multifidelity modeling and multidisciplinary optimization
TRANSCRIPT
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