Multidisciplinary Optimization Methods for Preliminary Design

J. J. Korte, R. P. Weston, and T. A. Zang Multidisciplinary Optimization Branch, MS 159, NASA Langley Research Center, Hampton, Virginia 23681 USA

AGARD Interpanel (FDP+PEP) Symposium "Future Aerospace Technology in the Service of the Alliance April 1997, Paris, France.
2/17/98

Outline
Definitions Requirements for using MDO in Preliminary Design Preliminary Design MDO Examples Summary

2/17/98

MDO Definition
Multidisciplinary Design Optimization (MDO) is a methodology for the design of complex engineering systems and subsystems that coherently exploits the

synergism of mutually interacting phenomena

Design = ( Discipline i) + MDO

i

MDO
2/17/98

MDO Conceptual Elements

Information Science & Technology Product Data Models Data & S/W Standards Data Management, Storage & Visualization S/W Engineering Practices Human Interface Design-Oriented MD Analysis Mathematical Modeling Cost vs. Accuracy Trade-off Smart Reanalysis Approximations Sensitivity Analysis MD Optimization

Discipline Optimization Optimization Procedures Design Space Search Decomposition

2/17/98

Product Data Model Example

(CAD Parametric Geometry Model)

2/17/98

Sensitivity Analysis
Computing derivatives of objective with respect to the design variables Methods
Finite differences
time consuming difficult to pick

Analytic
hard to code changes with each application fast

Automatic differentation
easy to use accurate can be time consuming
2/17/98

Automatic Differentiation of 3-Dimensional Navier-Stokes Flow Code (CFL3D)

High Speed Civil Transport Mach Number = 2.4, = 1
Wing Planform Design Variables (DV)
BC TC OS

IS

Aerodynamic Coefficients CL Lift CD Drag CY Side Force CMY Pitching Moment

RC

Sensitivity Derivatives - Derivatives of Aerodynamic Coefficients With Respect to Wing Planform Variables
CL DV CD DV Cy DV CMy DV

Time to Compute Sensitivity Derivatives (for 4 digits of Accuracy) Automatic Differentiation (Residual reduced 4 orders) = 10.75 units Finite Difference Method (Residual reduced 11 orders) = 15.00 units
2/17/98

Optimization Procedures
Direct Interface analysis cycle iteration sensitivity sensitivity Indirect Interface Using Approximations analysis

optimizer

iteration

local approx optimizer

2/17/98

Decomposition
System Level Optimization
(Coordinates Subproblems) Information Flow

Aerodynamics Optimization Subproblem

Structures Optimization Subproblem

...

Other Discipline Optimization Subproblem

2/17/98

Preliminary Design
Conventional Process
CAD-based geometry
surface internal layout

Higher-order analysis
CFD Finite Element

Discipline analysis & optimization

sequential or loosely coupled discipline-based figure of merits ( i.e., weight, thrust, drag, lift, etc. )

Emerging MD Enhancements
Parametric CAD definition Fully coupled multidiscipline analysis Multidisciplinary optimization
Figures of merit
2/17/98

system performance and cost multi-objective

10

Requirements for MDO Enhancements of Preliminary Design

Information Science & Technology
heavy duty hardware; fast CPU(s), large memory & disk space

common parametric geometry model

software support integration of proprietary, legacy, commercial, and research codes
code robustness, compatibility, & low algorithm noise

configuration control and data management collaborative work environment; person-person/machine

Design-Oriented MD Analysis
well posed interfaces for disciplines automated grid generation (CFD, FEM) discipline & MD sensitivities

MD Optimization
MDO problem definition design variables, objective(s), constraints MDO strategy

2/17/98

11

Preliminary MDO Examples

Aerospike Rocket Nozzle
Direct Optimization Approach

High-Speed Civil Transport (HSCT)

Approximation Optimization Approach

2/17/98

12

MDO Applied to Aerospike Nozzle Design

Multidisciplinary Optimization Branch NASA Langley Research Center

Aerospike Engine
2/17/98

13

Aerospike MDO Problem

Objective minimize Vehicle Gross-Lift-Off Weight Design Parameters 5 geometry variables 13 structural variables Constraints Stresses < allowable

2/17/98

14

Aerospike MDO Domain Decomposition

GLOW Contours

Isp

CFD DOMAIN

T/W TRAJECTORY DOMAIN

BASEFLOW MODEL DOMAIN FEM STRUCTURES DOMAIN

BASE BLEED

CFD-96-OPTIMIZATION DISK/WWF

2/17/98

15

2/17/98

16

Aerospike Nozzle Structural Design Parameters

2/17/98

17

Aerospike Nozzle Optimization

Sequential Optimization (Single Discipline Only) Aerodynamics Maximize Thrust Structural Minimize Weight

Multidisciplinary Optimization

Integrated Aerodynamics and Structures

Minimize Gross-Lift-Off Weight

Base-line Solution

2/17/98

18

Aerospike Objective Function

1.02

Gross-Lift-Off-Weight Ratio

1.00

Optimized Single Discipline Result

0.98

MDO

MDO

0.96

0.94 0 5 10 15 20

Iteration Number
2/17/98

19

MDO Applied to High-Speed Civil Transport (HSCT) Using FIDO

Mach 2.4 at 55,000 ft 6000-mile range 250 passengers

2/17/98

Framework for Interdisciplinary Design Optimization (FIDO)

20

2/17/98

21

2/17/98

22

HSCT MDO Problem Diagram

Flight Conditions Base Geometry Design Variables

Aero Grid Update Rigid Aero Analysis

C L , 0

FEM Node Update Weights Total Weight

Updated Values

Optimizer EoC Weights Sensitivity Derivatives EoC Aeroelastic Static Analysis

EoC CL, CD

SoC Weights SoC Aero/Struc Rigid Analysis Unloaded Shape

SoC = Start of Cruise EoC = End of Cruise = Program = Data

SoC CL, CD

Stresses Fuel Flow Rate

SoC Prop Analysis

EoC Prop Analysis Range

Performance
2/17/98

23

Key Steps in FIDO Aeroelastic Loop

Initial Shape from Design Variables Surface Shape Modifications Aerodynamic Grid Generation Surface Pressures

ADVMOD SURFACE VOLUME ISAAC

FEM Deflections

15,000 cells

Euler CFD

Loads Transfer : Aero to Structures

FEM Structural Analysis

TRN3D COMET
No

7300 Degrees of Freedom

Has Structural Deformation Converged?

Structural Response

STOP
Yes

Converged Shape

2/17/98

24

HSCT Design Optimization

4.0E+05
DEPENDENT DESIGN VARIABLES :

t inbd = C 0 +C1 (1- ) + C2 (1- ) 2 t outbd = C 0 +C3 (1- ) + C 4 (1- )2

3.8E+05 : INDEPENDENT DESIGN VARIABLES C 1, C 2, C 3, C 4
= 1.0

Aircraft Weight, lbs.

t outbd 3.5E+05
t inbd OBJECTIVE FUNCTION : =0

3.2E+05

Weight
CONSTRAINTS:

ginbd = f(K-S) goutbd = f(K-S) 3.0E+05 5 10 15 20

Cycle Number

2/17/98

25

Concluding Remarks
MDO is much broader than just MD-Analysis; it contains elements from information sciences, design-oriented analysis and optimization methods The MDO is the improvement in design obtained from multidisciplinary synergy of the disciplines as demonstrated by the Aerospike nozzle application Application of MDO to preliminary design requires sophistication in the computational infrastructure and MDO algorithms Adoption of MDO in industry design process requires demonstrations which quantify
MDO improvement in design reduction in time and effort in the design process
2/17/98

26