flexible parametric cad models for web-based low … · flexible parametric cad models for...
TRANSCRIPT
CEDC
Flexible Parametric CAD Modelsfor Web-based Low-Noise WingDesign (SIMDAT)
The pursuit of low noise civilaircraft has become a majordriving force for technologyinnovation in the aerospacedesign process. Design of lownoise aircraft is a complexprocess involving manydisciplines and seamlesscollaboration betweendisciplines is the key to thesuccess of such process. Datalies at the heart of thecollaboration scenario wherecomputational modelling andsimulations are essential todelivering low-cost, high-performance products. TheEC-funded SIMDAT projectfocuses on the use ofadvanced Grid technology topromote collaboration in fourapplication areas. In theaerospace application area,the design of a low-noisewing is chosen as the test bedfor available and emergingGrid technologies.
SIMDAT Aerospace– DataGrid forAerospace DesignThe SIMDAT aerospace datagrid aims to provide adistributed, collaborativeinfrastructure and tools tosupport design of complexaerospace systems, using thedesign of a low-noise, high-lift wing as an exemplar. Theproject involves three mainpartners, each specialising indifferent disciplines. Gridtechnologies, and Web Serviceaccess to proprietary analysiscapabilities in particular,
CEDC
variables provided bydesigners will produce thegeometry file. The geometryfiles in IGES format will thenbe passed to a meshgeneration Web service togenerate the mesh for meanflow CFD analysis andacoustic analysis. Theparametric definition of themodel will allow designers tomodify the design easily andstudy the tradeoffs betweendifferent choices. And after thewhole dataflow has beenautomated and validated, theparametric definition of themodel will allow it to becoupled with optimisers toexplore the design space inorder to locate near optimumsolutions. Reliable, fault-tolerant and secure dataexchange and serviceinvocations are the key toproviding a collaborativeprocess across institutionalboundaries.
Flexible ParametricCAD Models asServicesThe design of a flexibleparametric wing model is thefirst task of the project atSouthampton. The winggeometry can beparameterised in variousways, from a moremathematical definition to amore engineering intuitiveway based on geometricparameters such as leading-edge radius, maximumthickness, etc. Here the airfoilsection of the wing ismodelled using Non-UniformRational B-Splines (NURBS)and the locations of the controlpoints are used as designvariables, however, to bettercontrol the range of the designspace and reduce theprobabilities of producingunrealistic geometries, non-dimensional parameters areused in the definition of thecontrol points as shown inFigure 3. This will lead touniform ranges for all thedesign variables from 0 to 1and eliminate the need toconstrain the relative location
of control points, as is the casewhen absolute coordinates areused. 13 Control points willbe used in the definition ofone airfoil section and therewill be 20 design variables intotal for the airfoil section.The profile shapes andlocations for the flap will alsobe parameterised, while theshape of the fuselage will befixed in the problem. Anillustration of the final shapethat will be designed is givenin Figure 4.
Figure 3: NURBSParameterisation of AirfoilSections
Figure 4: Aircraft Wing Geometry(to be parameterized)
With the tools andtechnologies being developed,design and analysis serviceswill be composed together toform a task specific workflowthat can be used in the designprocess. Working in acollaborative way, designteams will be able draw onexpertise from variouspartners and deliver betterdesigns faster.
SIMDAT is funded by theEuropean Commission under theInformation Society TechnologiesProgramme(IST), contractnumber IST-2004-511438
Wenbin Song ([email protected])Computational Engineering Design GroupSchool of Engineering SciencesUniversity of SouthamptonHighfield, SouthamptonSO17 1BJ
provides an ideal vehicle forseamless collaborationtowards common targets.Figure 1 shows thearchitecture of the applicationscenario, in which the mainpartners and their area ofexpertise are illustrated.
Figure 1: SIMDAT AerospaceScenario (Boeing may join phaseII)
Figure 2: Dataflow for SIMDATAerospace Application
Services andDataflowSIMDAT Aerospace willprimarily contain threeservices: Design Services forproducing the geometrydefinition for analysis basedon input from designers;Aerodynamic Services formean flow analysis usingCFD; Aeroacoustic Analysisfor predicting the noise levelsfor the configuration.Structural analysis may alsobe introduced as a constraintin the design process. Thedata flow between services isillustrated in Figure 2. Thegeometry is defined in CATIAV5 as a parametric model,which, based on the input
DesignService
AeroacousticServices
AerodynamicSerivces
Mean flow
Geometry
Lift/Dra
g
Noise (PNL)Geometry
Structuralanalysis
source
Design problem specification
Design space specification
Product model generation
Product structure generation
Aerodynamics Services Structures Services Aeroacoustics Services
PSE
tools
resources
workflows
database
PSE
tools
resources
workflows
PSE
tools
resources
workflows
Design Optimisation Services
PSE
tools
resources
workflows
database
databasedatabase
Aerospace Data Interoperability Service
Design portals(GUI/scripting)
Design problem specification
Design space specification
Product model generation
Product structure generation
Aerodynamics Services Structures Services Aeroacoustics Services
PSE
tools
resources
workflows
database
PSE
tools
resources
workflows
PSE
tools
resources
workflows
Design Optimisation Services
PSE
tools
resources
workflows
database
databasedatabase
Aerospace Data Interoperability Service
Design portals(GUI/scripting)
This article may be found athttp://www.soton.ac.uk/~cedc/posters.html