Document
Aerodynamic Design of Integrated Propulsion - Airframe
Configuration of the Hybrid Wing - Body Aircraft
† * ** †
May - Fun Liou , Hyoungjin Kim , BJ Lee and Meng - Sing Liou
†
NASA Glenn Research Center
* ** SAIC & Vantage Partners, LLC,
Cleveland, OH
AIAA Aviation 2017
June 5 - 9, 2017, Denver, Co.
Outline
• Background & Objectives
• Aerodynamics of Hybrid Wingbody - Propulsion System
• Technical Backgrounds
– Geometric Parameterization
– Mesh generation & deformation
• Optimization: Aero Performance & Constraints
• Analysis of Optimal Design
• Viscous Effects on Aerodynamic Performance
• Conclusion
Background – Far Term (beyond 2035)
• HWB (hybrid wingbody) configuration requirements
Distributed Electric Propulsion System
• Turboelectric Distributed Propulsion ( TeDP )
– Mail - slot nacelle near trailing edge
– Boundary Layer Ingestion into embedded propulsor fans
– Propulsor fans driven by superconducting electric motors
– Wingtip mounted superconducting turbo - generators
Propulsor and inlet - nozzle systems Felder, J., Kim, H. D., Brown, G. V., and Chu, J., “An Examination of the Effects of Boundary Layer Ingestion on Turboelectric Distributed Propulsion Systems,” AIAA – 2011 – 0300
Development of Technologies for Hybrid
Wing/body with Distributed Electric Propulsion
- 2013 2014 2015 2016 2017 * N3 - X conceptual design N3X - Dep300 clean N3 - X with mailslot N3X - Dep300 with PAI Configurations N+2B inlet shape wing, 300 passenger nacelle nacelle (PAI) ** optimization cabin inlet A – BLI wall shaping mailslot n acelle cowl Inlet mailslot mailslot wall shaping cross wind analysis surface design GE R4 scaled single stage fan, Propulsor sizing /conceptual electri c fan design conceptual study of counter rotating fan unstructured iso – spring analogy unstructured aniso mesh Mesh crosschecked with overflow overset Parameterization NURBS CST / planform/inlet/nacelle NURBS Roe/AUSM+UP N3X - Analysis with body - force drag decomposition CFD Modeling SA/2 - eqs. turbulence models model trim modeling LUSGS & GMRES Optimization GBOM based on a djoint approach adjoint/NSGA - II Method *Jim Felder et al. AIAA – 2011 – 0300 Completed On - going & future works Current **Craig L. Nickol AIAA - 2012 - 0337
CFD flow - field of N3 - X with Fan Propulsor
st Flow field in the 1 slot with fan bodyforce model (2014) Total pressure contour FPR = 1.325 Static pressure contour Mass flow rate = 150.8kg/s/slot Mailslot top view Static pressure contour FLA ALF N3 - X powered by GE R4 distributed fan Kim et al. ISABE - 2015 - 20228 N3 - X mailslot modeling
Objective of the Present Work
• Further refine parameterization strategy for general complex
integrated propulsion - airframe system.
• Aerodynamic design under static stability constraints.
• Analysis and understanding of simulated flow - field
of the optimized configuration
Parameterization of Wing and Nacelle
Design sections of surface design and twist angles PAI configuration for present work
(we added 5 more sections for twist angle on red)
Free Stream Nacelle inlet Nacelle exit
Design with semi - span
• Horizontal region: 3D CST Main wing parameterization (CST – 4 sections) • Vertical region: 2D CST • Corner: elliptical shape connecting the two regions • Nacelle 6x8 • Main wing 60 (7x8+4 offset) • Twist 8 variables • Total : 124 design parameters Note: additionally, tip Section surface parameterization (CST) twist angle is used for • 8 parameters for each of upper & lower surfaces trim constraint and smooth spanwise • Minimization of L2 norm interpolation between • CST basis function (RHS) design sections, • Kulfan , B.,“Universal Parametric Geometry thickness constraint for Representation Method,” JA vol.45, No.1, 2008 cabin space are applied.
Parameterization of Airframe and Inlet
Example of aerodynamic shape optimization of nacelle
Inlet parameterization
Passage 1
Planform parameterization
Passage 4 N3 - X cowl shape design results: Comparison of sectional local Mach contours, Left: initial, Right: design.
(Kim et al. AIAA 2015 - 3805) Note: Theses inlet/nozzle and planform parameters are not used in the present work, it is used for previous design for the current baseline model and will be refined in the future study.
Mesh Generation & Deformation
• Mapping unstructured surface meshes on structured p3d (output of PAI configuration
generator)
• Spring analogy from surface mesh deformation to volume mesh deformation
Baseline surface mesh Baseline surface geometry (P3D) Mesh for RANS analysis Mesh for inviscid flow analysis
Longitudinal trim & static stability
𝐓𝐫𝐢𝐦 ∶ 𝐹 = 0 ; 𝑀 = 0 𝑖 . 𝑒 . 𝐷𝑟𝑎𝑔 = 𝑇ℎ𝑟𝑢𝑠𝑡 & 𝑃𝑖𝑡𝑐ℎ𝑖𝑛𝑔 𝑚𝑜𝑚𝑒𝑛𝑡 𝑎𝑡 𝑐 . 𝑔 . 𝑖𝑠 𝑧𝑒𝑟𝑜 .
𝑥 𝑐𝑔
Federal Aviation Regulations (FAR), Section 161 of PAR 23 : The
airplane must maintain longitudinal trim under each of the following
conditions: (1) A climb, (2) Level flight at all speeds, (3) A descent, (4) Approach.
Static margin : Pitching moment arm - Distance between Xc.g . and the Xa.c .; 𝐶 𝑀 𝛼
Mathematical expression - 𝐾 = −
𝑛 𝐶 𝐿 𝛼
Static stability : pitching moment changes caused by the perturbation in AOA revert the
aircraft back into trim, i.e. 𝐶 < 0 𝐾 >0
𝑀 𝑛 𝛼
Optimization : Aero Performance & Constraints
𝐶
Minimize: 𝐷
Subject to: 𝐶 = 𝐶 , 𝐶 = 𝐶 = 0 , Specified SM (baseline 4%MAC)
𝐿 𝐿 𝑀 𝑀 𝑇 𝑇
𝑅 ≥ 𝑅
𝐿𝐸 , 𝑛𝑎𝑐𝑒𝑙𝑙𝑒 𝐿𝐸 , 𝑏𝑎𝑠𝑒𝑙𝑖𝑛𝑒 𝑛𝑎𝑐𝑒𝑙𝑙𝑒
𝑡 𝑐 ≥ 𝑡 𝑐
𝑚𝑎𝑥 𝑚𝑎𝑥 , 𝑏𝑎𝑠𝑒𝑙𝑖𝑛𝑒 for each design section
𝐶 = 𝐶 + 𝐶 ∆ 𝛼 + 𝐶 ∆ 𝜃
Minimize: 𝐷 𝐷 𝐷 𝐷 𝑤𝑡 0 𝛼 𝜃 − 1 𝐶 𝐶 𝐶 𝐶 ∆ 𝛼 ∆ 𝐶 ∆ 𝐶 ∆ 𝛼 𝐿 𝐿 𝐿 𝐿 𝐿 𝐿 𝛼 𝜃 𝛼 𝜃 = , = 𝐶 𝐶 𝐶 𝐶 ∆ 𝜃 ∆ 𝐶 ∆ 𝐶 ∆ 𝜃 𝑀 𝑀 𝑀 𝑀 𝑤𝑡 𝛼 𝑀 𝑀 𝛼 𝜃 𝑤𝑡 𝜃 Cabin (301 Passengers) layout for thickness constraint Craig L. Nickol AIAA - 2012 - 0337
Clean - wing Design
• Front loaded optimized wing Optimized Baseline • X moved from 38.21%c ( ○ ) CG to upstream ( 36.73%c ● ) • SM=9%MAC • Shock strength at TE is reduced
Clean - wing Design
• Baseline (26.3cnts) : (Induced drag): (wave drag) =87%:13% Optimized Baseline • 15% ( - 3.4 cnts ) induced drag reduction • 85% ( - 2.9 cnts ) wave drag reduction C C C +C Di Dw Di Dw Baseline 87.24% 12.76% 100.00% Optimized 74.36% 1.87% 76.23% delta - 12.88% - 10.89% - 23.77% - 6.3 counts
Propulsion Airframe Integration Design
• Baseline (43cnts) : Optimized Baseline (Induced drag): (wave drag) =93%:7% • SM=4%MAC • X almost not changed CG even though the center of pressure changed significantly at outboard.
• Nacelle and inboard area dominate the longitudinal stability.
Propulsion Airframe Integration Design
• Baseline (43cnts) : (Induced drag): (wave drag) =93%:7% Baseline Optimized • 19% ( - 7.5cnts) induced drag reduction • 75% ( - 2.1cnts) wave drag reduction C C C +C Di Dw Di Dw Baseline 93.47% 6.53% 100.00% Optimized 75.75% 1.64% 77.39% Delta - 17.72% - 4.89% - 22.61% - 9.59 counts
Clean - wing vs PAI
• Lift contribution of nacelle affects longitudinal stability at inboard area.
• PAI baseline - 12% more lift, 35~39% more drag (vs. Cleanwing baseline) • X is predicted further downstream around 43.7%c while clean wing has CG at 36.7%c.
CG • More induced drag dominant design.
Drag Comparison Cdi+Cdw 10 C l Cdi 5 0 Cdw Baseline Optimized Baseline Optimized Cleanwing PAI Cdw Cdi Cdi+Cdw
Viscous Effects on Aerodynamic Performance
• Inviscid analysis is used for fast design optimization.
• RANS analysis for optimized PAI configurations.
Optimized Baseline Optimized Baseline Euler Analysis – Mach Contour RANS analysis – Ps Contour Euler C C C C +C RANS C C C C C +C +C L Di Dw Di Dw L Di Dw Dv Di Dw Dv Baseline 0.1934 39.64 2.77 42.41 Baseline 0.1503 35.7 3.85 57.5 97.1 Optimized 0.1934 32.12 0.70 32.82 Optimized 0.1520 22.9 0.70 56.6 80.5 Delta 0.00 - 7.52 - 2.07 - 9.59 Delta +0.0017 - 12.8 - 2.89 - 0.92 - 16.6 Delta% 0% - 18.96% - 74.88% - 22.6% Delta% +1.3% - 35.86% - 74.96% - 1.59% - 17.10% Euler C - cowl Dw RANS C - cowl Dw Baseline 1.29 Baseline 1.96 Optimized 0.53 Optimized 0.78 Delta - 0.76 Delta - 1.19 Delta% - 58.72% Delta% - 60.49%
Viscous Effects on Aerodynamic Performance -
cont’d
• Span - wise Lift Distribution
RANS analysis Euler analysis 0.16 0.16 0.14 0.14 0.12 0.12 0.1 0.1 0.08 0.08 Baseline 0.06 Adjoint 0.06 Elliptic 0.04 0.04 0.02 0.02 0.00 0.20 0.40 0.60 0.80 1.00 0.00 0.20 0.40 0.60 0.80 1.00 2y/b 2y/b
Viscous Effects on Aerodynamic Performance -
cont’d
• Surface Pressure Distribution y=60%b/2
y=0%b/2
y=95.4%b/2
y=30%b/2
Conclusion
• A design analysis tool for efficient geometry generation and optimal shape
design of the hybrid wing body propulsion airframe integration (PAI) has
been developed
• Preliminary PAI configurations of HWB are designed with Euler analysis
for fast turn around and rigorously investigated with RANS analysis .
– The RANS analysis results carries the improvement of performance consistently as
Euler analysis predicted .
• Aerodynamic optimization with lift, pitching moment constraints was
conducted ; the first trim, longitudinal stability consideration for HWB
PAI configuration
– Almost 10 counts of drag reduction could be achieved .
– Design starting from PAI concept is required due to that nacelle installation has
significant impact on aerodynamics, trim and longitudinal stability .
Future Works
- 2013 2014 2015 2016 2017 2018 - * N3 - X conceptual design N3X - Dep300 clean PAI N3 - X with mailslot N3X - Dep300 with N3X - Dep300 with N+2B inlet shape wing, 300 passenger nacelle nacelle (PAI) nacelle and propulsor Configurations ** optimization cabin inlet A – BLI wall shaping mailslot n acelle cowl propulsion system Inlet mailslot mailslot wall shaping cross wind analysis surface design sizing with fan/nozzle GE R4 scaled single stage fan, sizing /conceptual electri c fan design BLI tolerant fan Propulsor conceptual study of counter rotating fan unstructured – unstructured iso – spring analogy unstructured aniso mesh airframe/inlet/nozzle Mesh crosschecked with overflow overset structured - propulsor fan blade Parameterization NURBS CST / planform/inlet/nacelle NURBS parameterization (CST ) through flow model – Roe/AUSM+UP N3X - Analysis with body - force drag decomposition axi - symmetric (CSTALL) SA/2 - eqs. turbulence models CFD Modeling model trim modeling multi - stage CFD LUSGS & GMRES (SWIFT) Optimization GBOM based on a djoint approach adjoint/NSGA - II adjoint/NSGA - II Method *Jim Felder et al. AIAA – 2011 – 0300 Completed On - going & future works Current **Craig L. Nickol AIAA - 2012 - 0337
Acknowledgement
This work was supported by NASA’s Advanced Air Transport
Technology (AATT) Project .
NSGA - II – 8 twist angles (PAI)
Sections for twist angle variables (we added 5 more sections for twist angle on red) 6.00E-03 5.50E-03 Design with semi - span # gen = 1 # gen = 2 # gen = 3 5.00E-03 # gen = 4 # gen = 5 # gen = 7 4.50E-03 # gen = 8 # gen = 11 # gen = 19 4.00E-03 # gen = 40 Baseline Optimized 3.50E-03 Induced Drag coeficient *Note: This optimized point is from adjoint with 124 parameters 3.00E-03 5.00E-05 1.00E-04 1.50E-04 2.00E-04 2.50E-04 3.00E-04 3.50E-04 4.00E-04 4.50E-04
Wave Drag Coefficient
NSGA - II – 8 twist angles
0.16 Baseline Adjoint 0.14 Elliptic Gen 40 Run 1 0.12 Manual Nose Down (Elliptic) 0.1 0.08 0.06 0.04 0.02 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Local Incidence Angle Comparison
Twist Angle] (Deg.)
- -2 -4 2y/b -6 Incidence[AOA 0% 95.4% -8 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2y/b
Baseline NSGA II - G40 R1 Elliptic-Manual Nose down
Scaled Sensitivity of Nacelle Parameters
ADJ dv = 141 6.00E-04 4.00E-04 2.00E-04 0.00E+00 0 20 40 60 80 100 120 140 160 -2.00E-04 -4.00E-04 -6.00E-04 twist angles Nacelle Parameters twist angles at additional sections • The sensitivities of nacelle parameters scaled by 5 times and twist angle by 1.25 for both optimized cases.
• The shock strength on the nacelle scaled sensitivity case got weaker than the prime optimized design but the geometry resulted marginally larger drag due to increase of induced drag.
C C C +C Di Dw Di Dw Baseline 93.47% 6.53% 100.00% Optimized - 17.72% - 4.89% - 22.61% Nacelle SCLD - 14.36% - 6.14% - 20.50% Prime Optimized Nacelle Scaled