )
Eq. 3.2.2
A more complete description of the finite-difference expressions used in Eq. 3.2.2 can be found in17.
Eq. 3.2.2 is evaluated at each point i, leading to a system of TV equations in TV unknowns (the Azi
values). Note that at each point on the aerodynamic surface, Azi is coupled to values at neighboring
points. The resulting equations form a tridiagonal system that is solved for the values of Azi using the
well-known Thomas algorithm. Special treatment is required at two locations on the design
geometry contour. At the leading edge, an ambiguity arises as to the direction to apply the up winding
used to evaluate certain derivatives in Eq. 3.2.2. To eliminate this problem, the leading-edge point
is constrained to move as the average of both the upper- and lower-surface downstream points. This
also permits the local angle of attack to change during the design process. Also, at the trailing edge,
values of Az are needed to evaluate the second derivative terms. Currently, these values are set to
zero. Thus, during the design process, the trailing-edge thickness remains constant. Actually, the
baseline, or starting geometry used to initiate the design process, serves primarily to fix the trailing-
edge thickness of the final design geometry.

3.2.4 Application and Validation

The MGM design procedure as applied to airfoil shapes is illustrated in Figure 3.2.1. Here, the design
algorithm is coupled to a full-potential airfoil code18. The figure shows the results of an airfoil design
problem in which a symmetric baseline airfoil with a NACA 0012 contour is modified into a highly
cambered, aft-loaded supercritical airfoil shape. Surface pressures were first obtained by analyzing
a known supercritical airfoil, the GA(W)-1, and then applied as targets during the design process. To

17 Malone, J.B., Vadyak, J., and Sankar, L.N., "A Technique for the Inverse Aerodynamic Design of Nacelles and Wing

Configurations," AIAA Paper 85-4096, Oct. 1985.

18 Malone, J.B. and Sankar, L.N., "Numerical Simulation of Two-Dimensional Unsteady Transonic Flows Using the

Full Potential Equation," AIAA Journal, Vol. 22, Aug. 1984, pp. 1035-1041.
 115

compute these results,

boundary layer effects
were included. Revealed
in Figure 3.2.1-a is a
comparison of the initial
vs target and target vs
final airfoil contours. A
comparison of the
baseline pressures, target
pressures, and final
design pressures is given
in Figure 3.2.1-b. The
design pressures plotted
were obtained from a
separate analysis of the
final geometry in order to
verify that an adequate
design was obtained.
An application of the MGM
design procedure for 3D
nacelle configurations is
illustrated in Figure
3.2.2. For this sample
case, the design procedure
is coupled to a three
dimensional, full-
potential nacelle code for
arbitrary nacelle inlet
configurations19. This case
is for an asymmetric
nacelle operating at a
freestream Mach number
Figure 3.2.1 Input, Target, and Airfoil Contours and Surface
of 0.8, an angle of attack of Pressures for Supercritical Airfoil Design
2.0 degree, and an inlet
mass flow ratio of 0.7. The
object was to determine if the original nacelle contour could be recovered if its pressure distribution
was used as the target distribution and if a perturbed geometry was used as the initial geometry
estimate. Figure 3.2.2-a illustrates the upper symmetry meridian nacelle contours for the original,
designed, and initial estimate geometries. Figure 3.2.2-b illustrates the corresponding surface
pressure distributions. As is observed, the analysis successfully recovers the original contour.

3.2.5 Conclusion
A well-known design procedure, the [Garabedian-McFadden] method, has been modified to permit
design of airfoil and wing leading-edge regions and to extend the range of configurations that can be
handled by this technique. The modified design procedure has been incorporated by the authors into
several existing aerodynamics programs and sample design problems have been presented for airfoil

19 Vadyak, J. and Atta, E.H., "Approximate Factorization Algorithm for Three-Dimensional Transonic Nacelle/Inlet

Flow field Computations," Journal of Propulsion and Power. Vol. 1, Jan.-Feb. 1985, pp. 58-64.
 116

and nacelle geometries. Although

simple in nature, the changes to the
existing design algorithm have, in all
cases examined by the authors,
improved the overall versatility of the
method. Test problems that were
successful using the original method
were not adversely effected by the
modified scheme presented here.
However, several classes of problems,
an example of which is the case shown
here for the two-dimensional airfoil,
were successful only when the
modified algorithm was used during
the design process.

Figure 3.2.2 Input, Target, and Computed Nacelle

Contours and Surface Pressures for an Asymmetric Nacelle
Design Problem
 117

3.3 Case Study 3 - A Pseudo-Compressibility Method for Solving Inverse Problems

Based on the 3D Incompressible Euler Equations

Citation : M. Ferlauto (2015) A pseudo-compressibility method for solving inverse problems based on
the 3D incompressible Euler equations, Inverse Problems in Science and Engineering, 23:5, 798-
817, DOI: 10.1080/17415977.2014.939653

A numerical technique to solve the 3D inverse problems that arise in aerodynamic design is
presented by (Ferlauto, 2014)20. The approach, which is well established for compressible flows, is
extended to the incompressible case via artificial compressibility preconditioning. The modified
system of equations is integrated with a characteristic-based Godunov method. The solution of the
inverse problem is given as the steady state of an ideal transient during which the flow field assesses
itself to the boundary conditions, which are prescribed as design data, by changing the boundary
contour. The main aspects of the Eulerian-Lagrangian numerical procedure are illustrated and the
results are validated by comparisons with theoretical solutions and experimental results.

3.3.1 Introduction & Literature Survey

Aerodynamic design consists of finding the most convenient shape for a fluid-immersed body that
realizes some expected aerodynamic performances. There is a certain freedom in approaching
aerodynamic design. Several routes were exploited in the past by researchers and engineers, with
the same goals in mind. The most-followed approach iterates on a sequence of direct analyses.
Starting from an initial configuration, the geometry is modified and the aerodynamic effects are
either experimentally or numerically evaluated. Thanks to designer experience and intuition this
process is driven toward the expected performances. In Computational Fluid Dynamics (CFD) the
automation of this approach is known as the classical shape optimization21-22-23. This strategy
defines a functional, to measure the merit or cost of a certain aerodynamic solution, and a number of
geometric control parameters that are used to modify the solution. Then, the functional gradient
relative to the controls is evaluated numerically and its information is used to update the control
parameters and to march towards the functional extremum. Several methodologies have been
proposed in literature to evaluate the functional gradient and also to impose fluid-dynamic and
geometrical constraints. Another way of solving the aerodynamic problem is based on the
formulation of an inverse problem. In this case the aerodynamic surface is unknown and some flow
features are given on it.
The solution of the inverse problem enables designers to obtain the geometry of an aerodynamic
device satisfying the governing equations with all the given boundary conditions and other desirable
flow constraints. One typical inverse problem is that of finding the airfoil geometry, given the flight
speed and the pressure distribution on the profile24. The solution of the inverse problem is often a
complex task, if compared to the direct analysis approach. Nevertheless it offers the main advantage
of requiring a fewer number of flow field evaluations to determine the enquired geometry, and also

20 Michele Ferlauto, “A Pseudo-Compressibility Method For Solving Inverse Problems Based On The 3D

Incompressible Euler Equations”, Taylor & Francis in Inverse Problems in Science and Engineering ,2014.
21 O. Pironneau. On optimum design in fluid mechanics. Journal of Fluid Mechanics, 59:117–128, 1972.
22 A. Jameson, L. Martinelli, and N. A. Pierce. Optimum aerodynamic design using the Navier-Stokes equations.

Theoretical and Computational Fluid Dynamics, 10:213–237, 1998.

23 B. H. Dennis, Z.-X. Han, and G. S. Dulikravich. Optimization of turbomachinery airfoils with a genetic/sequential

quadratic programming algorithm. AIAA Journal of Propulsion and Power, 17(5):1123–28, 2001.
24 J. M. Lighthill. A new method of two-dimensional aerodynamic design. Aeronautical Research Council Reports

and Memoranda n. 2122, 1945.

 118

it allows for an easier imposition of the flow constrains (e.g. no flow separation, adverse pressure
gradient control25-26.
Approaches to the inverse problem solution are based on the potential flow theory and conformal
mapping techniques27-28-29, on stream-function base formulations 30-31, on boundary elements
replacing the body surface32, on the compressible potential flow theory. Several examples of inverse
problem solution methodologies are presented in 33 and references therein. A generalizable way of
solving aerodynamic inverse problems identifies the problem solution as the steady state of an
unsteady evolution during which the flow field tries to accommodate itself to the design data, which
are prescribed as boundary conditions on the unknown surface 34. To do this, the surface is allowed
to move. To give an idea of the physical scenario involved let us consider, for instance, the design of
an airfoil.
First, an arbitrary initial geometry is guessed and the related flow field at the selected flight Mach
number is evaluated. Then, the target pressure distribution along the profile is imposed. In general,
the boundary conditions will not be in agreement with the inner flow field. If we let the profile surface
move while it remains impermeable to the flow, walls start moving to decrease the pressure gap. If
the system reaches a steady state, and this feature relies on the existence of the steady solution for
the Euler equations, then the enquired geometry and the corresponding flow field are found. A
similar approach for the determination of the unknown shape is the elastic membrane motion
concept or the transpiration velocity 35.
One drawback of inverse problems is that they may be ill posed. If certain wall pressure distributions
are required on airfoils, the result is an open or self-intersecting profile. [Lighthill] discovered the
solvability conditions that have to be respected by pressure distributions within an incompressible
potential flow model, whereas the conditions for compressible flows and other issues where
investigated in and references therein36-37. The main advantage of numerical optimization over
inverse problem is that the first allows the maximization or minimization of global quantities, such
as lift or drag, in the presence of constraints, whereas for inverse problems the design is limited to
the selection of the pressure or velocity distribution on the boundary38, which is given on the basis

25 G.S. Dulikravich, B.H. Dennis, D.P. Baker, H.R.B. Orlande, and M.J. Colaco. Inverse problems in aerodynamics,
heat transfer, elasticity and materials design. International Journal of Aeronautical & Space Sciences, 2012.
26 A. Iollo, M. Ferlauto, and L. Zannetti. An aerodynamic optimization method based on the inverse problem adjoint

equations. Journal of Computational Physics, 173:87–115, 2001.

27 W. Mangler. Die berechnung eines tragfl¨ugelprofiles mit vorschriebener dr¨uckverteilung. Jahrbuch

Deutsches Luftfahrtvorschung, 1:46–53, 1938.

28 J. D. Staniz. Design of two-dimensional channels with prescribed velocity distributions along the channel walls.

NACA Report 1153, 1953.

29 P. Daripa and L. Sirovich. An inverse method for subcritical flows. Journal of Computational Physics, 1986.
30 L. Zannetti. A natural formulation for the solution of two-dimensional or axis-symmetric inverse problems.

International Journal for Numerical Methods in Engineering, 22:451–463, 1986.

31 JJ Keller. Inverse equations. Physics of Fluids, 11:513–520, 1999.
32 R. I. Lewis. Vortex Element Methods for Fluid Dynamic Analysis of Engineering Systems. Cambridge University

Press, New York, 1991.

33 A. M. Elizarov, N. B. Il’innskiy, and A. V. Potashev. Mathematical Methods of Airfoil Design. Akademie Verlag,

Berlin, 1997.
34 L. Zannetti. Time dependent method to solve inverse problems for internal flows. AIAA Journal, 1980.
35 A. Demeulenaere and R. Van den Braembussche. Three-dimensional inverse method for turbomachinery

blading design. ASME Journal of Turbomachinery, 120(2):247–255, 1998.

36 L. Zannetti and M. Pandolfi. Inverse design techniques for cascades. NASA CR 3836, 1984.
37 G. Volpe. Geometric and surface pressure restrictions in airfoil design. AGARD Report 780, 1990.
38 G.S. Dulikravich. A criteria for surface pressure specification in aerodynamic shape design. 28th Aerospace

Sciences Meeting, AIAA Paper 90-0124, 1990.

 119

of designer experience and therefore, somewhat arbitrarily. In addition, a limited control is possible
on the final geometry.
Optimization techniques based on the adjoint method can be adopted to drive inverse problems
towards the maximization or minimization of target functionals. From this point of view, the
numerical solution of an inverse problem becomes attractive, as an alternative route to shape
optimization and to automated design. In the present work a time-dependent technique of solving
inverse problems 39 is extended to the aerodynamic design in three-dimensional incompressible
flows. The numerical method is based on the artificial compressibility approach to make the system
of equations virtually hyperbolic 40-41. This choice allows us to easily set up an explicit, Godunov-
type numerical scheme42 and to reuse well established techniques of solving the compressible Euler
and Navier Stokes equations, as well as the related strategies of code optimization and parallelization.
Although our investigation has been here limited to inviscid flows, the general methodology has been
already extended to viscous flows and validated for the case of two-dimensional and axis-symmetric
compressible flows 43.
The plan of the paper is as follows: the mathematical model and the numerical technique are
explained in the context of a characteristic-based finite volume method; then the equation of motion
of impermeable surfaces is derived in either Cartesian or cylindrical coordinates; finally, some
numerical examples are presented and the accuracy of the method is studied through comparisons
against theoretical solutions and experimental results.

3.3.2 Mathematical Model

The pseudo-compressibility method is a preconditioning technique of overcoming the numerical
issues related to the uncoupling of pressure and velocity field in the incompressible Euler equations.
The governing equations are made artificially hyperbolic by adding a density time-derivative to the
continuity equation. Based on physical analogies, the time derivative of density is then linked to the
pressure by introducing an artificial compressibility factor c. In a Cartesian frame of reference, the
system of modified Euler equations is written as

∂𝐔 ∂𝐅 ∂𝐆 ∂𝐇
+𝚯( + + )=0
∂t ∂x ∂y ∂z
Eq. 3.3.1

𝜌 𝑐2 0 0 0
𝑢
𝐔={ } , 𝚯 = [0 1 0 0]
𝑣 0 0 1 0
𝑤 0 0 0 1

39 L. Zannetti and F. Larocca. Inverse methods for 3D internal flows. AGARD Report 780, 1990.
40 A. J. Chorin. A numerical method for solving incompressible viscous flow problems. Journal of Computational
Physics, 2:12–26, 1967.
41 A. Rizzi and L. E. Eriksson. Computation of inviscid incompressible flow with rotation. Journal of Fluid

Mechanics, 153:275–312, 1985.

42 D. Drikakis and W. Rider. High-Resolution Methods for Incompressible and Low-Speed Flows (Computational

Fluid and Solid Mechanics). Springer Verlag, Berlin, 2005.

43 M. Ferlauto and R. Marsilio. A viscous inverse method for aerodynamic design. Computers & Fluids, 2006.
 120

u v w
p + u2 uv uw
𝐅={ } , 𝐆={ } , 𝐇 = { vw }
uv p + v2
uw vw p + w2
Eq. 3.3.2
as usual, p is pressure, q = {u, v,w} is the velocity vector. All the variables are normalized to the
reference length lref , density ρref =ρ0, and pressure pref. The unsteady solutions of Eq. 3.2.1 does not
have a physical meaning, but, from the numerical point of view, the methods of solution for
compressible flow become available. At the steady state, the solution of Eq. 3.3.1 also satisfies
steady Euler equations. The proposed strategy of solving the inverse problem requires the numerical
integration of a 3D flow field on a time-dependent domain. Unsteady flow field are solved, in the
context of artificial compressibility methods, by introducing a dual time stepping technique [26].
Dual time stepping is here unnecessary, since we are interested in over-relaxing the system to the
steady state. Moreover, the time-dependent derivative of the metrics can be neglected without a
significant loss of numerical stability. According to the Gauss formula, an integral form of Eq. 3.3.1
is
∂
∫ UdΩ + 𝚯 ∫(𝐅i + 𝐆j + 𝐇k). 𝐧 dσ = 0
∂t Ω 𝚺
Eq. 3.3.3
where ∑ is the boundary of the volume Ω and n the outward normal. Eq. 3.3.3 is approximated using
a finite volume technique by discretizing the (x, y, z) space with hexahedral cells whose shape
depends on time. The integration in time is carried out according to a Godunov type two-step scheme.
A first order Flux Difference Splitting (FDS) is used at the predictor step: the variables (p, u, v and
w) are assumed as an averaged, constant value inside each cell. The fluxes F, G and H are evaluated
by solving the Riemann problems pertinent to the discontinuities that take place at the cell interfaces.
At the corrector level, the second order of accuracy is achieved by assuming a linear behavior of the
primitive variables inside the cells, according to an Essentially Non-Oscillatory (ENO) procedure.
Implicit residual smoothing is applied to increase the numerical stability. The resulting scheme is
second order accurate in space 44.

3.3.3 Dynamics of Materially Moving Boundaries.

Walls on which a flow variable, e.g. the pressure, is imposed, are treated as flexible and impermeable
surfaces whose motion is tracked in time until a steady state is reached. In inviscid flows, the velocity
vector must be tangent to the wall surface, which is considered a flow surface. It therefore moves
materially within the fluid. In this context, the no-through-flow condition represents the kinematic
constraint from which the motion of the wall surface is derived. Let us consider a new Cartesian
coordinate system Γ = (x1, x2, x3). A generic surface moving in time can be represented, for instance,
by equation x3 = b(x1, x2, t). Without any loss of generality, let us assume that b (x1, x2, 0) = 0. Thanks
to the change of variable B = x3 − b , the surface is now defined by the manifold B(x1, x2, x3, t) = 0

B(x1 , x2 , x3 , t) = 0
Eq. 3.3.4
This manifold moves materially within the fluid, that is

44D. Drikakis and W. Rider. High-Resolution Methods for Incompressible and Low-Speed Flows (Computational
Fluid and Solid Mechanics). Springer Verlag, Berlin, 2005.
 121

dB ∂B ∂B ∂b ∂b
= + 𝐪. ∇B = 0 or = u3 − u1 − u2
dt ∂t ∂t ∂x1 ∂x2
Eq. 3.3.5
where q = (u1, u2, u3) is the flow velocity expressed in the Γ reference system. In the followings, Eq.
3.3.4 is coupled to a properly defined Riemann problem. The resulting system is used to compute
the fluxes at the boundary and to evaluate the wall velocity. The latter is then integrated in time to
find the new wall contour. For the inverse problem on a nozzle or diffuser configuration, as shown
in Figure 3.3.1-a, the pressure distribution is imposed on the unknown boundary45. Aerodynamic
devices as wings or blades, instead, are composed by two surfaces that represent a closed contour.
In this case, the inverse problem formulation must ensure a correct profile closure and non-

Figure 3.3.1 Schematic View of the Imposed Boundary Conditions for Inverse Problem on a 3D Nozzle

overlapping conditions. Among possible approaches, in addition to suitable boundary conditions at

infinity, one may:
1. prescribe the distribution of thickness and pressure jump along the chord and inquire for
the geometry of the camber line;
2. prescribe the distribution of thickness and pressure on one side of the profile and inquire
the geometry of the camber line;
3. prescribe the pressure distribution around the profile and inquire for its geometry.

Case (3) is the most challenging since, when selecting the pressure distribution, additional closure
constraints must be imposed 46. Moreover the convergence rate is low because smaller time-steps
are required and also boundary crossing must be checked at each step. All of the mentioned cases
can be grouped in two approaches in imposing the boundary conditions: the case of a surface on
which a pressure distribution is imposed and the case of two aerodynamic surfaces subjected to a
prescribed pressure jump. In next sections the kinematic constraint is accomplished to the flow field
boundary conditions in order to derive the fluxes and the velocity on the unknown surfaces. For

45 (b) wave pattern of the Riemann problem at the unknown aerodynamic surface on which the pressure is
prescribed.
46 G. Volpe. Geometric and surface pressure restrictions in airfoil design. AGARD Report 780, 1990.
 122

further info, regarding moving boundary conditions, please refer to [Ferlauto] 47.

3.3.4 Numerical Results for 3D Nozzle

The inverse design of a nozzle has been chosen as test case for the inverse problem formulation
described in [Ferlauto]48. The pressure distribution imposed on the unknown aerodynamic surface
is derived from the potential flow theory. The planar solution is extended in three dimensions by
applying periodic boundary conditions in the transversal direction z. A converging-diverging nozzle
is obtained from the flow field of a standing vortex on the half plane. The flow can be represented by
the complex potential W

𝜉 − 𝑖𝑎
W(ξ) = u∞ − iκ log ( ) , Ψ(𝜉) = ℑ(𝑊(𝜉))
𝜁 + 𝑖𝑎
Eq. 3.3.6
where ξ = z + iy is the position, u∞ is the asymptotic velocity, κ the vortex strength and a is the distance
of the vortex from the x-axis. The two-dimensional nozzle is confined by the streamlines ψ(0) and
ψ(ib), where 0 < b < a. The velocity field can be deduced from Eq. 3.3.6 as

1 1
u − iw = u∞ − iκ ( − )
𝜁 − 𝑖𝑎 𝜁 + 𝑖𝑎
Eq. 3.3.7
while the pressure field is obtained from Bernoulli equation, once the total pressure level P0in at inlet
is fixed. The numerical test has been performed on a 90×30×20 grid for the case with b = 0.5 , u∞ =
0, κ = −1. We also set c = 3. With the scope of generating a fully 3D transient flow, we started from
an initial geometry characterized by a wavy upper wall. The theoretical values of total pressure P0in
= 1, the flow angles at inlet α = 0 and β = 0, the static pressure distribution at the outlet pex(y, z), and
the target static pressure pup = p(ψ(ib)) along the upper wall are imposed as boundary conditions. As
final solution, the projection of the upper wall geometry on the (x, y)-plane is expected to match the
theoretical curve ψ(ib). Initial and final geometry are shown in Figure 3.2.1 (a-b).

3.3.5 Inverse Problem on a Stator Blade

The numerical experiment here proposed is a check of the full correspondence, at the steady state,
between the inverse procedure and the direct solver solution on the same geometry. The blade
geometry computed by the inverse procedure is introduced in a direct solver and the flow field is
computed again, now by using standard boundary conditions. The two solutions are then compared.
Arbitrarily, we assumed the following expression for the blade thickness

x x − xl
t ( ) = b0 + κca (1 − ξ)√ξ(1 − ξ) , ξ=
ca ca
Eq. 3.3.8
where xl, xt are the leading and trailing edge axial stations, respectively, and ca = xt − xl is the axial
chord. The constants are set to b0 = 0.04 and k = 0.12 . Similarly, the blade load is prescribed as

47 Michele Ferlauto, “A Pseudo-Compressibility Method For Solving Inverse Problems Based On The 3D

Incompressible Euler Equations”, Taylor & Francis in Inverse Problems in Science and Engineering ,2014.
48 Michele Ferlauto, “A Pseudo-Compressibility Method For Solving Inverse Problems Based On The 3D

Incompressible Euler Equations”, Taylor & Francis in Inverse Problems in Science and Engineering ,2014.
 123

x
∆p ( ) = γ[1 − cos(2πξ)]
c
Eq. 3.3.9
with γ = 0.675 . The inlet and outlet boundary condition are: P0in = 1, α = 0, β = 0 and pex = 0.99,
respectively. Starting from the initial guess of a blade with planar camber line, the transient dynamics
of the system is integrated numerically until a steady state is reached. At each step the blade
geometry is recovered from the computed velocity of the blade surface. Only the dynamics of the
blade suction-side is computed. Pressure-side surface is deduced from the knowledge of thickness. A
snapshot sequence of the blade moving surface is depicted in Figure 3.3.2 (a-b-c)49. The transient
resembles the motion of a flag on the wind, fastened on one side. More precisely, the grid describes a

Figure 3.3.2 Inverse Problem on a Stator Blade

49(a-b-c) Evolution of the computational domain and of the blade geometry. Iso-pressure contours are also
shown. (d) Comparison of the target blade load with the corresponding distributions computed by the inverse
procedure and by the direct solver.
 124

sliding motion on the tangential direction, since the axial chord of the blade does not change during
the inverse process. At convergence, the blade geometry is found. By switching the code in direct
mode, the flow field is computed with standard boundary conditions, and the blade load is evaluated.
The solutions of the inverse and direct solver computations are in good agreement. The computed
blade load Δp, in direct and inverse modes and the target (Δp)t, defined in Eq. 3.3.9, are compared
in Figure 3.3.2-d. A pressure mismatch, with respect to the target blade load distribution of Eq.
3.3.9, can be observed close to the blade leading edge. This is a well-known local effect caused by the
poor resolution properties of the H-type grids close to the leading edge. A finer grid gives a more
detailed flow but is meaningless for the test proposed, since the inverse and the direct solvers are
acting on the same grid.

3.3.6 Conclusions
A numerical method of solving three-dimensional inverse problems in incompressible flows has been
presented. Details have been given for the derivation of the inverse problem over configurations of
interest in the aerodynamic design as nozzles, diffusers, blades or wings. The numerical method is
based on the pseudo-compressibility technique of preconditioning the unsteady Euler equations.
Time accuracy is unnecessary, so that dual time stepping is avoided. Moreover, the characteristic-
based formulation of the problem simplifies the parallelization of the code. The accuracy of the
numerical procedure has been tested through comparisons with theoretical solutions, numerical
computations performed with direct solvers, and finally, test against experimental data.
Discrepancies at the blade leading edge have been observed, caused by the H-type grid we used.
Further extension of the inverse procedure to the Navier-Stokes equations, in a way similar to 50 is
planned. An intermediate step required is the introduction of efficient overrelaxing techniques, to
increase the convergence ratio. For the inverse design of aerodynamic profiles the surface tracking
algorithm should be modified to deal with C-type grid generation systems that allow for an accurate
evaluation of the viscous stresses.

50 M. Ferlauto and R. Marsilio. A viscous inverse method for aerodynamic design. Computers & Fluids, 2006.
 125

3.4 Case Study 4 – Inversely Designed Scramjet Flow-Path (Chapter From

Hypersonic Vehicles - Past, Present and Future Developments)

Authors : Mookesh Dhanasar1, Frederick Ferguson1 and Julio Mendez2

Affiliations : 1North Carolina Agricultural and Technical State University, Greensboro, North
Carolina, USA
2Corrdesa LLC, Tyrone, Georgia, USA

Source : DOI: 10.5772/intechopen.85697

It can be argued that at the heart of functional hypersonic vehicle is its engine Invalid source
specified.51. Key to a functionally efficient scramjet engine lies in the design of its flow-path. The
flow-path is made up of the following sections:
➢ forebody inlet;
➢ isolator,
➢ combustor,
➢ nozzle.
Here we focuses on the design of the forebody inlet and the isolator sections of a scramjet engine. In
this framework, key to a functionally efficient scramjet engine lies in the design of its flow-path. This
flow-path design must consider a complex flow-field physics and the interaction of physical surfaces
with this complex flow-field. Many attempts to design efficient scramjet flow-paths have met with
some measured degree of success. This research uses a ‘inverse design’ approach, which is similar to
Darwin’s theory of evolution, where an organism adopts to survive in its environment; the scramjet
flow-path will be carved/extracted from the operational environment. The objective is to naturally
and organically capture, process and direct the flow from the environment; thus preparing it for the
combustion process. This approach uses the ideal 2D oblique shock relations, coupled with Non-
weiler’s caret wave rider theory and streamline marching techniques.

3.4.1 Introduction
Driven by the desire to improve air travel and shorten flight time, aircraft engines have evolved from
simple reciprocating internal combustion engines to advance axil flow jet engines. Jet engines fall
into several categories. These include air-breathing, turbine powered, turbojet, turbofan, ramjet
compression and scramjet compression engines. Ramjet and scramjet compression engines are
unique in that they represent the latest development on the evolutionary path of jet engines. The
ramjet, unlike conventional jet engines which uses turbine driven compressors to compress the
incoming air, uses shockwaves to achieve this goal. The compressed air is burnt in the combustor
under sub-sonic conditions. The scramjet is basically an air-breathing jet engine designed to fly at
hypersonic speeds between Mach 4 and 12 or speeds in the range of 1207–2995 m/s (2700–6700
mph).
A scramjet engine captures its airflow from the atmosphere and also compresses it across
shockwaves before the air enters the combustor. Fuel is injected into the combustor where
combustion occurs under supersonic conditions. The hot, high-pressure gas leaving the combustor is
then accelerated to high velocities in the nozzle to produce thrust as it exits the engine. Generally
speaking, the concepts associated with scramjet engines appear at first glance to be very simple. This
however is very misleading as attempts develop a working scramjet engine that has proven to be
quite an engineering challenge.

51Mookesh Dhanasar, Frederick Ferguson and Julio Mendez, “Inversely Designed Scramjet Flow-Path”, North
Carolina Agricultural and Technical State University, Greensboro, North Carolina, USA and Corrdesa LLC,
Tyrone, Georgia, USA, 2019.
 126

Several aspects of scramjet engine development are at various stages of development. These include
supersonic fuel-air mixing, aero-thermodynamic heat dissipation from both skin friction and internal
combustion, and other thermal management problems associated with operating an engine at
exceedingly high temperatures for extended periods of time. Combustion chamber components could
experience temperatures on the order of over 3033 K (5000°F). At these temperatures most metals
melt and fluids (air and fuel) ionize, making the physics of their associated behavior unpredictable.
This chapter focuses on the design concepts for the forebody, inlet, and isolator sections of an
innovative scramjet engine geometry and some of its flow physics.

Figure 3.4.1 Pod-Mounted Scramjet Concept

3.4.2 Inverse Scramjet 2D Centerline Design Approach

As stated earlier, the scramjet concept represents the latest evolution in the series of air-breathing
jet engines. Combustion in these engines occurs under supersonic conditions. Scramjet engines are
seen as the propulsion system that is at the heart of hypersonic vehicles/platforms. Every scramjet
conceptual engine design and engines flown to-date all have a common set of components or sub-
sections. Figure 3.4.1 presents these components/sub-sections for a pod-mounted conceptual
scramjet design. These components/sections are the forebody section, the inlet section, isolator
section, combustor section, and the diffuser-nozzle section. Ideally, the engine concept presented
should be able to function over a wide range of Mach numbers. This gives rise to the idea of a
morphing ramjet/scramjet or dual mode scramjet configurations as presented in Figure 3.4.252.
Figure 3.4.2 a, presents the dual mode scramjet engine, Figure 3.4.2 b, the pure scramjet mode
and Figure 3.4.2 c, the pure ramjet mode. A typical dual mode scramjet configuration as that
presented in Figure 3.4.4 was inversely carved out of supersonic and hypersonic flow-fields. The
design framework used in the design of the forebody, inlet and isolator sections forms the core of this
study.

3.4.3 Scramjet Inverse Design Approach

The inverse design approach relies on extracting the configuration of interest from the environment
in which it operates. For this design process the centerline geometry of a given 2D scramjet
configuration is explicitly constructed using the following design inputs:
• freestream Mach number M∞,
• scramjet forebody length L,
• shock angle β,

52Dhanasar M. Development of A Benchmark for the Design and Analysis of A Tip-To-Tail Ramjet-Scramjet
Propulsion System [dissertation]. Greensboro: North Carolina Agricultural and Technical State University; 2009.
 127

• caret angle α,
• cruising flight altitude H∞,
• and isolator backpressure ratio Pin/Pexit.

Figure 3.4.2 Dual-Mode Scramjet Concept

All freestream flow-field properties are extracted from the Mach number and altitude53-54. This
information is used in the construction, analysis and definition of the three fundamental aerodynamic
zones, namely; the ‘primary shock’ zone AB, the ‘reflected shock zone’, BC, and the ‘isolator zone’,
CD as presented in Figure 3.4.4. Also presented in Figure 3.4.4 is a 2D conceptual representation
of the flow-field physics associated with supersonic flow interaction over a wedge and in a constant
area duct. Details of this flow-field physics and its exploitation in the inverse design approach are
explained in the next section. Also addressed is the derivation of the actual 3-dimensional forebody
section. This is a two-step process, where in step one, the 2D construction of the ‘forebody’, domain
A-D, is conducted. Step two is where the 3D geometry is obtained.

53Anderson JD. Fundamentals of Aerodynamics. 3d ed. New York: McGraw-Hill; 2001.

54Heiser WH, Pratt DT. Hypersonic airbreathing propulsion. AIAA Education Series. Virginia, USA: American
Institute of Aeronautics and Astronautics; 1994. ISBN: 1-56347-035-7.
 128

Figure 3.4.3 Illustration of the Cross Section of the Scramjet

3.4.3.1 Aerodynamics of the 2D ‘Forebody’ Configuration

Consider the 2D cross-sectional illustration of the scramjet forebody-inlet-isolator section presented
in Figure 3.4.4. Now consider a supersonic flow travelling parallel to the x-axis of a 2D wedge.
Supersonic aerodynamics dictates that the flow is deflected first by the oblique shock wave, AB2,
originating from the leading edge, A, of the wedge. The flow is deflected a second time by a reflected
shock wave, B2C1 emanating from the cowl lip at point B2, of the inlet. The flow enters the isolator
duct and travels once more in a direction that is parallel to the x-axis. To ensure that the flow in the
isolator duct remain supersonic the freestream Mach number must be greater than 3.0 and the shock
wave angle β, greater than 12 and less than 30 degrees. The flow-field behavior within the isolator
duct is of paramount importance. This flow-field may consist of a system of oblique or normal shocks,
as visualized in Figure 3.4.4. Driving this behavior is the flow-field vicious interactions with the
isolator duct walls. The isolator’s non-dimensional length, L/H, and the pressure differential at the
duct’s entrance and exit also enhance the flow-field’s behavior.

Figure 3.4.4 Conceptual 2D Centerline Cross-Section of the Forebody-Inlet-Isolator Scramjet Section

with Flow Physics Representation
 129

3.4.3.2 Derivation of the 2D ‘Forebody-Inlet-Isolator’ Configuration

The ‘forebody-inlet-isolator’ concept presented in Figure 3.4.4 relies on determining the geometric
design points located at stations A, B, C and D, along the x-axis of the scramjet. This is accomplished
by use of the oblique shock relations described in55-56 and the ‘isolator’ relations that were
experimentally derived in 57-58. It is assumed that in Figure 3.4.4 the flow travels in the x-direction,
and that the construction of the ‘forebody’ configuration starts at design point, A. The following
account details the logic used to define the locations of design points A, B, C and D.
3.4.3.3 Design Point at Station A
The design point at station A is considered the origin of the scramjet design coordinate system,
therefore, design point A coordinates are evaluated as follows, Ax = 0, Ay = 0, and Az = 0.
3.4.3.4 Design Points at Station B
Using the input data, the location of design point B, can be computed with the use of the following
relations: Bx = L, By = 0, and Bz = 0. In addition, using trigonometric relationships design point B1 is
evaluated as follows: B1x = Bx, B1y = Bxtan(ϴ) and B1z = 0. The coordinates for design point B2 are
evaluated in the following manner: B2x = Bx, B2y = Bxtan(ϴ), and B2z = 0.
The wedge angle is represented by theta (ϴ) and the shock angle is represented by beta (β). Using
the Mach number and the shock angle beta (β), the wedge angle theta (ϴ) can be obtained with the
use of the Theta-Beta-Mach (ϴ-β-M) relationship given as seen in Error! Reference source not
found.). In Error! Reference source not found. the constant γ is set at a value of 1.4.

2
M∞ Sin2 β − 1
θ = atan {2cotβ [ 2 ]}
M∞ (γ + Cos2β) + 2
Eq. 3.4.1
3.4.3.5 Design Points at Station C
The design points at station C is extracted from the wedge angle ϴ, and the flow-field properties
behind the primary shock wave, AB2, as seen in Figure 3.4.4. Determination of the location of design
point C is a little more involved and is approached systematically as outlined in the following steps:
➢ First, the Mach number, M, behind the primary shock wave, AB2 (see Figure 3.4.4), is
obtained using,

(γ − 1)
1 {1 + [ 2 ] [MSin(β)2 ]}
M=[ ]
Sin(β − θ) [γMSin(β)2 − (γ − 1)/2]
Eq. 3.4.2
➢ This Mach number, coupled with the free stream parameters are then used with the oblique
shock relations derived in59 for the evaluation of all of flow field properties behind the

55 Ferguson F, Dhanasar M, Williams R, Blankson I, Kan kam D. Supersonic and hypersonic slender air-breathing
configurations derived from 2D Flow fields. 46th AIAA Aerospace Sciences Meeting and Exhibit; 7–10 January
2008; Reno, NV. DOI: 10.2514/6.2008-163.
56 Curran ET, Murthy SNB. Scramjet Propulsion. Vol. 189. Virginia, USA: AIAA; 2000. ISBN: 1-56347-322-4.
57 Billig FS. Research on supersonic combustion. Journal of Propulsion and Power. 1993;9(4):499-514.
58 Waltrup PJ, Billig FS. Prediction of recompression wall pressure distributions in scramjet engines. Journal of

Spacecraft and Rockets. 1973; 10(9):620-622.

59 Anderson JD. Modern Compressible Flow: With Historical Perspective. 3rd ed. New York, USA: McGraw-Hill;

2004.
 130

primary shock, AB2. The flow-field properties, pressure, P, temperature, T, density ρ, and total
pressure, Pt,2, are evaluated using Eq. 3.4.3 - Eq. 3.4.6.
➢
P 2γ(M∞ Sinβ)2 − (γ − 1)
=
P∞ (γ + 1)
Eq. 3.4.3

T [2γ(M∞ Sinβ)2 − (γ − 1)][(γ + 1)(M∞ Sinβ)2 + 2]

=
T∞ [(γ + 1)2 (M∞ Sinβ)2 ]
Eq. 3.4.4

ρ (γ + 1)(M∞ Sinβ)2
=
ρ∞ (γ − 1)(M∞ Sinβ)2 + 2
Eq. 3.4.5

𝛾 1
(γ + 1)(M∞ Sinβ) 2 (γ + 1)
p 𝛾−1 𝛾−1
=[ ] [ ]
p∞ (γ − 1)(M∞ Sinβ)2 + 2 2𝛾(M∞ Sinβ)2 − 𝛾 − 1
Eq. 3.4.6

➢ B2C1 as seen in Figure 3.4.4 represent the reflected shock wave. This reflected shock wave
is a the result of a flow-field behind the primary shock wave, AB2, with a supersonic Mach
number, M, once more being deflected by an imaginary wedge, with wedge angle ϴ at design
point B2. This imaginary wedge is oriented in such a manner that it ensures that the deflected
flow travels parallel to the x-axis, Figure 3.4.4. At this stage updated values for the wedge
angle ϴ and the Mach number, M, are obtained using Eq. 3.4.1 and Eq. 3.4.2. A reflection
shock angle is now be defined as ϕ = β1- ϴ. In this expression, β1 is the reflected shock angle.
This reflected shock angle is generated by the interaction of the flow-field with Mach number
M and the imaginary wedge with angle ϴ. Note that β1 is obtained using Eq. 3.4.1 and
replacing the value of the freestream Mach number, M∞, with that of the supersonic Mach
number, M.
➢ The flow-field properties behind the reflected shock wave B2C1 are now obtained in a similar
manner as described in ‘b’ above. Eq. 3.4.2 is used to obtain M1, which is the Mach number
behind the reflected shock. In Eq. 3.4.2 the freestream Mach number, M∞, is replaced with
Mach number M. It is very important to note here that M1 represents the Mach number at the
entrance to the isolator section of the scramjet. Eq. 3.4.3 - Eq. 3.4.6 are used to derive the
additional flow-field properties of pressure, temperature, density and total temperature, p1,
T1, ρ1 and To, behind the reflected shock. Note that in these equations the value for the
freestream Mach number, M∞, is now replaced with the value of the Mach number, M, from
the flow-field properties behind the primary shock.
➢ Having obtained the parameters ϴ, β and β1 all design points at station C can now be derived.
The y-coordinate and z-coordinate are defined as Cy = 0, and Cz = 0, respectively. The x-
coordinate is obtained with the help of trigonometric relations, and is defined as:
 131

tan(β) − tan(θ)
Cx = [1 + ]B
tan(θ) − tan(β1 − θ) x
Eq. 3.4.7
➢ The coordinates of point C1 are determined as follows: C1x = Cx, C1y = Cxtan(ϴ), and C1z = 0.
➢ Similarly, the coordinates of point C2 are determined from: C2x = Cx, C2y = B2y, and C2z = 0.
3.4.3.6 Design Points at Station D
The evaluation of the coordinates of the design points at station D is also a multistep process.
➢ First a non-dimensional expression for the ‘normal total’ pressure value, Pn, in, is derived, Eq.
3.4.8. This expression is a function of isolator entrance conditions, where M1 is treated as the
Min, and the static pressure, P1, as Pin. Note here that the values of M1 and P1 are obtained
from the flow-field properties behind the reflected shock B2C1.

Pn,in 2γM12 − (γ − 1)
=[ ]
Pin (γ + 1)
Eq. 3.4.8
In determining the isolator length for a design process, the ratio of the entrance to exit pressures,
Pin/Pout, over the range between Pin and Pn,in has to be evaluated. This value is needed to determine
the length of an isolator that can reliably prevent all ‘unstart’ conditions. In this design process, the
ratio, Pout/Pn,in, representing the isolator exit pressure, Pout, to the ‘normal total’ pressure value, Pn,in,
is prescribed. Using this approach, the value for Pin/Pout can be determined by using :

Pout pout pn,in

=( )( )
Pin pn,in pin
Eq. 3.4.9

➢ The system of 1D conservation laws result in the following expression for the isolator exit
Mach number, Mout 60- 61;

2 2 −0.5
γ2 Min [1 + ((γ − 1)/2)Min ] γ−1
Mout ={ 2 − }
(1 − γMin − pout /pin )2 2
Eq. 3.4.10
Similarly, with the exit Mach number known, the non-dimensional length of the isolator can be
evaluated based on the following experimental relationship developed in 62-63 :

L √θ/H 50[(pout /pin ) + 1] + 170[(pout /pin ) − 1]2

( ) = { 2 }
H Isolator Reθ1/4 Min −1
Eq. 3.4.11

60 Billig FS. Research on supersonic combustion. Journal of Propulsion and Power. 1993;9(4):499-514.
61 Waltrup PJ, Billig FS. Prediction of recompression wall pressure distributions in scramjet engines. Journal of
Spacecraft and Rockets. 1973; 10(9):620-622
62 Billig FS. Research on supersonic combustion. Journal of Propulsion and Power. 1993;9(4):499-514.
63 Waltrup PJ, Billig FS. Prediction of recompression wall pressure distributions in scramjet engines. Journal of

Spacecraft and Rockets. 1973; 10(9):620-622.

 132

where Reϴ is the inlet Reynolds number based on the momentum thickness. Also, the symbol, H,
represents the isolator height that is determined from the y-coordinates of points C2 and C1, in a
manner such that, H = C2y – C1y.
➢ The coordinates of point D are computed as follows: Dx = Cx + LIsolator, Dy = 0, and Dz = 0.
➢ The coordinates of point D1 are computed as follows: D1x = Dx, D1y = C1y, and D1z = 0..
➢ The coordinates of point D2 are computed as follows: D2x = Dx, D2y = C2y, and D2z = 0.

Finally, with the coordinates of all the design points at all stations, A, B, B1, B2, C, C1, C2, D, D1, and D2,
fully defined, the sketch illustrated in Figure 3.4.4 can be constructed.

3.4.4 Computer Aided Design (CAD) Design

3.4.4.1 Overview of the 3D Design Process
The 3D design process has as its origin in the inversely design two-dimensional geometry extracted
from a 2D hypersonic flow-field. This is then coupled with the Nonweiler’s wave rider approach 64 of
inversely carving stream surfaces from inviscid flow-fields. A caret wave rider, Figure 3.4.5, is
chosen as an example because it represents a 3D geometry that was obtained from a 2D flow-field.

Figure 3.4.5 Nonweiller Caret Wing Wave Rider Configuration

This caret wave rider geometry is constructed from a single planer shock wave, AB3B4, as seen in
Figure 3.4.5. A unique feature of this construction process is that at any cross-section of the wave
rider geometry there is a wedge that is supported by an oblique shock wave, with these wedges being
parallel to the flow.

64Nonweiler TRF. Aerodynamic problems of manned space vehicles. Journal of the Royal Aeronautic Society.
1959;63:521-528.
 133

In reality, the caret wave rider is carved from an inverse design approach that relies on the inviscid
streamline principle. This principle states that any inviscid streamline can be replaced by a solid wall.
The principle also states that replacing the inviscid streamline with a solid wall has no effect on the
external flow. Planar inviscid stream surfaces are formed from these inviscid streamlines. These
inviscid stream surfaces are then brought together to construct 3D inviscid wave rider geometries
and stream tubes. An examination of Figure 3.4.5 demonstrates how the streamlines form planar
stream surfaces, such as, upper inviscid surfaces, ABB3 and ABB4, or lower stream surfaces, such as,
AB1B3 and AB1B4. This approach is further explained the next sub-section and is demonstrated by the
construction of a supersonic 3D wedge followed by a 3D supersonic caret shaped geometry. This
caret-shaped geometry will then be used to generate supersonic star-shaped geometries of interest.
3.4.4.2 2D Forebody Construction (Side View)
The review of begins with the construction of the supersonic wedge. Established ideal oblique 2D
shockwave relationships are used to construct the supersonic 2D forebody. There are two ideal
oblique shock relationships which can be used, the Theta-Beta Mach relationship, or the Beta-Theta
Mach relationship. In this review, the Theta-Beta Mach65-66-67 relationship, described earlier to be
used in the construction of the supersonic 2D forebody. For a prescribed Mach number, shock angle,
Beta, at a given altitude, a wedge angle, Theta, is extracted. The next step is to set a forebody length.
Having all the geometric data the 2D forebody with the attached shock is constructed as presented

Figure 3.4.6 Preparation for Extracting Information for 2D Base View

65 Anderson JD. Fundamentals of Aerodynamics. 3d ed. New York: McGraw-Hill; 2001.

66 Anderson JD. Hypersonic and High Temperature Gas Dynamics. American Institute of Aeronautics and
Astronautics. Virginia, USA: McGraw-Hill; 1989.
67 Anderson JD. Modern Compressible Flow: With Historical Perspective. 3rd ed. New York, USA: McGraw-Hill;

2004.
 134

in Figure 3.4.6.
3.4.4.3 2D Inlet Construction (Side View)
The inlet construction is an extension of the 2D forebody construction. The oblique shock AB hits the
cowl lip at point B and is reflected as shown in Figure 3.4.6. Line BC represents the reflected shock
from the interaction of the oblique shock and the cowl lip. Ideal oblique shock relationships are used
to determine the reflected shock angle, Beta reflected. Note that the line AB1 which represents the
lower surface of the forebody continues to point C where it intersects with line BC. At this point in
the design process the 2D forebody and the inlet are constructed.
3.4.4.4 Streamline Preparation of Flow-Field
A 2D base view of the forebody-inlet components are constructed from geometric information
obtained from the 2D side view. The streamline cross-marching method used preserves both the
geometric information and the 2D flow-field information. The oblique shockwave, line AB, is first
divided up into N number of equal parts, in this case six, as seen in Figure 3.4.6. Streamlines are
then constructed emanating from the oblique shockwave. Each streamline has a starting point on the
oblique shockwave, and ends on the reflected shockwave, line BC as presented in Figure 3.4.6. The
longest streamline is represented by line AC and is the lower surface of the forebody-inlet. The
shortest streamline is represented by point 6; here the streamline starts and stops at the same point.
The streamlines emanating from the oblique shockwave and ending on the reflected shockwave
travel parallel with respect to the lower surface of the forebody-inlet as presented in Figure 3.4.6.
All streamlines are now processed by the reflected shockwave, BC, and travel parallel to the surfaces
beginning at points C and B as shown in Figure 3.4.6. The 2D base view can be extracted from the
flow field.
3.4.4.5 Wedge Geometry Extraction From Flow-Field
The base view for the 2D wedge is now extracted for the 2D forebody-inlet and the associated 2D
flow-field. A zy-coordinate system is set up and a wedge width is prescribed. Streamlines emanating
from the reflected shockwave, BC, are now mapped onto the zy-coordinate system as presented in
Figure 3.4.7. Having completed the construction of the 2D side view and the 2D base view, the
designer now has that can be used to generate the 3D forebody-inlet geometry for a 3D wedge.

Figure 3.4.7 Generation of 2D Base View for a Wedge

 135

3.4.5 The Caret Geometry

The caret geometry forms the basis of the design of the star shaped geometries in this study. A similar
process is used to obtain the caret-shaped 2D base view. Now instead of providing a wedge width, a
star angle, Phi, is provided as presented in Figure 3.4.8. For the four-point-star, Phi is 45 degrees.
Reflecting points AB Point C, about the z-axis will generate the 2D base view for the caret-shaped
wave rider geometry. As before, all data required for the 3D construction of the 3D caret shaped
forebody-inlet have been extracted from the flow-field.

Figure 3.4.8 Generation of 2D Caret-Shaped Geometry, Base View

3.4.5.1 3D Stream Tube Construction using the Wave Rider Approach

The scramjet forebody-inlet-isolator design concept as being proposed suggests a new use for wave
rider geometries. Here, the focus is not only of the wave rider shape, but also on the external flow-
field supporting the wave rider configuration. As seen in Figure 3.4.5, attention is on the external
2D flow on the wave rider lower surfaces, that is, AB1B3 and AB1B4, and the flow entering and exiting
the planes, AB3B4 and B1B3B4. With this alternative perspective, the innovation lies in the fact that
the flow moving across the lower surface of the wave rider is treated as the flow entering a stream
tube through surface, AB3B4 and leaving through the plane, B1B3B4. Recall at this point that the flow-
field is two dimensional, confined to the xy-plane and can be treated as a collection of 2D slices that
are parallel to each other. The flow within the stream tube is bounded by the lower inviscid
surfaces,AB1B3 and AB1B4 and an imaginary line surface, B3B4. (see the development in [Dhanasar et
al.]68.
A completed stream tube consisting of the forebody-inlet-isolator sections is presented in Figure
3.4.9. This stream tube is carved/extracted from a supersonic flow-field travelling parallel to the x-
axis, which is compressed by two oblique shock waves; resulting in the flow once again traveling in
a direction parallel to the x-axis. Further examination of Figure 3.4.9 identifies the primary shock
wave plane as AB3B4, which supports two compression surfaces, ACB3 and ACB4. At this stage the flow
field is no longer parallel to the x-axis. A reflected shock wave is constructed to form the plane, CB3B4.

68Mookesh Dhanasar, Frederick Ferguson and Julio Mendez, “Inversely Designed Scramjet Flow-Path”, North
Carolina Agricultural and Technical State University, Greensboro, North Carolina, USA and Corrdesa LLC,
Tyrone, Georgia, USA, 2019.
 136

Figure 3.4.9 Wave Rider Derived Stream Tube

This specially designed plane, CB3B4, now straightens the flow leaving the shock surface, CB3B4, so
that it once again travels parallel to the x-axis. The reflected flow now forms the stream tube
comprising of the following planar surfaces, CDD3B3, CDD4B4, and B3B4D4B3.
3.4.5.2 Transforming Stream Tubes to ‘Star’ Shaped Geometries
The preceding section saw the design of a single stream tube. These single stream tubes can now be
used to create star-shaped geometries of interest, an example of which is presented in [Dhanasar et
al.]69. It is a four point star geometry, so termed because it is a collection of four stream tubes that is
assembled in a manner to create a ‘closed form’ geometry of interest. The fundamental concept in
moving from a 2D geometry, Figure 3.4.4, to the 3D geometries, (see [Dhanasar et al.]70), lies mainly
on identifying the coordinates along the z-axis. Determination of the location of points, B3, B4, D3 and
D4, is of significant importance. These points are responsible for the development of a closed form
geometry/closed tube with the ability of preserving the aerodynamics associated with the inviscid
flow-field behavior. Additionally, the ‘y’ and ‘z’ coordinates of these points rely of the choice of angle
α, an example of which is the angle D3DD4 as seen. In generating the four point star configuration the
angle α is set to 90 degrees.
3.4.5.3 Validation Section
This section focuses on the validation of the forebody-inlet-isolator sections associated with the
proposed scramjet engine concept.
3.4.5.4 2D Simulations
Both Euler and viscous studies were conducted on the scramjet forebody, inlet, and isolator sections.

69 Mookesh Dhanasar, Frederick Ferguson and Julio Mendez, “Inversely Designed Scramjet Flow-Path”, North
Carolina Agricultural and Technical State University, Greensboro, North Carolina, USA and Corrdesa LLC,
Tyrone, Georgia, USA, 2019.
70 Mookesh Dhanasar, Frederick Ferguson and Julio Mendez, “Inversely Designed Scramjet Flow-Path”, North

Carolina Agricultural and Technical State University, Greensboro, North Carolina, USA and Corrdesa LLC,
Tyrone, Georgia, USA, 2019.
 137

2D Euler flow studies were conducted using the air vehicles unstructured solver (AVUS) 71. AVUS is
a three-dimensional finite volume unstructured-grid Euler/Navier-Stokes flow solver. 2D isolator
viscous simulations were conducted using an in-house computational scheme, the integral
differential scheme (IDS)72. The following contour plots (Figure 3.4.10 & Figure 3.4.11) represent
the solution of the AVUS software, whose units are in the SI. The IDS is built on the premise of
reducing numerical and modeling errors. As such, the IDS implements the dimensionless form of the
Navier Stokes equations, and therefore it reduces the round-off error.

Figure 3.4.10 AVUS Euler Results Density Contours

3.4.5.5 2D Euler Simulations Using AVUS

The four-point star configuration, was selected as the test case. The scramjet forebody-inlet-isolator
model was exposed to a Mach 5 hypersonic freestream flow-field at a zero angle of attack. Figure
3.4.10 - Figure 3.4.11 presents the 2D Euler simulation results along the centerline of the four point
star configuration. On examining these figures, the following observations are made. Figure 3.4.10
presents velocity distribution data for the geometry, where it is observed that the behavior of the
flow imitates the conceptual flow-field presented in Figure 3.4.4. That is, freestream flow is first
processed by the 2D oblique shock, travels parallel to the wedge surface, is processed again by the
reflected shock and travels parallel to the isolator duct walls. Once more we observe the organized
nature of the 2D flow, which is supported by the constant property in the respective zones. The
development of the shock train within the isolator duct is also captured in these figures.
3.4.5.6 2D Isolator Viscous Simulations
A 2-D viscous simulation was conducted on the isolator section using the integral differential scheme
(IDS), currently under development at North Carolina Agricultural & Technical State University. At
the heart of the IDS numerical scheme is the unique combination of both the differential and integral
forms of the Navier-Stokes equations (NSE). The differential form of the NSE is used for explicit time
marching, whereas integral form of the NSE is used to evaluate the spatial fluxes. The IDS scheme has

71 Air Force Research Laboratory, CFD Research Branch, Wright-Patterson AFB, OH, Air Vehicles Unstructured
Solver (AVUS).
72 Ferguson F, Mendez J, Dodoo-Amoo D. Evaluating the hypersonic leading edge phenomena at high Reynolds

and Mach numbers. Recent Trends in Computational Science and Engineering. Rijeka, Croatia: IntechOpen;
2017.
 138

Figure 3.4.11 Forebody-Inlet-Isolator Validation Study with 2D Slices

the ability to capture the complex physics associated with fluid flows. It does this by using a ‘method
of consistent averages’ (MCA) procedure which ensures the continuity of the numerical flux
quantities. The objective of this initial simulation was to observe the flow behavior. Further details
on the physics and computational numerical scheme associated with the IDS can be found in73. Flow-
field properties presented include Mach number distribution, pressure distribution, density
distribution, and temperature distribution. Examination of theses flow-field properties supports the
fact that the flow-field is behaving in a manner as it was designed to.
3.4.5.7 3D Simulations
3D computational simulations were also conducted on the scramjet forebody, inlet, and isolator
sections. Computational tools used were Fluent and AVUS. In the case of the 3D Euler computational
simulation, a single 3D stream tube, Figure 3.4.9, was exposed to a Mach 6 flow-field. A similar
process was implemented with AVUS and 2D slices of flow-field data are extracted and presented in
Figure 3.4.12. On examining these 2D slices of 3D data, it is observed that the stream tube is
processing the flow in an organized consistent manner that is aligned with its design. Arguable, this
is an Euler analysis, however it is worth pointing out that the stream tube 2D design process holds.
This is further supported by Error! Reference source not found. which presents data on the Mach
distribution at the isolator exit.

73Ferguson F, Mendez J, Dodoo-Amoo D. Evaluating the hypersonic leading edge phenomena at high Reynolds
and Mach numbers. In: Recent Trends in Computational Science and Engineering. Rijeka, Croatia: IntechOpen;
2017.
 139

Figure 3.4.12 Centerline 2D Pressure Contours

Figure 3.4.13 Mach Contours at the Isolator Exit

3.4.5.8 Combustor Diffuser Nozzle Sections

A dual mode scramjet configuration, as presented in Error! Reference source not found., in addition
to having a four-section consisting of a forebody, inlet, isolator sections; also has an aft-section
consisting of a combustor, diffuser and nozzle sections. Whilst the focus of this chapter has been on
the design of the fore-section, a brief discussion on the design of the aft-section and its integration is
warranted for the sake of completion in the design of the dual mode scramjet. Design of the aft-
section focuses on four fundamental design sections; a transition section, a combustor section, a
diffuser section and a nozzle section, Figure 3.4.14 and Figure 3.4.15. The transition section, as
implied in the name, is designed to prepare the Error! Reference source not found., flow before it
enters the combustor. This section takes the flow leaving the isolator duct and guides it towards
 140

Figure 3.4.14 2D-3D Geometric Construction from Prescribes Isolator Cross-Section and
Aerodynamic Inputs

combustor. The primary design goal is to ensure that the flow entering the combustor is as organized
as possible. The combustion section is where fuel is added, mixed and burned. The diffuser section
is used to help control the combustion process as the scramjet operates across its dual mode, that is,
switching from ramjet mode to scramjet mode. The nozzle section is used to accelerate the exhaust
gases as the flow leaves the dual mode scramjet.
3.4.5.9 Scramjet Flow-Path
A completed scramjet flow-path can now be obtained with the assembly of both the forebody-inlet-
isolator and combustor-nozzle sections. Two samples are presented in Figure 3.4.15-Figure
3.4.18 presents a scramjet that has a square combustor configuration and Error! Reference source
not found., a circular combustor configuration. Referring back, one can observe that a variety of
geometries can be Figure 3.4.15 generated by manipulating the design points A through H, and the
design variables, x3–x11, in any combination.

3.4.6 Conclusion
In this chapter we explored an inverse design approach used in designing scramjet configurations.
The forebody, inlet and isolator sections formed the core focus of the chapter. Ideal oblique 2D shock
relations along with Billig’s isolator relations were first used to generate a centerline 2D geometry.
Streamline marching techniques coupled with Nonweiler’s caret wave rider theory were used to
geometrically construct 3D stream tubes. These 3D stream tubes were later used in the construction
of various star-shaped forebody-inlet-isolator sections. Initial 2D Euler and 2D viscous studies were
performed on the forebody-inlet-isolator sections and the results presented. Initial 3D Euler studies
were also conducted on a single 3D stream tube. All results presented demonstrated the uniform
nature of the flow-field within the stream tube, supporting the inverse design approach. A 3D viscous
analysis of the 3D stream tube is yet to be undertaken.
 141

Figure 3.4.15 Illustration of the Transition-Combustor-Nozzle Element

Figure 3.4.16 A Dual Mode Ramjet-to-Scramjet Concept – Courtesy of Dhanasar M

Figure 3.4.17 4-Pts Scramjet with Circular Combustor C-Sections

Figure 3.4.18 4-Pts Scram Jet with Square Combustor C-Sections

142
 143

4 Structural Design Perspectives of Aircrafts

4.1 Generalities and Load Factor
The structure of an aircraft must be strong enough to carry all the loads to which it might be
subjected, including the repeated small to medium loads experienced in normal flight and the big
loads experienced during extreme conditions (Megson )1. To fly an airplane's exterior must have an
aerodynamic shape Into this shape must be fitted members having a high strength-to-weight ratio
that arc capable of sustaining the forces necessary to balance the airplane in flight. Airplanes are
generally designed for a specific purpose that dictates the structural design required. The airplane
structure must be capable of withstanding much more force than that imposed by its own weight.
When the purpose of a particular design is established, it provide structure according to strict
standards established by the Federal Aviation Administration to ensure safety. In general, airplanes
are designed to withstand 1.5 times the maximum expected forces. To be certified by the Federal
Aviation Administration, the structural strength (load factor) on airplanes must conform with the
standards set forth by Federal Aviation Regulations.
The loads imposed on the wings in flight are stated in terms of load factor. Load factor is the ratio of
the total load supported by the airplane's wing to the actual weight of the airplane and its contents,
i.e.. the actual load sup-ported by the wings divided by the total weight of the air-plane. For example,
if an airplane has a gross weight of 2000 lb (907 kg) and during flight is subjected to aerodynamic
forces that increase the total load the wing must support to 4000 lb (1814 kg), the load factor is 2.0
(400W2000 = 2). The airplane wing is producing lift that is equal to twice the gross weight of the
airplane. Another way of expressing load factor is the ratio of a given load to the pull of gravity, to
refer to a load factor of 3 as "thee g's," where g refers to the pull of gravity. In this case the weight of
the airplane is equal to I g. and if a load of three times the actual weight of the airplane were imposed
upon the wing due to curved flight, the load factor would be equal to three g's. Refer to Aircraft Basic
Science for more information. All airplanes are designed to meet certain strength requirements,
depending upon the intended use of the air-plane (Egbert Torenbeek)2. In addressing any problem
in continuum or solid mechanics, three factors must be considered3:
1. Newtonian equations of motion, in the more general form recognized by Euler, expressing
conservation of linear and angular momentum for finite bodies (rather than just for point
particles), and the related concept of stress, as formalized by Cauchy,
2. geometry of deformation and thus the expression of strains in terms of gradients in the
displacement field, and
3. Relations between stress and strain that are characteristic of the material in question, as well
as of the stress level, temperature, and time scale of the problem considered.
These three considerations suffice for most problems. They must be supplemented, however, for
solids undergoing diffusion processes in which one material constituent moves relative to another
(which may be the case for fluid-infiltrated soils or petroleum reservoir rocks) and in cases for which
the induction of a temperature field by deformation processes and the related heat transfer cannot
be neglected. These cases require that the following also be considered:
4. Equations for conservation of mass of diffusing constituents,
5. First law of thermodynamics, which introduces the concept of heat flux and relates changes
in energy to work and heat supply, and
6. Relations that express the diffusive fluxes and heat flow in terms of spatial gradients of
appropriate chemical potentials and of temperature.
To examine the mathematical structure of the theory, considerations (1) to (3) above will now be
further developed. For this purpose, a continuum model of matter will be used, with no detailed
 144

reference to its discrete structure at molecular, or possibly other larger microscopic, scales far below
those of the intended application.

4.1.1 References
1 T. H. G. Megson, “Aircraft Structures for engineering students”, 4th Edition, Elsevier Ltd., 2007.
2 Egbert Torenbeek, “Advanced Aircraft Design Conceptual Design, Analysis and Optimization of

Subsonic Civil Airplanes”, A John Wiley & Sons, Ltd., Publication, 2013.
3 Mechanics of solids, Encyclopedia Britannica, 2018.

4.2 Composite Material in Aircraft Structure

Composite structures can lower the weight of an airliner significantly74. The increased production
cost, however, requires the application of cost-effective design strategies, and investigated by
[Kaufmann et al.75. Hence, a comparative value is required which is used for the evaluation of a design
solution in terms of cost and weight. The direct operating cost (DOC) can be used as this comparative
value; it captures all costs that arise when the aircraft is own. In this work, a cost/weight optimization
framework for composite structures is proposed. It takes into account manufacturing cost, non-
destructive testing cost and the lifetime fuel consumption based on the weight of the aircraft, thus
using a simplified version of the DOC as the objective function. First, the different phases in the design
of an aircraft are explained. It is then focused on the advantages and drawbacks of composite
structures, the design constraints and allowable, and non-destructive inspection. Further, the topics
of multi-objective optimization and the combined optimization of cost and weight are addressed.
Manufacturing cost can be estimated by means of different techniques; here, feature-based cost
estimations and parametric cost estimations proved to be most suitable for the proposed framework.
We comprises a parametric study in which a skin/stringer panel is optimized for a series of
cost/weight ratios (weight penalties) and material configurations. The weight penalty, defined as the
specific lifetime fuel burn, is dependent on the fuel consumption of the aircraft, the fuel price and the
viewpoint of the optimizer. It is concluded that the ideal choice of the design solution is neither low-
cost nor low-weight but rather a combination thereof. Next, we suggests the inclusion of non-
destructive testing cost in the design process of the component, and the adjustment of the design
strength of each laminate according to the inspection parameters. Hence, the scan pitch of the
ultrasonic testing is regarded as a variable, representing an index for the (guaranteed) laminate
quality. It is shown that the direct operating cost can be lowered when the quality level of the
laminate is assigned and adjusted in an early design stage.

4.3 Design Phases

The design of aircrafts can roughly be divided into three design phases, the conceptual design phase,
the preliminary design phase and the detail design phase (Kaufmann et al.)76. They all have a
distinctive multidisciplinary character, and the resulting design is a trade-off made by all
contributing engineers.

4.3.1 Conceptual Design

In the conceptual design phase, numerous design alternatives are compared and evaluated, based on

74 This thesis is mailed outside Sweden and an explanation of the Swedish Licentiate might be necessary. Here,
we have an intermediate academic degree called Licentiate of Technology. This is a degree that can be obtained
half-way between an MSc and a PhD. The examination is less formal than for a PhD but it requires the
completion of a thesis and a public seminar.
75 M. Kaufmann, D. Zenkert and P. Wennhage, “Cost/Weight Optimization of Aircraft Structures”, KTH School of

Engineering Sciences, Licentiate Thesis, Stockholm, Sweden 2008.

76 See Previous.
 145

cost/weight/range trade off studies. The result is an initial aerodynamics and propulsion concept,
including overall dimensions, weights and global loads. A notable study would be the work done by
[Mansouri]77 in regard to structural sizing of within the conceptual design stage. Others, the work
done by [Komarov et al.]78 and [Torenbeek]79.

4.3.2 Preliminary Design

A good overall structural concept is initiated during preliminary design80. At the very beginning of a
preliminary design effort, the designer writes a set of specifications consistent with the needs. It
should be clearly understood that during preliminary design it is not always possible for the designer
to meet all the requirements of a given set of specifications. In fact, it is not at all uncommon to find
certain minimum requirements unattainable. It is then necessary to compromise. The extent to which
compromises can be made must be left to the judgment of the designer. However, it must be kept in
mind that to achieve a design most adaptable to the specified purpose of the airplane, sound
judgment must be exercised in considering the value of the necessary modifications and/or
compromises (Niu)81. In the preliminary design phase, a global finite element model is built up from
which local loads and loading conditions are derived. An illustration of typical loads on an airliner is
given in Figure 4.3.1. As can be seen, aero-elastic loads such as tensile, compressive and torsional
loads in wing, fuselage and empennage represent only a fraction of the load cases the structural
engineers have to consider. Other loads arise from the cabin pressure (hoop stress), bird strike into
the cockpit, the trailing edge or the empennage; impact loads on the tail are a result of an abrupt take-

Figure 4.3.1 Typical Load of Airliner (Courtesy of Airbus)

77 Reza Mansouri, “Structural Analysis at Aircraft Conceptual Design Stage”, Master of Science In Aerospace
Engineering, The University of Texas at Arlington, May 2014.
78 Komarov Valeriy A. Borgest Nikolay M. Vislov Igor' P. Vlasov Nikolay V. Kozlov Dmitriy M. Korolkov Oleg N.

Maynskov Vladimir N, “Conceptual Aircraft Design”, The Ministry of Education and Science of the Russian
Federation, Samara State Aerospace University, 2011.
79 Egbert Torenbeek, “Advanced Aircraft Design - Conceptual Design, Analysis and Optimization of Subsonic Civil

Airplanes”, Delft University Of Technology, The Netherlands, A John Wiley & Sons, Ltd., Publication, 2013.
80 Markus Kaufmann, “Cost/Weight Optimization of Aircraft Structures”, KTH School of Engineering Sciences,

Licentiate Thesis, Stockholm, Sweden 2008.

81 Michael Chun-Yung Niu, “Airframe Structural Design”, Conmilit Press Ltd. 1988.
 146

off or a too inclined landing. Note the very high local stresses that can be found in the landing gear
ribs, the side-stay fittings and the pylon structure. The task of the structural engineers is then to
design the inner structure of the aircraft. The design is basically constrained by the aerodynamic
configuration; it has to withstand all loads and should be as light as possible. Different levels of
detailed are investigated in the preliminary design phase, see Figure 4.3.2. First, the structural
arrangement of the major parts, such as ribs and spars in the wing and lap joints and butt joints in

Figure 4.3.2 Levels in Detailed (Courtesy of H. Assler, Airbus Deutschland GmbH)

the fuselage have to be defined. Then, the structure is designed on a panel level. The strength and
stiffness of the structural members are defined and verified by means of finite element models, while
changes in the configuration (e.g. stiffener distance, rib stiffeners, etc.) are still possible. According
to [Assler]82, the design process is nuanced by a variety of factors at this stage. Examples are:
➢ Airworthiness regulations
➢ Environmental considerations
➢ General aircraft requirements (mission profile, maintenance, DOC, etc.)
➢ Specific requirements for structural details
➢ Available materials and technologies
➢ Manufacturing capacities and capabilities
➢ Non-destructive testing and investigation capabilities

H. Assler. “Design of aircraft structures under special consideration of NDT”. Jorg Volker, editor, 9th European
82

NDT Conference (ECNDT), Berlin, 2006.

 147

➢ Design costs

4.3.3 Detail Design

In the detail design phase, the structure is analyzed by means of high-fidelity models, and the
fabrication, tooling and assembly processes are defined. The result is a detailed work breakdown
structure including all structural parts, mounting, bolts and rivets, clips, doors, brackets, etc. Every
part has to fulfill its particular requirements, based on structural failure, fatigue, corrosion resistance,
lightning strike, sealing, conduction, maintenance or testing.

4.4 System Integration

As the development of a new aircraft type involves
great costs and, in turn, a great financial risk.
Aircrafts are not designed, fabricated and
assembled by one single manufacturer anymore;
the latter have been replaced by consortiums of
system integrators and suppliers (Kaufmann, et al.)
83. In Figure 4.3.1, the supply hierarchy in the

commercial aerospace industry is shown. On the

top of the supply pyramid, a system integrator is
responsible for the coordination, the overall design Figure 4.3.1 Supply Hierarchy in the
and the final assembly of the aircraft. In a Commercial Aerospace Industry
subordinate position, other members of the
consortium take part in the development of sub-systems (such as wing structures), its
manufacturing, sub-assembly and delivery. Further down in the hierarchy are equipment and
component supplier. They do not take part in the design process and are restricted to the
manufacturing and delivery of components. Estimated profit margins are the highest in the top
position of the pyramid and, accordingly, the higher the position in the hierarchy, the greater the risk
for that company in the overall project. The embedding of a subsystem supplier (e.g. Saab,
Bombardier or Alenia) occurs already in the concept design phase, sometimes already within pre-
development or research projects. The benefits of this approach are the early exchange of knowledge
and experience and a good integration of the otherwise widespread design teams. As the risks are
shared, the partners are under pressure to continuously increase the efficiency in terms of cost,
weight and producibility. The main drawbacks are communication problems across the company
borders, such as cultural differences, hierarchical misunderstandings and delayed information flows;
further mentioned is the need for an extensive product lifecycle management database (PLM) in the
case of the A380, the PLM is used by more than 5000 engineers and contains more than 3000 CAD
drawings of about 150'000 parts.

4.5 Design of Composite Structures

The fuel consumption stands for a considerable part of the operating cost of an airliner. The fuel
costs can be reduced by the development of more efficient engines, by the minimization of the
aerodynamic drag, by optimizing the light trajectory of the aircraft, or by reducing its mass. The latter
was the main motive to change from metals to composite materials, and the portion of composite in
the total structural weight is continuously increasing, see Figure 4.5.1.
The first application of composites to commercial aircrafts were radar domes in 1940. Since then,
composite materials gradually replaced their metallic counterparts. In 1975, NASA developed a

83M. Kaufmann, D. Zenkert and P. Wennhage, “Cost/Weight Optimization of Aircraft Structures”, KTH School of
Engineering Sciences, Licentiate Thesis, Stockholm, Sweden 2008.
 148

Figure 4.5.1 Portions of Composite material in Airbus Aircrafts (Courtesy of Airbus Industries)

series of composite parts for research purposes, and the elevators of the B727 and B737 and the
vertical fin of the DC10 were redesigned. Secondary structure (i.e. the leading edge, trailing edge,
aps, ailerons and rudder) was made of carbon fibers in the Boeing 777, and with the center wing box
of the A380, composites were even used for primary (load-carrying) structures. The latter enabled
weight savings of 1500 kilogram compared to the aluminum baseline.
In Figure 4.5.2, the specific stiffness’s
and the specific strengths of different
materials are presented84. One can see
that the specific strength of a carbon
fiber reinforced plastic is much higher
than the respective values of
aluminum or titanium, whereas the
specific moduli are approximately the
same. A composite component that is
designed for stiffness will therefore
have a higher safety factor against
material failure than its metallic
counterpart. This characteristic
accounts for the good fatigue behavior
of composites, cutting down the
maintenance cost for the airlines.
Another advantage is the possibility to
Figure 4.5.2 Specific Strength and Stiffness of Different
tailor composites specifically for a Metals and Alloys (Courtesy of Kaufmann)
desired function. This can be done by

84quasi-isotropic glass fiber reinforced plastic (Glass/QI) and quasi-isotropic carbon fiber reinforced plastic
(Carbon/QI).
 149

either adjusting the fiber angle distribution or by unifying the form and thus reducing the number of
parts.

4.6 Constraints and Allowable

The structure has to withstand given external loads. For composites loaded in tension, material
failure might be the limiting constraint; for composites loaded in compression, material failure or
stability concerns form the topology of the structure. The situations under which the integrity of the
structure needs to be proved are described in regulations published by the aviation authorities;
structural constraints, for example, are based on the airworthiness requirements, defined in JAR
25.613 Material Strength Properties and Design Values. This document is released by the Joint
Aviation Authorities, a European body representing the civil aviation regulatory authorities of a
number of European states. Similar regulations exist also in the United States (FAR), therefore it is
often referred to the FAR/JAR regulations. The determination of an allowable stress and strain limit
is based on the statistical evaluation of specimen tests. A typical design strength, for instance, is the
stress level where at least 90% of the population pass with a confidence of 95%. These tests have to
be performed for filled or unfilled holes, as stress concentrations around fastener and bolt holes can
be the cause for material failure.
A composite laminate can fail by different modes, e.g. by fiber failure, micro-buckling, matrix failure
or fiber/matrix deboning. Other effects are delamination’s due to pull-off loads, free-edge effects, a
poison’s ratio mismatch or compressive buckling. Therefore, most engineers (and aeronautical
engineers in particular) tend to use rather conservative failure criteria. Unlike metal structures,
composite structures are limited by strain and not by stress concerns. The strain limits of composite
laminates are only indirectly related to the strain levels of the matrix or the fibers; it is rather a design
strain based on coupon tests as stated above. One of the limiting load cases of coupon test is the
Compression After Impact (CAI) test, simulating tool drops and runway debris. There, the remaining
compressive failure strain of a damaged composite panel is evaluated. Being about 0.4%, this strain
level can be much lower than the one of the unmatched coupon. Although not being the most
elaborate failure theory, the maximum strain criterion is still widely used in the aerospace industry.
The maximum strain criterion is given as three independent sub-criteria

ε̂1,c < ε < ε̂1,t , ε̂2,c < ε < ε̂2,t , |γ12 | < γ̂12
Eq. 4.6.1
The indices c and t denote compression and tension, respectively; 1 and 2 denote the ply's
longitudinal and transversal direction, and ^ denotes the allowable strain value. Apart from material
failure, the design of composite laminates is governed by a series of other rules. Examples given are
the requirement for symmetrical stacking, a minimum amount of 10% of each ply angle, or the
sequence of the proper stack. Further, it has to be noted that aircraft engineers maintain a stacking
sequence consisting of 0, 90 degrees and +-45degree plies. Although there exist stacking sequences
that allow a significant weight reduction, cortication issues prevent their use yet.

4.7 Non-Destructive Testing

The Federal Aviation Administration's (FAA) regulations of airworthiness require the quality
assurance of each assembled part. Unlike metallic structures, composite parts are fabricated in-situ
and the grade of these structures are highly depending on the process robustness and workmen
skills. Typical manufacturing generated defects in composites can be voids, porosity, fiber
misalignment, wrinkling, poor cure, resin-rich and/or resin-poor areas, and the quality of the
adhesive layer itself is crucial for the structural performance of the structure. Therefore, every
composite part has to undergo rigorous non-destructive testing (NDT) prior to the assembly. For
metal aircraft structures, non-destructive testing is also part of the damage tolerance concept. As
 150

micro-cracks are basically tolerated, an airliner is regularly checked for structural integrity (no
occurrence of crack propagation). In so-called D checks, complete overhauls of six to ten years
rhythm, the paint is removed and cracks or delamination’s are sought. Apart from these regular
checks, the integrity is also tested after bird strikes, hard landings or similar incidents. All these
inspections are very costly due to the downtime of the aircraft. Therefore, the need for fast and robust
NDT methods is evident.

4.7.1 Ultra-sonic Methods

Due to the nature of composite structures, flaws can occur in monolithic structures (porosity,
delamination, cracks), the adhesive layers (deboning), or sandwich cores (density irregularities,
cracks). While flaws in the outer skin can be detected with single-sided access, the under-laying
defects often need through transition scanning. Thick structures are generally more difficult to test
than thin structures. The most common method for the inspection of aerospace structures is
ultrasonic testing (UT). There, a transducer is passed over the area being tested. Ultrasonic waves
penetrate the structure, while the receiver records the reflected (pulse-echo mode) or transmitted
(through transmission mode) sound waves, see Figure 4.7.1. The screen on the diagnostic machine
will show these results in the form of amplitude and pulse readings, as well as the time it takes for
the waves to return to the transducer. The so-called A-scan, the presentation of the amplitude of the
wave as a function of time, is sufficient for the manual detection of flaws such as voids, cracks,
delamination, etc. Scanning along a given route leads to the B-scan presentation with the in-depth
position of the aw as a function of the scan distance. The C-scan in turn represents an aerial defect
image of the scanned part by scanning a 2D-pattern, while the D-scan combines the in-depth
information of the B-scan with the C-scan.
The advantage of UT is the high sensitivity, permitting the detection of extremely small flaws. The
penetrating power of UT is higher than the one other methods, thus allowing the detection of flaws
deeper in the part. Unfortunately, the damping effects of local in-homogeneities of composite
structures reduce the reflected energy significantly. UT needs a high level of work skills and
experience; this is often mentioned as a costly drawback. Thick structures are difficult to inspect due
to high damping characteristics and high structural noise. While air-coupled ultrasonic is possible in
principle, this is restricted to through transmission mode. For pulse-echo mode, however,

Figure 4.7.1 Ultrasonic Test Setup in Through Transmission Mode (Courtesy of Kaufmann et al.)
 151

improvements are necessary and water is still common as the coupling medium between probe and
specimen.

4.8 Optimization Process

4.8.1 Multi-Objective Optimization
Our optimization problem contains the optimization of different goals, such as low-cost, low-weight
and testability. As we have to find the optimum of different disciplines (e.g. structural calculation
and economic aspects), one has to deal with a multi-objective, multidisciplinary design optimization.
Several ways to capture multi-objective design problems exist. Two or more objectives in our case
low-cost and low-weight should lead to an optimal geometry; therefore, they have to be incorporated
mathematically into one objective function. The standard formulation of a multi-objective design
problem is given as

Min F(x) = [f1 (x), f2 (x), … . . , f𝑛 (x)]T ,

subject to h(x) = 0 , g(x) ≤ 0 , a ≤ x ≤ b
Eq. 4.8.1
In our case, f1 and f2 represent the cost and the weight of a composite aircraft part. Stability and failure
criteria are represented by the inequality constraint g(x), whereas a and b are lower and upper limits
of the variable vector x (Kaufmann et al.)85.

4.8.2 Multilevel Programming

Sometimes, the objectives can be ordered hierarchically in terms of importance. Hence, a top level
objective function is defined and the set of points that minimize this first level objective is sought. In
a second step, the set is reduced to the points that minimize the second level objective, and the
method proceeds until the lowest level objective has been minimized. An example for multilevel
optimization of aircraft structures is given by [Gantois & Morris]86. Here, however, is concerned with
the trade-off behavior of a combined cost/weight optimization. It is not possible to order the two
objectives, as they are on the same level and
cannot be separated hierarchically. Thus, first
weight, then cost or first cost, then weight
approaches would not make sense here. In the
following, an overview on the different
techniques is given.

4.8.3 Pareto Optimality

A topic closely related to multi-objective
optimization is Pareto optimality. Imagine a full
search exploration of all possible designs of a
structural part, e.g. the points given in Figure
4.8.1. Each design solution is represented by a
variable set x, the manufacturing cost f1 and a
weight f2. As can be seen, the points that are a
Figure 4.8.1 Design solutions, the
combination of lowest cost and lowest weight are
corresponding Pareto points and the Pareto
marked with cross symbols; they are called frontier

85 M. Kaufmann, D. Zenkert and C. Mattei, “Cost optimization of composite aircraft structures including variable
laminate qualities”. Submitted to Composites Science and Technology in February 2008.
86 K. Gantois and A. J. Morris. “The multi-disciplinary design of a large-scale civil aircraft wing taking account of

manufacturing costs”. Structural Multidisciplinary Optimization, 28:31- 46, 2004.

 152

Pareto points. Mathematically, the Pareto points are defined by the definition:
A point x*∊ C is said to be (globally) Pareto optimal or a (globally) efficient solution or a non-dominated
or a non-inferior point for the multi-objective optimization problem (4.2) if and only if there is no x*∊ C
such that fi(x) ≤ fi(x*) for all i ∊ 1, 2, , , , n, with at least one strict inequality.
In other words, a Pareto point is a point in the design space for which there is (a) no possible design
solution with a lower weight and the same manufacturing cost or (b) no possible design with the
same weight and a lower manufacturing cost. The curve that connects all Pareto points is called
Pareto frontier. The Pareto frontier is now of great importance in the multi-objective optimization,
as it represents the trade-off behavior of the two objectives. A lot of optimization algorithms deal
with the generation of a complete Pareto frontier, thus providing a choice of possible solutions to the
optimization problem.

4.8.4 Minimizing Weighted Sums

More promising is an approach called weighted sums. Here, the two or more objectives are
incorporated into one objective function and weighted by predefined parameters α i. These
parameters represent the trade-off between the objectives and result in a one-shot design solution.
By varying αi, however, a range of solutions with different cost/weight trade-offs can be obtained.
The objective function is then given as

F(x) = ∑ αi fi (x), αi > 0 , i = 1,2, … . , n

i=1
Eq. 4.8.2
The approach of minimizing weighted sums is criticized by some researchers, as the generation of
evenly distributed Pareto points fails. However, in the case of a combined cost/weight optimization,
one would use the parameters to establish a relationship between the manufacturing cost and the
structural weight, and the generation of a Pareto frontier is not necessary. This approach is described
in detail below, where this cost/weight relationship represents the "fuel burn cost per structural
mass".

4.8.5 Weight and Performance Optimization

A lot of work has been done within the field of weight optimization (Kang, Kim and Walker)87-88. As
a second example, researched on the topology optimization of stiffened panels with different stiffener
configurations, i.e. zero, one or two longitudinal or two perpendicular stiffeners. The aim was further
to maximize the buckling load for a given plate thickness by varying the ply angle of an angle ply
symmetric layup.

4.8.6 Integrated Cost Optimization

When it comes to cost optimization of aircraft structures, research started relatively late, mainly due
to the lack of sophisticated cost models. In 1997, [Sobieszczanski-Sobieski and Haftka] wrote "Very
few instances can be found in which aerospace vehicle systems are optimized for their total
performance, including cost"89. Since then, progress has been made by different members of the

87 J.-H. Kang and C.-G. Kim. “Minimum-weight design of compressively loaded composite plates and stiffened panels

for post buckling strength by genetic algorithm”, Proceeding of 14th International Conference on Composite
Materials, 2003.
88 M. Walker. “The effect of stiffeners on the optimal ply orientation and buckling load of rectangular laminated

plates”. Computers and Structures, 80(27-30):2229 - 2239, 2002.

89 J. Sobieszczanski-Sobieski and R. T. Haftka. “Multidisciplinary aerospace design optimization: survey of recent

developments”. Structural Optimization, 14:1 - 23, 1997.

 153

scientific community. One of the earlier studies that integrated costing information into the design
process has been performed by [Geiger and Dilts]90. In 1996, they presented a conceptual model for
an automated design-to-cost approach; the main aim was the provision of a framework that helped
the structural engineer with the decision making process. [Heinmuller & Dilts] applied this
framework to the aerospace industry. They explained the design-to-cost concept as a trade-off
between operational capability, performance, schedule and cost. In 1997, they concluded that
"enabling automated design-to-
cost in a typical aerospace
manufacturing company will be
a difficult and time consuming
process".
Three material configurations
for a wing of a high-speed civil
transport aircraft have been
taken into account. It has been
concluded that lower operating
cost could be achieved by a
costly design with higher
reliability (less maintenance,
downtime, etc.), and , vice versa Figure 4.8.2 Trade-off Between Acquisition and Operating Costs
that the lowest acquisition cost
does not always signify the
lowest life-cycle cost (LCC). (Figure 4.8.2). There, the point depicted as Minimum Life Cycle Cost
would be the best alternative for both the manufacturer and the airline. [Marx et al.] included R&D
costs, manufacturing and sustaining costs, and revenue [Kaufmann et al.]91.

4.8.7 Combined Cost/Weight Optimization

As seen above, most research has been done within the field of weight optimization, or by minimizing
the manufacturing cost while maintaining a given structural performance (goal programming).
However, when both the reductions of cost and weight are sought, these two objectives have to be
incorporated into the formulation of one objective function. Different approaches to the multi-
objective optimization have been presented and it has been concluded that two most common
implementations were the generation of a Pareto frontier (i.e. by means of a genetic algorithm), and
the minimization of a weighted sum. The approach of a weighted sum has some advantages when it
comes to cost/weight optimization. First, it can easily be implemented in an optimization framework,
and second, the combination of cost and weight (weighted by parameters α i) gains an economical
significance. First, stiffened composite panels have been optimized separately for minimum cost Cmin
and minimum weight Wmin, and in a second step, the objective functions

C − Cmin W − Wmin
F(x) = α1 + α2
Cmin Wmin
Eq. 4.8.3
were applied. Here, C and W represented the actual cost and weight, respectively. The idea has been

90 T. S. Geiger and D. M. Dilts. “Automated design-to-cost: Integrating costing into the design decision”. CAD

Computer Aided Design, 28(6-7):423 - 438, 1996.

91 M. Kaufmann, D. Zenkert and C. Mattei, “Cost optimization of composite aircraft structures including variable

laminate qualities”. Submitted to Composites Science and Technology in February 2008.

 154

resumed by [Kelly and Wang]92 and [Wang et al.]93. They proposed a simplified objective function in
the form F(x) = cost + 500 x weight where 500 US$/kg represents the weight parameter α2. From an
engineering point of view, this can be interpreted such that a kilogram structural mass is worth 500
US$. This methodology was applied to the optimization of a closed box structure, an aileron and a
Krueger ap. Structural constraints were formed by the von Mises criterion, local and global buckling
coefficients, and the manufacturing cost consisted of material, fabrication and assembly costs. The
objective function is given as

F(x) = α1 . manufacturing cost + fuel burn(weight)

or F(x) = α1 . manufacturing cost + α2 . (weight)
Eq. 4.8.4
where α1 is 2 or 3.5 and the fuel burn has been replaced by α2 = 300 US$/kg times the weight. Here,
only metallic structures have been considered.

4.8.8 Interdisciplinary Wing Design – Structural Aspects

The following paper describes a multidisciplinary approach to design a wing with almost optimal
aerodynamic efficiency during the entire cruise flight [Anhalt et al.]94. Therefore a tight collaboration
between structural mechanics and aerodynamics is necessary. Aerodynamic aspects are described
here to illustrate the interdisciplinary nature of the design process, but they are not explained very
deeply. Here, we focuses on structural aspects, e.g. description of the tasks of the wings structural
members, their placement within the wing and the modeling of the actual wing structure95.

4.9 Flutter Phenomena96

4.9.1 Definition
Flutter is a complex phenomenon that must in general be completely eliminated by design or
prevented from occurring within the flight envelope. Flutter is a dynamic instability occurring in
flight, at a speed called the flutter speed, where the elasticity of the structure plays an essential part
in the instability. Flutter is a resonance effect where, at certain speeds and conditions, oscillations of
parts of the plane are amplified by the passing air, and grow in amplitude. This can lead to loss of
control and damage to the plane. Severe flutter can rip the wings or stabilizers off a plane. Flutter is
not something that simply happens at one particular speed. The onset and amplification of flutter
depend upon a variety of factors, including the loading of the plane (in particular the fuel loading in
the wings), and the turbulence of the atmosphere. Even then the flutter requires some initial
oscillation (excitation) of the affected structure in order for it to start.

4.9.2 History
The history of flutter testing lists the problems of getting sufficient excitation to induce flutter.
Solutions to this problem included flying around looking for areas of high turbulence, attaching small
rockets to the wings, or having rotating weights inside the wings. Structural excitation is a necessary

92 D. Kelly, K. Wang, and S. Dutton. “A guided trade-off for cost and weight for generating optimal conceptual

designs”. AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, 2003.

93 K. Wang, D. Kelly, and S. Dutton. “Multi-objective optimization of composite aerospace structures”. Composite

Structures”, 57(1):141 - 148, 2002.

94 C. Anhalt, H. P. Monner, E. Breitbach, “Interdisciplinary Wing Design – Structural Aspects”, SAE International,

2003.
95 M. Kaufmann, D. Zenkert and C. Mattei, “Cost optimization of composite aircraft structures including variable

laminate qualities”. Submitted to Composites Science and Technology in February 2008.

96 Krishna Kumar Subramanian. Aircraft Engineer, Aircraft Systems Educator, Qura.
 155

part of the flight flutter testing methodology. The first formal flutter test was carried out by Von
Schlippe in 1935 in Germany. His approach was to vibrate the aircraft at resonant frequencies at
progressively higher speeds and plot amplitude as a function of airspeed. A rise in amplitude would
suggest reduced damping with flutter occurring at the asymptote of theoretically infinite amplitude.
This idea was applied successfully to several German aircraft until a Junkers JU90 fluttered and
crashed during flight tests in 1938. Regulations dealing specifically with flutter, deformation, and
vibration on transport category airplanes were first introduced when part 04 of the Civil Air
Regulations (CAR) became effective in the mid-1940s. The flutter requirement was extensively
revised in 1964 to require compliance with the single failure criteria for the entire airplane as well
as adding special provisions for turboprop airplanes. Regulations dealing specifically with flutter,
deformation, and vibration on transport category airplanes were first introduced when part 04 of the
Civil Air Regulations (CAR) became effective in the mid-1940s.

4.9.3 Detection
Detection of impending aeroelastic instabilities cannot be made without adequate excitation.
Adequate excitation provides energy to excite all of the selected vibration modes with sufficient
magnitudes to accurately assess stability from the response data. Today’s aircraft designs undergo
sophisticated aeroelastic analyses to ensure that the design is free of flutter within the flight
envelope. These analytical results are often verified by wind-tunnel flutter models and ground
vibration tests. Flight flutter testing provides the final verification of the analytical predictions
throughout the flight envelope. In the early years of aviation, no formal flutter testing of full-scale
aircraft was carried out. The aircraft was simply flown to its maximum speed to demonstrate the
aeroelastic stability of the vehicle. Flutter and other aeroelastic instability phenomena have had a
significant influence on airplane development and the airworthiness criteria governing the design of
civil airplanes.

4.9.4 Aeroelasticity
The term aeroelasticity has been applied by aeronautical engineers to an important class of problems
in airplane design. It is often defined as a science which studies the mutual interaction between
aerodynamic forces and elastic forces, and the influence of this interaction on airplane design.
Aeroelastic problems would not exist if airplane structures were perfectly rigid. Modern airplane
structures are very flexible, and this flexibility is fundamentally responsible for the various types of
aeroelastic phenomena. Structural flexibility itself may not be objection-able; however, aeroelastic
phenomena arise when structural deformations induce additional aerodynamic forces. These
additional aerodynamic forces may produce additional structural deformations which will induce
still greater aerodynamic forces. Such interactions may tend to become smaller and smaller until a
condition of stable equilibrium is reached, or they may tend to diverge and destroy the structure. The
term aeroelasticity, however, is not completely descriptive, since many important aeroelastic
phenomena involve inertial forces as well as aerodynamic and elastic forces.
We apply a definition in which the term aeroelasticity includes phenomena involving interactions
among inertial, aerodynamic, and elastic forces, and other phenomena involving interactions
between aerodynamic and elastic forces. The former will be referred to as dynamic and the latter as
static aeroelastic phenomena.

4.9.5 Flutter
Flutter has perhaps the most far-reaching effects of all aeroelastic phenomena on the design of high-
speed aircraft. Modern aircraft are subject to many kinds of flutter phenomena. The classical type of
flutter is associated with potential flow and usually, but not necessarily, involves the coupling of two
or more degrees of freedom. The nonclassical type of flutter, which has so far been difficult to analyze
on a purely theoretical basis, may involve separated flow, periodic breakaway and reattachment of
 156

the flow, stalling conditions, and various time-lag effects between the aerodynamic forces and the
motion. Preventive measures and cures usually involve either increased stiffness or decreased
coupling by adjustments in mass distribution, or a combination of both. The most important stiffness
parameter affected by flutter considerations is wing torsional stiffness. It is not uncommon for the
flutter condition to control the selection of wing skin thickness.
Of course, wing structural design is controlled by either a strength or a stiffness criterion. For
example, if the torsion carrying structure of a wing is designed by a stiffness requirement, the wing
would probably consist of a structure which carries its normal stresses in the wing skin with a
minimum of stringers and flanges. This type of wing structure would require several spanwise webs
in order to stabilize the heavily loaded cover skin. For a wing designed initially by strength
considerations to carry a given load factor, it is obvious that a higher torsional stiffness and hence a
higher flutter speed will result if the ratio of stiffener area to skin area is reduced to a minimum. In
addition, the use of higher strength alloys, which have no corresponding increase in modulus of
elasticity, tends to make flutter more critical for wings designed for strength only. Heavy mass items
in the wing are often located by considerations of optimum conditions for flutter prevention.
For example, a given mass distribution may require higher wing stiffness and hence higher wing
structural weight to prevent flutter than some other mass distribution. For this reason, analytical and
model studies are often made in the design stages in order to determine the optimum mass
distribution for flutter prevention.
Wing planform and aspect ratio also have significant effects on flutter characteristics. Decreases in
wing aspect ratio and increases in sweep tend to raise flutter speeds, whereas increases in aspect
ratio and decreases in sweep, including sweep forward, reduce flutter speeds. Flutter considerations
may affect control surface design in the determination of aerodynamic and mass balance, hinge
location, and the degree of irreversibility required in the actuating system. Modern airplane design
requires aeroelastic stability evaluations including flutter, divergence, control reversal and any
undue loss of stability and control as a result of structural deformation. The aeroelastic evaluation
must include whirl modes associated with any propeller or rotating device that contributes
significant dynamic forces. Compliance with this section must be shown by analyses, wind tunnel
tests, ground vibration tests, flight tests, or other means found necessary by the regulator.