f"

William W. Follett

MDO Project Lead

Raj RajagopalRDCS Project Lead

::_. Rocketdyne Propulsion and Power

_i:i _ Boeing- Canoga Park

_i"i! : . Presented at:

mW_._?_"_-_ _ Air Force Research Lab (AFRL) Electronic Prototyping Review

_!{_,._ Boeing - SeattJe

_4i_:_! : July2s 2oo_

AGENDA

• Vision

• Approach

ji.i!;_i • Optimization and Robust Design',i • Knowledge Based Systems_- • Collaborative Technologies

_) • Examples

__ :.,- S.gmmary

https://ntrs.nasa.gov/search.jsp?R=20020022115 2019-08-26T04:04:54+00:00Z

:::" f

'Real time" analysis enable rap_

design, multidisciplinary anaiysis

complete mission predictions &

manufacturingsimulations

i ',_ _ ,_,._, . - -'J ,_C C

_: _I_!_ - _

RPP engineers work fluently with

collaborators across the globe, providing

the best so_Jtions by combiningworldwide resources

Automated Reasoning

Expert systems capture design

knowledge. Automated data analysis,

optimization, and robust design

: generate new knowledge.

t( SsemsanMDOCapureanExpand Design Knowledge

Approach

_i!ntegrate design technologies into state of the art design systems

B_: -- Reduce time & cost required to create new products

_! Capture engineering expertise

,_2 Provide users with design environments customized to their problem

_l_p_._¢_n t__ hhology thrusts:

• -tii'fiization ar_l Robust Design

OllaDoratlve_ i ecnnologles

_,_ ...,_r'_:_;/_'-._:_'_ ': -;, ..

Technology Thrust: Robust DesignChallenge: Reduce De_,elopment Coxt

Traditional Design Process

Time

Robust Design Process

Cost

Eliminate Failure Modes

Engineering Supporl

Design

Certification

Time

Test-Fail-Fix Cycle

Representing 73% of Cost

REDUCE BYFACTOR OF 4

- ..... Robust Design Computation SystemRDCS Features

i:

<_': .... File S_'sl_fil M onilor Help

o,[kflow aids robust design ....................................

_ables fromdatabases IlilJ_=_-__il_0delsincluda Vadablesl Math[

i_dai'dized and automated ------ Model! Des-.ii_.I_sls processes _-- ProcessillgW_ -_-_" -_: '-- .... , Compute

im_m_'_ , ._;' _-,_,x., : Results t_U)[OCesses _SVallaDle _

/.u_J_#meg__omputlng

RDCS Probabilistic Analysis - Results

Plot Tvr,e t_)p_m:,:ot

X _is kabe!

I1

Technology Thrust: Optimization

Implement Optimization Tools to Support Increased Design Complexit),

Technology Thrust: Optimization

hnplement Optimization Tools to Support Increased Design Complexi_

Technology Thrust: Knowledge Based Systems

• Future role of analysts/designers can become that of

i: i:i: owner/developer of expert shells or 'knowledge stations'

., • The expert shells interact with each other

'. semi-autonomously

Intelligent Design Advisor CapturesHypersonic Propulsion Expertise

:__ ",'__;iL-,T----,_"r.................

IDA System Architecture -- Engine Analysis Module

IDA System Architecture

/

' Inner Driver

Technology Thrust: C.llab.rative Technologies

Collaborative optimization algorithms• Enable MDO among geographically dispersed teams

• Address MDO problems of unprecedented size and corpJexrty

_,_;_X, • Internet-Based Parallel Processing Tools_. t ° Enable controlled access among design team members to__;._ :. geograp.hically dispersed computational resources

_:'_:_,_ • _',.-, . >L,,.';; ., ..___'},-"_ t.:J._Eacditate effect ve use of ava lable computat ona resources for

ii_a_un_l_e.g., parallel process,ng)

l_V, irt0al collocation of personnel.:[[;: '_r_:_:- _:_;_"! _-_ ,.>_: :

i:;_:_;'L_5 / " : : - " , BO'_/_O

_liul ..... =.lrl ....... , .............

AGENDA

_<--. • Vision

: ° Approach

_:- ° Optimization and Robust Design

l_i{ ° Expert Systems. TechnologiesCollaborative

_ .t._ ? d_ *" " " - ' ,B'OEI.,_

Potential to Reduce RLV Vehicle Weight Demonstrated through

Multidisciplinary Optimization" (MDO)

I Baseline- OptimizedSingleDisciplineWeight0.00I

MDO

J

leduction achievedbyMDO ,.

° Ref AIAA 97-3374, NASA TM 110326

10 15

CASINO Reduces Cost and Complexity of

Aerospike Design by an Order of Magnitude

;I31[ & Automation

+ +

Captured Design & Parallel Computing

Analysis Knowledge & Optimization

• For 2000: _'_"

• $5M added engineeering value |• 4000% return on investment in first _ear

• 61% improvement in predicted

payload-to-ISS

Multidisciplinary Optimization is Now a Design Tool

Triprope/lant Nozzle 3-O Supersonic Stator Vane

• Maximize thrust and minimize hot spots

load reduction, • Match target pressure profiles• 90% improvement achieved

Jetpump

Integrated Development

Multidisciplinary Linked Analysis & Design

Hydrodynamics analysisblade and environment definition Mechanical design (Pro-E)

Pressure application

Model

(ANSYS)

Stressresults

Impeller C_clc Time Reduced - 6 Months to 1 V<eek

i

Rule-based System for Combustion Cliamber Design

New Project I Existing Project

Integrated Design Systems reduce Cost &

Cycle Time for Engine Development

Synthesize MDO, ES and collaborative technologies intoturnkey analysis and design systems

• User-oriented systems in language that user understands

- - Capture key engineering expertise

Reduce design and life cycle costs

_"* Improve product quality

:, Reduce expertise level required by the user

ties:(BIVDS)- Scramjet engine design

. TUrbine optimization program

) network analysis shell

.=Information Powergrid

pike nozzle design

,_'OE/,4Y_"

Summary

The focus of the AA MDO team is to reduce product developmentcost

• Capture & automate best design & analysis practices• Increase availability of low-cost, high-fidelity analys,s

:, • Implement robust design to reduce costs associated with the_'_._.:: Test-Fall-Fix cycle

__e_ntly focusing on several technologies to improve the

_grl " ,.__ i.

les I:-I/L :

.......... I I IIII Ill ......