Integrating multi-body simulation and CFD: toward complex ...€¦ · – Catia V5: Parametrization. ... easy-to-use environment. CATIA, ICEM, CFX, ADAMS have been integrated in a

Integrating multiIntegrating multi--body simulation and CFD: body simulation and CFD: toward complex multidisciplinary design optimisationtoward complex multidisciplinary design optimisation

Federico UrbanESTECO

Italy

Martin MühlmeierAUDIGermany

Stefano PieriDepartment of EnergeticsUniversity of TriesteItaly

Courtesy of

This work show-cases how to carry out a multi-disciplinary design process couplingall the tools generally involved in a complete multi-body analysis.The integration will be managed by modeFRONTIER MDO package

The track: Le MansThe race-car:

Audi R8

PresentationPresentation OutlineOutline

Courtesy of

PresentationPresentation OutlineOutline

Optimisation Goals;The physics behind the problem: Aerodynamics Multi-bodyDynamics;The numerical analyses for simulating the real-life complexity: Tools;How to put together the numerical tools for achieving the result: Methods;Optimisation Results.

Courtesy of

GoalsGoals

INCREASE VEHICLE PERFOMANCES

Objective: Minimization of Lap-timeInput Variables: Geometrical Entities

Courtesy of

AerodynamicsAerodynamics

Aerodynamics is a crucial issue in the design of a high-speed vehicle.

Porsche911-GTO (McNish) Le Mans '98

It is useful and “safer” to simulate in advance the Aerodynamics at different race scenarios

Courtesy of

NumericalNumerical AnalysesAnalyses

For achieving a complete and reliable numerical simulation we shouldconsider the impact of the aerodynamic forces on the mechanicalbehaviour of the vehicleIn practice, carry out the coupled numerical solution with CFD tool(CFX) and Multi-Body tool (Adams) using a three-dimensional model.

Courtesy of

ToolsTools

The Numerical Tools SCENARIO:• 3-D Analysis• CAE Tools:

– Catia V5 : Parametrization.– Icem CFD 4.3: Mesh Generation. – CFX 5.6: Fluid-Dynamics Simulation.– MSC-Adams: Multi-Body Dynamics Simulation.– modeFRONTIER: “the wrapper”

• Multiple CFD analyses at various positions of the car

Courtesy of

ToolsTools: : CatiaCatia V5V5

The parametric model:

• Coordinates of the fundamental points of the diffuser (Xpar, Zpar).

• Inclination and the Height of the rear wingprofile (H2, Alpha).

Courtesy of

ToolsTools: : IcemIcem CFD 4.3CFD 4.3

The finite-element model: Hybrid mesh (tetra+prisms) - 3 millions cells

Courtesy of

ToolsTools: CFX 5.6 : CFX 5.6 -- BoundaryBoundary conditionsconditions

INLET

v=44 m/sOUTLET

Wall no slip v=0Wall no slip (v=44 m/s)

SimmetryRotating wheels

Courtesy of

ToolsTools: CFX 5.6 : CFX 5.6 -- ResultsResults

PressureVelocity

Courtesy of

ToolsTools: MSC: MSC--ADAMS ADAMS CarMotorsportCarMotorsport. .

Simulating vehicle dynamics

• The global model is represented by means of a group of mechanical elements with specificcharacteristics.

• The structure of the complete car model has a pyramidal layout.

Courtesy of

MethodsMethods: The Idea!: The Idea!

Original Geometry

CFD Simulations

Dynamic Simulation

Aerodynamic MatricesNew Geometry

LAP time

Optimization

Algorithm

Courtesy of

MethodsMethods: The Design : The Design FlowFlow

Parmacro.CATScript

H2, Alfa, Xpar, Zpar

AUDI_box.model cfx5macro

ICEM CFD CFXCATIA ADAMS LAP time

Courtesy of

MethodsMethods: Design : Design ProcessProcess RequirementsRequirements

In order to achieve the most reliable map of aerodynamic forces to be used as boundary conditions along the multi-body analyses, eachdesign is evaluated in 12 different vehicle body positions

Courtesy of

MethodsMethods: The Design : The Design FlowFlow

Parmacro.CATScript

H2, Alfa, Xpar, Zpar

H1, BetaBody

n < 12AUDI_box.model

Matrix 3x4 Matrix 6x8

ICEM CFD CFX ifCATIA ADAMS LAP time

D.A.C.E. (Design and analysisof computer experiment), external Response SurfaceModeller

DACE

With 12 simulations, it is possible to extrapolate the complete 6x8 matrix of the aerodynamic forces required by ADAMS.

Courtesy of

MethodsMethods: The Design : The Design FlowFlow

Parmacro.CATScript

H2, Alfa, Xpar, Zpar

H1, BetaBody

n < 12AUDI_box.model

Matrix 3x4 Matrix 6x8

ICEM CFD CFX ifCATIA ADAMS LAP timeDACE

modeFRONTIERCourtesy of

MethodsMethods: : ProcessProcess ItegrationItegration with with modeFRONTIERmodeFRONTIER

Courtesy of

MethodsMethods: : ProcessProcess ItegrationItegration with with modeFRONTIERmodeFRONTIER

Courtesy of

MethodsMethods: Complete : Complete ProcessProcess FlowFlow

Courtesy of

MethodsMethods: Input : Input ParametersParameters RangesRanges

50 mm-50H2 (mm)

2-2Alfa (degrees)

-94,124-244,124Zpar (mm)

28862736Xpar (mm)

Upper BoundLower BoundParameter

Courtesy of

ResultsResults: : HistoryHistory ChartChart –– Lap Time vs. Design Lap Time vs. Design IDID

Simplex Algorithm

RESULTDesign 17 improves2,36 sec. the lap time

Courtesy of

Minimum height Maximum inclination

ResultsResults: CFD : CFD PlotsPlots –– Original vs OptimalOriginal vs Optimal

Cd coefficient: -15 %Cl coefficient: +6.0 %

Cd coefficient: -11 %Cl coefficient: -12 %

Courtesy of

VELOCITY

FUEL CONSUMPTION

ResultsResults: : DynamicsDynamics DiagramsDiagrams –– Original vs Original vs OptimalOptimal

OptimalOriginal

Courtesy of

Conclusive Conclusive remarksremarks

• modeFRONTIER managed a real-life multi-disciplinary optimization problem in a easy-to-use environment. CATIA, ICEM, CFX, ADAMS have been integrated in a process integration framework.

• The design chain worked successfully achieving virtually 2 seconds reduction of the lap time.

• Each the Process Integration issue and the Design Optimization problem have been reliably solved

Courtesy of