Multifidelity Modeling andMultidisciplinary Optimization Techniques for

Environmentally-Sensitive Aircraft Design

AEROSPACE COMPUTATIONAL DESIGN LABORATORY

Karen WillcoxAerospace Computational Design LaboratoryDepartment of Aeronautics and Astronautics

Massachusetts Institute of Technology

UTIAS-MITACS International Workshop on Aviation and Climate ChangeUniversity of Toronto Institute for Aerospace Studies

May 29, 2008

Acknowledgements

• Krzysztof Fidkowski, Andrew March, Theresa Robinson,Ian Waitz (MIT)

• Ilan Kroo (Stanford University)

• Funding• NASA Subsonic Fixed Wing Project,

Technical Monitor Steve Smith• FAA Office of Environment and Energy,

Program Manager Maryalice Locke

Any opinions, findings, and conclusions or recommendations expressed in thismaterial are those of the author(s) and do not necessarily reflect the views of

the FAA, NASA or Transport Canada.

Outline

• Environmental tradeoffs and interdependencies

• Multidisciplinary design optimization andmultifidelity modeling

• Examples:• Integrating design and operations• Advanced technologies for environmentally-sensitive

aircraft design• Integrated decision-making: Aviation Environmental

Portfolio Management Tool

Environmentally-Sensitive Aircraft Design

• Future aircraft must satisfy an increasingly diverse set ofdesign requirements, many driven by stringentenvironmental constraints

• Requirements necessitate the use of advancedtechnologies, new operational strategies, and novelconfigurations• Traditional conceptual design models rely heavily on empiricism and

past experience

• Research Objectives• Evaluate trade-offs between cost

and environmental impacts duringaircraft conceptual design usingmultidisciplinary optimization

• Develop multifidelity optimizationframework that can accommodatecurrent and new designs

• Investigate new design concepts

System Interdependencies are Key

900m

150m

650m

AirportPerimeter

2000 m fromtouchdown

3357 & 4164 m frombreaks-off point

PlumeCenter

Height

Width

An integrated approach to design anddecision-making is essential.

OperationsTechnology

NoiseEmissions

DesignRegulation

Performance

MaximumTakeoffWeight

BWB

A3XX-50R

18%

19%BWB

A3XX-50R

OperatorsEmptyWeight

Cost

Multidisciplinary Design Optimization (MDO)

• Multidisciplinary approachto design

• Numerical optimizationprovides a systematic wayto explore the design space

• Problem formulation is notobvious and requiresengineering judgment

• Challenges:• Incorporation of environmental, financial, operational effects• Incorporation of new technologies, unconventional aircraft concepts• Use of high-fidelity tools in early design at the system level• Accounting for uncertainty and variability

Source: Kroo (Stanford),http://adg.stanford.edu

MDO: Aircraft Conceptual Design

• Traditional multidisciplinaryconceptual design tool must besupplemented with:• Environmental models

• Operational design space

• Models of new technologies

• Higher-fidelity (physics-based)models

• Uncertainty must also bequantified and accounted forin the decision-making process. Program for Aircraft Synthesis

and Sizing (PASS), I. Kroo,N. Antoine, StanfordUniversity.

Example: Integrating Technology and Operations

• Models that represent operations in existing design toolsare not suitable for detailed assessment of environmentalperformance

• Existing detailed operational assessment models are notsuitable for a vehicle design framework

• Must account for uncertainty in operations

900m

150m

650m

AirportPerimeter

2000 m fromtouchdown 3357 & 4164 m from

breaks-off point

LTO Analysis Requires Detailed Low-SpeedAerodynamics Model

Low-speed aero model for the EnvironmentalDesign Space, A. March (MIT).

Environmental Assessment RequiresParameterization of Aircraft Operations

• Operational design variables:- flap setting- throttle setting- true airspeed- end-of-procedure altitude

• Objective functions:- time to climb- fuel burn- noise

Parameterization of departure trajectory for optimization

Takeoff noise certification points

Optimization Results: Tradeoffs AmongEnvironmental Performance Metrics

363 sq. mi. 364 sq. mi. 357 sq. mi.

Example: Advanced Technologies forEnvironmentally-Sensitive Aircraft Design• Tighter coupling between aerodynamics, structures, and

control early in the design process can yield a concept withdramatically improved fuel efficiency and thus dramaticreductions in CO2 emissions.

• Expected design characteristics:• Very high bypass ratio engines• Low wing sweep for extended laminar flow. Restricts section

thickness which places more critical demands on wing structure.• Slatless wing for reduced noise and simpler compatibility with

laminar flow. Leads to reduced wing loading and higher gust loads.• High wing span for reduced lift-dependent drag.• Gust load alleviation via active control of

wing trailing-edge devices.

Conceptual Design of Environmentally-SensitiveAircraft Requires New MDO Approaches

• More detailed controls models thanusual in conceptual design

• Close integration of aerodynamics,structural dynamics, and controls

• For some disciplines, physics-basedmodels of the desired fidelity aretoo expensive to be used within theoptimization process

• Proposed approach: multifidelityoptimization with a hierarchyof models

Gust encounter studiesK. Fidkowski (MIT)

Conceptual Design: Multifidelity Models

• Reduced complexity of f(¢) and c(¢)• Simplified physics• Model reduction• Other surrogate models

(data fit, multigrid, etc.)

• Reduced complexity of u• Different design descriptions• Decreased resolution of

design representation

J.J. Alonso, I.Kroo, StanfordUniversity

Medium-fidelity:ASWING(lifting line)

High-fidelity:Cart3d(Euler CFD)

Low-fidelity: analytical(Wagner/Küssner)

Conceptual Design: Multidisciplinary DesignOptimization with Multifidelity Models

Collaborative work withK. Fidkowski (MIT), I. Kroo(Stanford), E. Cramer (Boeing),F. Engelsen (Boeing)

Integrated Decision-Making: AviationEnvironmental Portfolio Management Tool

APMT

Policy scenarios•Certification stringency•Market-based measures•Land-use controls•Sound insulation

Market scenarios•Demand•Fuel prices•Fleet

Environmental scenarios•CO2 growth

Technology andoperational advances•CNS/ATM, NGATS•Long term technologyforecasts

Cost-effectiveness•$/kg NOx reduced•$/# people removed from65dB DNL•$/kg PM reduced•$/kg CO2 reduced

Benefit-cost•Health and welfareimpacts•Change in societalwelfare ($)

Distributionalanalyses•Who benefits, who pays•Consumers•Airports•Airlines•Manufacturers•People impacted bynoise and pollution•Special groups•Geographical regions

Global, Regional, Airport-local

inpu

tsoutputs

(reports available at www.partner.aero) An FAA/NASA/TC-sponsored Center of Excellence

APMT PARTIALEQUILIBRIUM BLOCK

NOISE IMPACTS

LOCAL AIR QUALITYIMPACTS

CLIMATE IMPACTS

APMT COSTS & BENEFITS

New Aircraft

Emissions

Noise

APMT BENEFITSVALUATION BLOCK

Monetized Benefits

CollectedCosts

Emissions

Emissions & Noise

Policy and Scenarios

AEDT

Fares

DEMAND(Consumers)

SUPPLY(Carriers)

Operations

Schedule&

Fleet

Airport-levelEmissions

Tools

GlobalNoise

Assessment

Airport-levelNoiseTools

GlobalEmissionsInventories

EDS VehicleNoise

Design Tools

VehicleEmissions

Design Tools

TechnologyImpact

Forecasting

VehicleCost

Assessment

DesignTools

Interface

APMT: ApproachAviation Environmental Portfolio Management Tool (APMT)Aviation Environmental Design Tool (AEDT)Environmental Design Space (EDS)

BB&C

Vital LinkPolicy Analysis

HARVARD SCHOOL OFPUBLIC HEALTH

APMT: Representation of Uncertainty and Preferences

Use of “Tornado” charts to communicateuncertainty sources and importance of policymaker choices

Preliminary Results Only--Do not cite

0 0.5 1 1.5 2 2.5 3 3.5 4

NPV (USB$2005)

Discount

Rate

Quiet

Level

Housing

Growth

Rate

NDI

Contour

Uncertainty

5% 1%

Historic

-2%

Historic

+2%

Valuation:

Scenario:

Scientific:

Input Distributions

1% 3.5% 5%

HistoricLow High

50 dB 55 dB52.5 dB

mean = .6651%

std=.104%

-2 dB 0 dB 2 dB

Nominal case

50 dB55 dB

0.561% 0.769%

-2 dB 2 dB

Significance

Level 65 dB 52.5 dB Quiet

Lev.60 dB 65 dB

1.90 0.5 1 1.5 2 2.5 3 3.5 4

NPV (USB$2005)

Discount

Rate

Quiet

Level

Housing

Growth

Rate

NDI

Contour

Uncertainty

5% 1%

Historic

-2%

Historic

+2%

Valuation:

Scenario:

Scientific:

Input Distributions

1% 3.5% 5%

HistoricLow High

50 dB 55 dB52.5 dB

mean = .6651%

std=.104%

-2 dB 0 dB 2 dB

Nominal case

50 dB55 dB

0.561% 0.769%

-2 dB 2 dB

Significance

Level 65 dB 52.5 dB Quiet

Lev.60 dB 65 dB

1.9

Inputs

Benefits Valuation Noise Module

APMT: Expected Outcomes and Practical Applications

• Expected outcomes• Deliver APMT component of an integrated policy-

analysis framework to support ICAO/CAEP and JPDOdecision-making

• Aviation Environmental Portfolio Management Tool (APMT)• Aviation Environmental Design Tool (AEDT)• Environmental Design Space (EDS)

• Practical applications• Support ICAO/CAEP decision-making

(goal = CAEP/8, 2010)• Support JPDO/NGATS decision-making• Identify high leverage uncertainties and research needs

Summary

• Integrated approach to decision-making is key:we cannot focus on just a single environmental metric

• Significant opportunities exist for reducing environmentalimpact of future aircraft• Requires a highly-integrated, multidisciplinary approach to design

• Uncertainty must be quantified and accounted forin the decision-making process