Aerodynamic Design Optimization for Natural Laminar Flow Airfoils

P. Paredes1 P. Mysore2 K. Jacobson3 B. Diskin4 N. Hildebrand4 M. Choudhari4

1 National Institute of Aerospace

2 Georgia Institute of Technology
3 Aeroelasticity Branch, NASA Langley Research Center
4 Computational AeroSciences Branch, NASA Langley Research Center

AIAA Aviation 2024, AIAA Paper 2024-3529

CFD-01: Boundary Layer Transition Modeling and Applications I
9:30AM-9:50AM PDT, Room: Academy Ballroom 410. Las Vegas, NV. July 29, 2024
 Motivation

Outline

1 Motivation

2 Theory

3 Results

4 Summary and Concluding Remarks

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 1 / 14

 Motivation

Natural Laminar Flow Technology

Natural laminar flow (NLF) technology is a passive laminar flow control (LFC) strategy that seeks to delay the
onset of boundary-layer transition (BLT) through shape optimization to reduce the viscous drag of the aerodynamic
vehicle
Viscous drag accounts for about 50% of the total drag of an aircraft1
Laminar flow on wings at high Re can provide upto O(10%) drag reduction on commercial aircrafts
Drag reduction improves fuel efficiency that increases flight range and payload capacity, reduces operational costs,
and lowers carbon emissions
Boeing has recently included NLF technologies2,3 :
I 787-8 NLF Nacelles
I 737Max NLF Winglet and Nacelles
I 777X NLF Nacelles

(Crouch 2015, AIAA Paper 2015-2472)

1
R. Rashad and D.W. Zingg. Viscous drag reduction on transport aircraft. AIAA Paper 1991-0685. 1991.
2
J.D. Crouch. Broundary-layer transition prediction for laminar flow control. AIAA Paper 2015-2472. 2015.
3
M.R. Malik et al. “Application of Drag Reduction Techniques to Transport Aircraft”. In: Encyclopedia of Aerospace Engineering. John Wiley & Sons, Ltd, 2015.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 1 / 14
 Motivation

Natural Laminar Flow Design Optimization

Optimization is critical to maximize the benefits of NLF technology
Knowledge-based inverse-design optimization:
I Crossflow Attenuated NLF (CATNLF)4
? Successful design of transonic wind tunnel model, the CRM-NLF5

Genetic algorithm design optimization:

I Coupling of Euler solver with BL code and LST N-factor calculation6
Physics-based adjoint design optimization:
I Coupling of Euler solver with BL code and PSE N-factor calculation7
I Coupling of RANS solver with BL code and N-factor surrogate8,9,10
([Link]

4
R.L. Campbell and M.N. Lynde. Natural laminar flow design for wings with moderate sweep. AIAA Paper 2016-4326. 2016.
5
M.N. Lynde and R.L. Campbell. Computational design and analysis of a transonic natural laminar flow wing for a wind tunnel model. AIAA Paper 2017-3058. 2017.
6
D. Simanowitsch et al. Comparison of Gradient-Based and Genetic Algorithms for Laminar Airfoil Shape Optimization. AIAA Paper 2022-0008. 2022.
7
O. Amoignon et al. “Shape Optimization for Delay of Laminar-Turbulent Transition”. In: AIAA Journal 44.6 (2006), pp. 1009–1042.
8
J.-D. Lee and A. Jameson. Natural-Laminar-Flow Airfoil and Wing Design by Adjoint Method and Automatic Transition Prediction. AIAA Paper 2009-0897. 2009.
9
R. Rashad and D.W. Zingg. Aerodynamic shape optimization for natural laminar flow using a discrete-adjoint approach. AIAA Paper 2015-3061. 2015.
10
Y. Shi et al. “Natural Laminar-Flow Airfoil Optimization Design Using a Discrete Adjoint Approach”. In: AIAA Journal 58.11 (2020), pp. 4702–4722.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 2 / 14
 Motivation

Objectives

Transition Modeling: Integration of physics-based BLT modeling based on N-factor method into
FUN3D/SFE calculations (similar to previous work with FUN3D/FV and LASTRAC11 )
I Direct coupling of RANS transitional solutions with N-factor calculation (no BL code)
I Transition location based on N-factor envelope calculation based on linear stability analysis (LSA)

Design Optimization: Fully-coupled physics-based adjoint design optimization of BLT control

strategies to target improvement of aerodynamic performance of the vehicle
I Sensitivities to transition location incorporated through fully-coupled adjoint of CFD, mesh
deformation, and transition prediction solvers

11
N. Hildebrand et al. Coupling of the FUN3D Unstructured Flow Solver and the LASTRAC Stability Code to Model Transition. AIAA Paper 2022-1952. 2022.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 3 / 14
 Theory

Outline

1 Motivation

2 Theory

3 Results

4 Summary and Concluding Remarks

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 4 / 14

 Theory

NLF Design Optimization with Adjoint-Based Sensitivities

Final Design Adjoint-Based
Sensitivities
Baseline Geometry Model
Optimizer
Adjoint Grid
Surface Grid Update of Design Deformation
Generation & Parameters
Parametrization Adjoint Flow Solver
Surface Grid Adjoint CFD Adjoint
Deformation Solver with Conversion
Laminar-Turbulent Flow Solver Imposed from LSA to
Volume Grid CFD Solver Conversion Transition CFD Grid
Generation/Deformation with Imposed from CFD to
Adjoint LSA
CFD Grid Transition LSA Grid
Transition
LSA Grid LSA Transition Prediction
Prediction
CFD: Computational Fluid Dynamics
LSA: Linear Stability Analysis (N-factor method)

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 4 / 14

 Theory

Coupled Adjoint Formulation

Adjoint problem formulated by defining the Lagrangian L based on the objective function J to be
optimized
Lagrangian contains J and set of residuals from the forward coupling
Lagrangian multipliers λ are the adjoint variables

shape deformation grid deformation LSA grid

z }| { z }| { z }| {
L(d, xs , x, ξ, ξT , γ, Q, q̄) = J(x, Q, d) + λTD D(d, xs ) +λTG G(xs , x) +λTξ Kξ (xs , ξ)
+λTγ Γ(ξ, ξT , γ) +λTR R(d, x, γ, Q) +λTq Kq (x, ξ, Q, q̄) +λTT T(ξ, q̄, ξT )
| {z } | {z } | {z } | {z }
intermittency CFD flow solution LSA base flow transition location

Adjoint equations derived by differentiating L wrt design variables, d, and collecting and setting to zero
the coefficients of the derivatives wrt state variables, xs , x, ξ, ξT , γ, Q, q̄
Adjoint-based sensitivities:
dL ∂J ∂D ∂R
= + λTD + λTR
dd ∂d ∂d ∂d
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 5 / 14
 Theory

Surface and Volume Grid Deformation

Surface grid deformation: Free-form deformation (FFD) method12 used for surface parametrization and
deformation
RAE 2822 airfoil with FFD box of 9x3 control points
0.2

z/c
0.0

−0.2
0.0 0.2 0.4 0.6 0.8 1.0
x/c
13
Volume grid deformation: Linear elasticity model used to displace volume nodes as a function of
surface deformation
RAE 2822 airfoil Deformed airfoil

12
T.W. Sederberg and S.R. Parry. “Free-Form Deformation of Solid Geometric Models”. In: SIGGRAPH Computer Graphics 20 (1986), pp. 151–160.
13
E.J. Nielsen and W.K. Anderson. “Recent improvements in aerodynamic design optimization on unstructured meshes”. In: AIAA Journal 40.6 (2002), pp. 1155–1163.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 6 / 14
 Theory

Transition Prediction Coupled with FUN3D

CFD solver: FUN3D/SFE solver14 based on second-order finite-element discretization with negative
version of Spalart-Allmaras RANS model
Transition prediction: Dual N-factor based on LST of planar TS and stationary CF instabilities
Laminar-turbulence model: Intermittency function γ is a multiplier of the turbulence production term15,16
( h
i
3.36(ξ−ξT ) 2 3/4
1 − exp −0.413 lT (ξT )
, RelT = 5.2Reξ if ξ ≥ ξT
γ(ξ, ξT ) = T

0 if ξ < ξT ,

Iterative method to model transitional flow

1 Calculate mean flow solution with defined γ(ξT ,i )

2 Convert near wall laminar-flow solution from CFD to LSA grid

3 Perform LSA and define a new γ(ξT ,i+1 )

4 Check convergence, i.e., |ξT ,i+1 − ξT ,i | < T , and go to step 1 if not satisfied

14
W.K. Anderson, J.C. Newman, and S.L. Karman. Stabilized finite elements in FUN3D. AIAA Paper 2017-0077. 2017.
15
S. Dhawan and R. Narasimha. “Some Properties of Boundary Layer Flow During the Transition from Laminar to Turbulent Motion”. In: Journal of Fluid Mechanics 3.4 (1958),
pp. 418–436.
16
G.J. Walker. “Transition Flow on Axial Turbomachine Blading”. In: AIAA Journal 27.5 (1989), pp. 595–602.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 7 / 14
 Theory

The Dual N-factor Criterion20

The logarithmic amplification ratio or N-factor is defined as
Z ξ
N(ξ, ω, β) = − σ(ξ 0 , ω, β) dξ 0
ξI

σ(ξ 0 , ω, β): Disturbance growth rate calculated with spatial LST eigenvalue analysis
Transition onset predicted where the dual N-factor indicator function reaches 1, IN = Nd (NTS , NCF ) = 1
 aTS  aCF
NTS NCF
IN = Nd (NTS , NCF ) = +
NTS,c NCF ,c

NCF : N-factor envelope of stationary CF waves (ω = 0)

NTS : N-factor envelope of TS waves approximated by the envelope of planar waves (β = 0) in a rotated
orthogonal coordinate system aligned with u∞
Demonstrated and validated for a subsonic aircraft model17 and the transonic CRM-NLF model18
Discrete adjoint implementation of spatial LST eigenvalue analysis similar to previous work19
17
A. Leidy et al. Measurements and Computations of Natural Transition on the NASA Juncture-Flow Model with a Symmetric Wing. AIAA Paper 2023-0441. 2023.
18
P. Paredes et al. Transition Modeling based on the Dual N-factor Method for the CRM-NLF Wind Tunnel Configuration. AIAA Paper 2023-3532. 2023.
19
C. Klauss et al. Eigenvalue sensitivity computations for linear stability theory. AIAA Paper 2023-3271. 2023.
20
D. Arnal. “Boundary layer transition: Prediction based on linear theory”. In: Special course on Progress in Transition Modeling. AGARD. R-793. 1994, pp. 1–62.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 8 / 14
 Results

Outline

1 Motivation

2 Theory

3 Results

4 Summary and Concluding Remarks

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 9 / 14

 Results

Flow Conditions and Optimization Parameters

RAE 2822 airfoil used to evaluate coupled adjoint formulation for design optimization
I Unswept, 2D BL case: Λ = 0◦ , CL,d = 0.3, symmetric spanwise BCs
I Swept, 3D BL case: Λ = 30◦ , CL,d = 0.225, zero-order extrapolation spanwise BCs
Optimizer: SNOPT, gradient-based optimizer with sequential quadratic programming method
Objective function: J = CD
Constraints: CL = CL,d , A ≤ A0 (area), tcp > 0.5tcp,0 (thickness)
Design variables: AoA, vertical displacement of 14 FFD control points (black circles)
NASA FUN3D/SFE solver: NS solver for unstructured grids with finite-element discretization
CGT21 : c-type structured grid with 1, 521 × 2 × 151 points (721 in airfoil surface, 75 in BL at midchord)
RAE 2822 airfoil with FFD box Near-wall grid view
0.2
z/c

0.0

−0.2
0.0 0.2 0.4 0.6 0.8 1.0
x/c
21
W. Chan et al. Best practices in overset grid generation. AIAA Paper 2002-3191. 2002.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 9 / 14
 Results

Computational Details and Verification of Adjoint-Based Sensitivities

Dual N-factor parameters for transition modeling:
I NTS,c = 9, NCF ,c = 7, aTS = 1, aCF = 2.5
Convergence thresholds:
I RANS residual: R = 10−13
I Linear grid deformation: G = 10−13
I LST eigenvalue problem with Newton-Raphson iterations: LST = 10−9
I Transition location: T = 10−8
Adjoint implementation verified by comparing adjoint-based and real-step central FD sensitivities
Adjoint Central FD (h = 10−6 )
d(CD )/d(AoA) 1.208 × 10−3 1.207 × 10−3
d(CL )/d(AoA) 7.856 × 10−2 7.862 × 10−2
d(CD )/d(FFDz ) 1.875 × 10−1 1.887 × 10−1
d(CL )/d(FFDz ) −1.391 −1.402

Relative errors are O(10−3 ) with selected FD step size and convergence thresholds
Similar relative errors found for a linearized frequency-domain gust analysis with FUN3D/SFE22

22
K.E. Jacobson, A.S. Thelen, and B.K. Stanford. Linearized Frequency-Domain Gust Analysis and Adjoint-Based Sensitivities. AIAA Paper 2024-2591. 2024.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 10 / 14
 Results

Computational Details and Verification of Adjoint-Based Sensitivities

Dual N-factor parameters for transition modeling:
I NTS,c = 9, NCF ,c = 7, aTS = 1, aCF = 2.5
Convergence thresholds:
I RANS residual: R = 10−13
I Linear grid deformation: G = 10−13
I LST eigenvalue problem with Newton-Raphson iterations: LST = 10−9
I Transition location: T = 10−8
Adjoint implementation verified by comparing adjoint-based and real-step central FD sensitivities
Adjoint Central FD (h = 10−6 )
d(CD )/d(AoA) 1.208 × 10−3 1.207 × 10−3
d(CL )/d(AoA) 7.856 × 10−2 7.862 × 10−2
d(CD )/d(FFDz ) 1.875 × 10−1 1.887 × 10−1
d(CL )/d(FFDz ) −1.391 −1.402

Relative errors are O(10−3 ) with selected FD step size and convergence thresholds
Similar relative errors found for a linearized frequency-domain gust analysis with FUN3D/SFE22
For a fixed ξT , sensitivities become d(CD )/d(AoA) = 9.840 × 10−5 and d(CD )/d(FFDz ) = −2.760 × 10−4
I Highlights importance of accounting for the effects of ξT variations
22
K.E. Jacobson, A.S. Thelen, and B.K. Stanford. Linearized Frequency-Domain Gust Analysis and Adjoint-Based Sensitivities. AIAA Paper 2024-2591. 2024.
Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 10 / 14
 Results

Unswept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 0◦ , AoA0 = 0.72◦ , CL,d = 0.3
2D boundary layer: NCF = 0, Nd = NTS /9
CD /CD,0 , CL /CL,d , xT /xT ,0 Shape and Cp of #0 & #32 N-factors of iteration #32
1.4
1.8 0.1 Nd,u
CD /CD,0 N /N

z/c
CL /CL,d 0.0 1.2 N T S,u/N T S,c
CF,u CF,c
1.6 xT,l /xT,l,0
xT,u /xT,u,0 −0.1
0.0 0.2 0.4 0.6 0.8 1.0 1
−1.0
1.4 x/c
0.8
−0.5

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5 Design #0

0.2
Design #32
0.6 1.0 0
0 5 10 15 20 25 30 35 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Design # x/c x/c

Optimized airfoil #32: AoA32 = 1.057◦ , CD,32 /CD,0 = 0.7 (30% drag reduction)
Additional design iterations required to reach optimum shape for NLF

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 11 / 14

 Results

Unswept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 0◦ , AoA0 = 0.72◦ , CL,d = 0.3
2D boundary layer: NCF = 0, Nd = NTS /9
CD /CD,0 , CL /CL,d , xT /xT ,0 Shape and Cp of #0 & #32 N-factors of iteration #32
1.4
1.8 0.1 Nd,l
CD /CD,0 NT S,l /NT S,c

z/c
CL /CL,d 0.0 1.2 NCF,l /NT S,c
1.6 xT,l /xT,l,0
xT,u /xT,u,0 −0.1
0.0 0.2 0.4 0.6 0.8 1.0 1
−1.0
1.4 x/c
0.8
−0.5

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5 Design #0

0.2
Design #32
0.6 1.0 0
0 5 10 15 20 25 30 35 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Design # x/c x/c

Optimized airfoil #32: AoA32 = 1.057◦ , CD,32 /CD,0 = 0.7 (30% drag reduction)
Additional design iterations required to reach optimum shape for NLF

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 11 / 14

 Results

Unswept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 0◦ , AoA0 = 0.72◦ , CL,d = 0.3
2D boundary layer: NCF = 0, Nd = NTS /9
CD /CD,0 , CL /CL,d , xT /xT ,0 Shape and Cp of #0 & #32 N-factors of iteration #32
1.4
1.8 0.1 Nd,u
CD /CD,0 N /N

z/c
CL /CL,d 0.0 1.2 N T S,u/N T S,c
CF,u CF,c
1.6 xT,l /xT,l,0
−0.1 Nd,l
xT,u /xT,u,0 1
−1.0 0.0 0.2 0.4 0.6 0.8 1.0 NT S,l /NT S,c
1.4 x/c NCF,l /NT S,c
0.8
−0.5

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5 Design #0

0.2
Design #32
0.6 1.0 0
0 5 10 15 20 25 30 35 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Design # x/c x/c

Optimized airfoil #32: AoA32 = 1.057◦ , CD,32 /CD,0 = 0.7 (30% drag reduction)
Additional design iterations required to reach optimum shape for NLF

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 11 / 14

 Results

Swept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 30◦ , AoA0 = 0.72◦ , CL,d = 0.225
3D boundary layer: Nd = (NTS /9) + (NCF /7)2.5 (NCF > 0 only in lower side)
CD /CD,0 , CD /CD,d , xT /xT ,0 Shape and Cp of #0 & #4 N-factors of iteration #4
1.4
1.8 0.1 Nd,u
CD /CD,0

z/c
CL /CL,d 0.0 1.2 NT S,u /NT S,c
1.6 xT,l /xT,l,0 NCF,u /NCF,c
xT,u /xT,u,0 −0.1
0.0 0.2 0.4 0.6 0.8 1.0 1
−1.0
1.4 x/c
Design #0 0.8
−0.5 Design #4

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5
0.2

0.6 1.0 0
0 1 2 3 4 5 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6
Design # x/c x/c

Optimized airfoil (iteration #4): AoA4 = 0.658◦ , CD,4 /CD,0 = 0.855 (14.5% drag reduction)
Additional design iterations expected to further reduce drag

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 12 / 14

 Results

Swept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 30◦ , AoA0 = 0.72◦ , CL,d = 0.225
3D boundary layer: Nd = (NTS /9) + (NCF /7)2.5 (NCF > 0 only in lower side)
CD /CD,0 , CD /CD,d , xT /xT ,0 Shape and Cp of #0 & #4 N-factors of iteration #4
1.4
1.8 0.1 Nd,l
CD /CD,0

z/c
0.0 1.2 NT S,l /NT S,c
CL /CL,d
1.6 xT,l /xT,l,0 NCF,l /NCF,c
xT,u /xT,u,0 −0.1
0.0 0.2 0.4 0.6 0.8 1.0 1
−1.0
1.4 x/c
Design #0 0.8
−0.5 Design #4

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5
0.2

0.6 1.0 0
0 1 2 3 4 5 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6
Design # x/c x/c

Optimized airfoil (iteration #4): AoA4 = 0.658◦ , CD,4 /CD,0 = 0.855 (14.5% drag reduction)
Additional design iterations expected to further reduce drag

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 12 / 14

 Results

Swept RAE 2822 Airfoil

∗
Flow conditions: M∞ = 0.19, Rec = 5.6 × 106 , T∞ = 300 K, Λ = 30◦ , AoA0 = 0.72◦ , CL,d = 0.225
3D boundary layer: Nd = (NTS /9) + (NCF /7)2.5 (NCF > 0 only in lower side)
CD /CD,0 , CD /CD,d , xT /xT ,0 Shape and Cp of #0 & #4 N-factors of iteration #4
1.4
1.8 0.1 Nd,u
CD /CD,0

z/c
CL /CL,d 0.0 1.2 NT S,u /NT S,c
1.6 xT,l /xT,l,0 NCF,u /NCF,c
xT,u /xT,u,0 −0.1 Nd,l
0.0 0.2 0.4 0.6 0.8 1.0 1
−1.0 NT S,l /NT S,c
1.4 x/c
Design #0 0.8
NCF,l /NT S,c

−0.5 Design #4

N
1.2
0.6

Cp
1.0 0.0
0.4

0.8 0.5
0.2

0.6 1.0 0
0 1 2 3 4 5 0.0 0.2 0.4 0.6 0.8 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6
Design # x/c x/c

Optimized airfoil (iteration #4): AoA4 = 0.658◦ , CD,4 /CD,0 = 0.855 (14.5% drag reduction)
Additional design iterations expected to further reduce drag

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 12 / 14

 Summary and Concluding Remarks

Outline

1 Motivation

2 Theory

3 Results

4 Summary and Concluding Remarks

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 13 / 14

 Summary and Concluding Remarks

Summary and Concluding Remarks

Implementation of physics-based aerodynamic design optimization of unswept and swept airfoils for
natural laminar flow
I Adjoint-based design optimization formulation with NASA FUN3D/SFE, linear grid deformation, and
physics-based transition modeling based on linear stability analysis
I Iterative transition modeling method with NASA FUN3D/SFE based on dual N-factor criterion with LST
spatial eigenvalue problem that accounts for linear amplification of planar TS and stationary CF waves
Preliminary optimization results with unswept and swept airfoils at subsonic conditions achieve significant
drag reductions by shifting transition locations downstream over both airfoil sides
Future work:
I Perform additional design iterations to reach optimal airfoil shapes
I Extend the analysis to transonic and supersonic conditions
I Apply the optimization to 3D NLF wing design

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 13 / 14

Thank you for your attention
Acknowledgements
I This material is based upon work supported by
? the NASA Transformational Tools and Technologies (TTT) Project of the Transformative

Aeronautics Concepts Program (TACP) under the Aeronautics Research Mission Directorate
(ARMD)
? the Air Force Office of Scientific Research under award number FA9550-20-0023

I Computational Resources supporting this work are provided by

? Department of Defense (DOD) High Performance Computing Modernization Program (HPCMP)
? NASA High-End Computing (HEC) Program through the NASA Advanced Supercomputing (NAS)
Division at Ames Research Center
? NASA K cluster at the Langley Research Center

Aerodynamic Design Optimization for Natural Laminar Flow Airfoils 14 / 14