Multidisciplinary Optimization for Industrial Aeronautical Applications. Dr. Michel Delanaye General Manager Contact:

Transcription

1 Multidisciplinary Optimization for Industrial Aeronautical Applications M. Delanaye, F. Lani, I. Lepot Dr. Michel Delanaye General Manager Contact:

2 Aim Aim & Outline Present some real applications of multi-objective multidisciplinary optimization for complex design problems in aeronautics Outline Quick presentation of Minamo platform Example 1: design of counter rotating ti open rotors (aeroacoustics) Example 2: design of composite structural parts (cost, manufacturing)

3 Who we are?

4 R&T private company focused on simulation and modelling process modeling integrity analysis On demand structure design Multidisciplinary analysis flow aerodynamic design

5 Minamo basics

6 Surrogate assisted Genetic Algorithm workflow User Specifications Gradients for complex objectives usually non available! Approximate Model ANN, RBF, Optimization GA, SA, DATABASE Accurate Model CFD / Structure / Exp. /... END Performance Check

7 Surrogate based optimization Obje ective Predicted Optimum Actual Optimum Approximate model Initial Accurate Results Design Variable

8 Surrogate based optimization Obje ective Predicted Optimum Actual Optimum Artificial Neural Networks Radial Basis Functions Kriging... Approximate model Initial Accurate Results Design Variable

9 Efficient design exploration techniques Reduce the number of samplings (high fidelity simulation) Accuracy of response surfaces (models) Methodologies to assess the quality and reliability of the models Efficient multi-objective GA Key issues

10 Derivative free optimization with Minamo Illustration of adaptive sampling capability

11 Efficient Design Exploration Misplaced optimum Candle function Global optimum Exact solution Standard LHS sampling Auto adaptive LCVT sampling

12 Monitoring/Reliability Leave-One-Out (Open gap/stall point) DOE Optimization Large DoE scatter - Stabilization after about 50 design iterations: 2 different promising design families pointed out, satisfying the manufacturing constraints LOO Reliability Assessment: Isentropic efficiency correlation coefficient (DoE) (optimization)

13 ANOVA Sobol indices First order sensitivities and interaction volume (if required higher order sensitivities) quantification Relative Importance of Parameters 03 0,3 Sobol Indices 0,25 0,2 0,15 0,1 Section 5 first camber parameter Isent_Eff_Large_Gap_1.10 Isent_Eff_Small_Gap_1.10 Isent_Eff_Large_Gap_1.13 Isent_Eff_Small_Gap_1.13 0,05 0 Illustration on NEWAC optimization accounting for engine wear

14 Direct CAD access CATIA CAPRI mesher Master script (Python or C++, called by Minamo) CA D Model Ref. mesh Mod dified CAD model Modified CA D Model Re ef. mesh Me esh CAPRI client GeomSim Discrete, MeshSim Linux/Windows TCP/IP SOAP IGG/AutoGrid 5/Samcef/Abaqus CAE, etc Windows/Linux CAPRI server CATIA V5 SolidWorks UG NX Pro/Engineer OpenCASCADE

15 Mono- and multi-objective evolutionary algorithms, including memetic approaches Online surrogates approach Space fill and auto-adaptive DoE (LHS, CVT, LCVT, ) Efficient adaptive nonlinear global and local models (ANN, RBFN, Kriging, SVM, ) Response surfaces reliability through leave-k-out cross-validation Constraints activity interactive monitoring Quantitative Variance Analysis tool (ANOVA): Sobol sensitivity indices estimation Data mining capabilities for efficient multi-criteria decision making (self organizing maps, ) Simulation coupling through Python scripting Native and neutral CAD access Available in standalone version or as engine plug-in in other software (Optimus) Summary of Minamo Features

16 Example 1: Design of Counter-Rotating Open Rotors

17 Open Rotor Advanced Concept Studies Potential to meet drastic fuel burn decrease for future passengers aircraft at horizon > 2020 Challenges: Noise mitigation while ensuring high efficiency Aeromechanical Optimization of a Counter-Rotating Open Rotor (CROR) 17 17

18 Objective function: Maximization of global TOC Minimization of acoustic TO Constraints: Aerodynamic constraints: Thrust TOC and TO Torque TOC and TO (DD architecture) Streamlines contraction/r1 tip vortex Mechanical constraints Max VM stress/ linear FE, rig-scale Simplified flutter criterion Geometric constraints/feasibilitiy: LE/TE thickness, max thickness, reverse mode, LE curvature change criterion, curvilinear length of TE, CG position RPM R1 = RPM R2 Optimization Specification Variable blade re-staggering (including between both OPs) Clipping of rotor 2 is fixed, no contouring modification Failed or unstable simulations handled by a success switch 18

19 In-house Cenaero blade modeler Profile shape (6 equidistant sections along blade height) Maximum thickness Maximum thickness position Chord length Stagger angle Skeleton angle at LE/TE Stacking (6 equidistant sections along blade height + 2 additional sections at 90 and 95%) CG axial position CG tangential position Blade pitch angles including variable delta restaggering between TOC and TO 102 parameters conception space 19

20 CROR Optimization Chain Setup Maximize Minimize Aero./Mech. constraints Design of Experiments Approximate Models Auto-adaptive RBF networks Minamo END Evolutionary MO Optimization i elsa RANS FE SAMCEF Post treatment utilities Performance check Success switch DATABASE ONLINE modeling Infill criteria

21 CROR CFD Setup elsa (ONERA) RANS simulations with mixing plane Absolute velocity formulation Non reflecting farfield fi BCs k-ω Wilcox turbulence model Jameson's scheme Full Multigrid (2 levels) Mesh convergence analysis (~ nodes) Cost functions convergence Farfield boundaries positioning Regeneration robustness assesssed through dedicated DoE

22 Lessons Learned Phase 2 Acoustic cost function behaviour Acoustic A ti criterion it i may be b seen as minimization of R1 wake, up to the height of R2 REFERENCE OPTIMIZED ŵ = projection of relative speed fluctuation on the normal to the mean flow Wake of R1 reduced up to the height of R2 R1 tip vortex strengthened strengthened, but above R2 tip

23 Pareto front (MM+CFD): acoustic cost vs global Phase 3 V 1.2 V 2.0 Phase 2

24 Pareto front: Acoustic cost vs global V 1.2 evel Acoustic l 102 dimensions ~ 500 function calls Aeromechanical Optimization Phase 3 V 2.0 Phase 2 Mechanical constraints most stringent Gain: >10 db Effi: > 1.5 % Aco oustic cos st functio TO Global efficiency Global TOC

25 Promising case (selected V2.2) Rotor 1 SS Pressure Distribution OPTIMIZED REFERENCE Non dimensional static pressure

26 Example 2: Design of composite structures

27 Structural preliminary design and optimization of composite structures 1. Objectives Multi-disciplinary / multi-physics technical objectives: Maximize the stiffness Minimize the mass Maximize the buckling resistance Minimize the max / average shear angle of plies due to draping onto complex areas Minimize drag (Fluid-structure interaction and CFD computations)) Economical objective: Minimize the manufacturing cost (provided a realistic cost model)

28 Structural preliminary design and optimization of composite structures 1. Constraints Design rules Manufacturing constraints Maximum allowable mass Damage criterion & models Technological (eg: problematic vibration, water ingress, flame retardant, buckling, ) and geometrical constraints 2. Design space Materials Constituents of the plies (fibres and matrix) Ply thickness and orientation Stacking sequence Draping seed point, curve, splits, order Geometry Position, angles, dimensions Addition or removal of elements (quasi topological approach)constraints

29 Structural preliminary design and optimization of composite structures Multi-objective, multi-constraint optimisation with a discrete & continuous variables appropriate & efficient genetic algorithm enhanced with metamodels (neural network, radial basis functions) for accelerated convergence Direct access to CAD required for shape optimisation CAD Mesh BCs Materials associativity Materials database with appropriate formalism Finite element solver with appropriate features (multi-layer shell elements, failure criteria, ) Other solvers for the technical objectives Dedicated cost model introducing manufacturing constraints In-house libraries & software interfaces

30 Typical optimization workflows for composite structures Minamo Catia CPD Simulayt Composite Link Abaqus Link in Catia Simulayt Composite Modeler Abaqus Seer DFM Minamo Catia FMS-FMD FMD CAPRI Simmetrix CADMesh Mecamesh Samcef (also Samcef Gateway) Seer DFM Automa ation Lev vel Minamo Catia FMS-FMD CAPRI Simmetrix CADMesh MSC Nastran Seer DFM

31 Design and optimization of a satellites dispenser Initial metallic design design violated constraints => composite

32 Design and optimization of a satellites dispenser 2 Objectives: minimum mass minimization of the frequency margin of the first lt laterall vibration mode 352 Constraints: Maximum allowed cost Static, dynamic Buckling Design space: laminates & geometry (25 design variables reduced to 13 after ANOVA)

33 Optimization Loop Based on Samcef Detailed view: opti.prj Minamo TM Minamo.dat - client server ops Client (LINUX) Serveur (Windows) - I/O operations - meshing - Functioning of in house libraries i cost.out post.cc response.out CADMesh Package parameter_modification.py model.stl Script Python Catia V5 geometry_modification.cc Geometry_modification.cc model_reconstruction.py Script Python Simmetrix Serveur (Windows) Client (LINUX) pli.dat SEER DFM mesh_translation mesh.dat tsai.txt mass.dat model.dat disp.txt model.res Samcef V !parameter definition INPUT pli.dat INPUT mesh.dat!end of the definition...

34 Design and optimization of a satellites dispenser Target 1. A slight increase of mass 2. 7% improvement of the margin on the frequency 3. Cost reduced by 12%

35 Sensitivity of structural integrity to draping Nacelle subjected to pressure and with clamping conditions at the interface with the pylon Quasi-isotropic isotropic CFRP laminates with 10 plies Design space: Seed point for plies 1-5 SM = static margin based on the Tsaï-Hill (th) criterion

36 Sensitivity of structural integrity to draping Computational chain Minamo + Catia CPD Simulayt Composite Link Abaqus Link in Catia Simulayt Composite Modeler Abaqus Flat pattern Shear

37 Sensitivity of structural integrity to draping Variable: Seed Point Ply 1 Ply 2 Ply 3 Ply 4 Ply 5 SM Region 1 12% 2% 2% 0% 4% SM Region 2 8% 1% 9% 3% 3% SM Region 3 2% 12% 6% 2% 2% SM Region 4 1% 7% 0% 1% 5% SM Region 5 8% 2% 0% 1% 1% SM Region 6 7% 1% 14% 1% 2% SM Region 7 6% 23% 2% 22% 7% SM Region 8 3% 0% 1% 8% 4% Coupled optimization o of plies orientations and Orientation of plies + draping seed points position influence the behavior of the structural integraty. manufacturing (automated fiber placement.

38 Sampling and meta-modeling Ongoing developments Further development of auto-adaptive adaptive sampling Kriging + Expected Improvement Criterion Surrogate models coupling: local/global - weighted average RBFN (auto-)adaptive fine tuning Support Vector Machines Optimization - Hybridization: Memetic approaches / efficient global-local coupling Exploitation of collective knowledge with multi-parent crossovers (UNDX). Gradient knowledge (SPSA, FDSA, ) to be incorporated in genetic operators, e.g. gradient-based mutation.

39 Conclusion Multiobjective, multidisciplinary optimization for complex design problems is not a dream! But requires: Efficient and powerful optimization engines Careful accuracy, steering and reliability ad-hoc methodologies Fully automated CAD to objective functions computations Enough computing power Thank you for your attention