Sunteți pe pagina 1din 4

AB2030 Rev.07.

09

Designing for Cost Reduction


Red Cedar Technology, Inc.

Introduction Wenormallythinkofoptimizationintermsof reducingmassorincreasingperformanceofa product,butoftentheunderlyinggoalistoreduce productcostwhileachievingatargetperformance. Productcostmaybeafunctionofmanyfactorssuch asmaterialtype,amountofmaterialused, manufacturingprocess,numberofparts,assembly details,andmore. Therelationshipbetweencostanddesignvariables (e.g.,shape,material,gaugethickness,locationof features,etc.)ishighlyimplicit.Inotherwords,itis oftendifficulttounderstandthedirecteffectthat changingoneormorevariableswillhaveoncost. Butmathematicaldesignoptimizationhastheability toefficientlyexplorethesecomplexrelationships. Forthisreason,itisapowerfultoolforachievinga costeffectiveproductdesign. Itistemptingtoreducedesigncomplexityby focusingonjustonetypeofdesignvariable.For example,costreductioneffortsmayconsideronly reducinggaugethickness,orchangingmaterialtype, orchangingshapetoreducematerialusage.By themselves,anyoftheseapproachesmayyieldsome benefitintermsofcostreduction.Butthebiggest costsavingsareusuallyachievedbytaking advantageofthestrongcouplingamongshape, materialproperties,gaugethickness,andother variables. Forexample,thebestpartshapeforonematerial maybeverysuboptimalforanothermaterial becausethestiffnessandstrengthcharacteristicsof eachmaterialaredifferent.Consideringbothshape andmaterialtogetherduringthedesignoptimization processoftenleadstoamoreoptimaldesign,one withsignificantlylesscost. Further,theleastexpensivematerialmaynotalways leadtothelowestcostdesign,sinceperformance andmassareoftenstronglyaffectedbyachangein material.Simultaneouslyoptimizingthedistribution andtypeofmaterialwithineachcomponentisthe

bestwaytoachievebothhighperformanceandlow cost. Here,wedemonstratehowtocoupleacostmodel andafiniteelementmodelwithinanoptimization processtoobtainaminimumcostdesignthatalso meetsperformancerequirements.Theapplication exampleisanofficechairlegwithbothmaterialand shapedesignvariables.HEEDSProfessionalisusedto automatethedesignevaluationandoptimization process,Abaqusisusedtoperformfiniteelement analysis,andExcelisusedtocalculatethecostof eachproposeddesign. Analysis Model Asinglelegofatypicalfivelegofficechairwas designedtominimizeitscost.Thebaselinedesignis showninFigure1.Thematerialandtheshapeofthe chairlegaretreatedasdesignvariables.Thepossible materialchoicesare:Aluminum,CastIron,Low StrengthSteel,andHighStrengthSteel.Thesix shapedesignvariables,asshowninFigure2,arethe height,thicknessandlocationoftheinternalrib,the linearlyvaryingsidewallthicknessprofileandthe floorthickness.Theparametricgeometrymodeland associatedfiniteelementmodelaredeveloped withinAbaqusCAE.

Figure1.Chairlegtobeoptimized.

Page|1

DesigningforCostReduction Table1.Baselinedesignfiniteelementresults.
DesignParameter Back wall thickness Material RibLocation(mm) RibHeight(mm) Rib height Front wall thickness RibWidth(mm) FloorThickness(mm) FrontWallThickness(mm) Rib location Rib width BackWallThickness(mm) Response Floor thickness Volume(mm3) MaxStress(MPa) MaxDisplacement(mm) MaxRotation(radians) Value Steel_LowStrength 100 15 15 8 6 9 Value 2.848E5 350.14 1.005 0.0216

Figure2.Chairlegshapeparameterstobedesigned.

Thenonlinearstaticfiniteelementanalysisis performedusingAbaqus/Standard.Forallmaterials exceptcastiron,mildnonlinearityisassociatedwith materialplasticityastheloadingisallowedto exceedthematerialyieldstressslightlyinthemost efficientdesigns.Thelegisloadedasshownin Figure3.Itisfixedatthebaseofthechair,anda pointforceandapointmomentareappliedatthe castersockettosimulaterollingofthechairona frictionalsurface.Figure4showsthescaled deformationunderloading.Table1containsthe geometricdetailsandanalysisresultsforthe baselinedesign.

Excel Model for Cost Analysis and Failure Index Computation Thecostofeachdesigndependsonthetypeof material,itsdensityandtheamountofmaterial used.Therelativecostofeachdesignwascalculated asthemassofthedesignmultipliedbytherelative materialcost.Therelativecostisadimensionless (normalized)quantitythatprovidesarelative,rather thananabsolute,measureofcost.Thedensityand relativematerialcostofeachmaterialareprovided inTable2. Themaximumstressineachdesignwascalculated usingtheAbaqusfiniteelementmodel.Becausethe allowablestressforadesigndependsonthe materialchoice,astressfailureindexwasusedto assessperformancerelativetothestressconstraint. Thisindexwasdefinedasthemaximumstress dividedbythematerialsallowablestress.Avalue lessthan1.0forthestressfailureindexindicateda designthatsatisfiedthestressconstraint.The allowablestresswasdefinedasthestressat0.1% plasticstrainforthegivenmaterial(orforCastIron, theyieldstress).ThisdataisprovidedinTable2.
Table2.Materialdatausedinthestudy.
Material Aluminum LowStrength Steel HighStrength Steel CastIron Density 6 3 (10 *kg/mm ) 2.71 7.85 7.85 7.30 Relative MaterialCost 1 (kg ) 7.9 5.3 6.7 4.7 Allowable Stress (MPa) 162.2 434.8 586.2 379.0

Figure3.Chairlegloadingandboundary conditions.

Figure4.Scaledchairlegdeformation.

Page|2

DesigningforCostReduction

Thecostandperformancemodelwascreatedusing MicrosoftExcel.Aworkbookwasdevelopedto calculatetherelativecost,mass,andstressfailure indexforagivenchairlegdesign.Theworkbook requiredasinputthematerialtypeandthe maximumstressandvolumefromtheFEmodel. HEEDSprovidedthesevaluestotheworkbook duringoptimization.Figure5showsthecalculations forthebaselinedesigndefinedinTable1.

Chair Leg Optimization Theoptimizationstudyconsideredbothdiscrete (material)andcontinuous(shape)variables.The searchwasperformedusingtheSHERPAalgorithm withinHEEDSProfessional.Theoverallprocessflow isdescribedinFigure6.Thegoalofthestudywasto minimizethecostofthechairleg,withconstraints onthestressfailureindex,deflection,androtation. Themathematicaloptimizationstatementis:
Objective: Subjectto: Byvarying: minimizerelativecost stressfailureindex<1.0 maxdisplacement<5mm maxrotation<0.15radians 10mm<riblocation<250mm 4mm<ribheight<20mm 4mm<ribwidth<30mm 1mm<floorthickness<10mm 4mm<frontwallthickness<10.75mm 4mm<backwallthickness<10.75mm Materialfromthediscreteset{Aluminum, CastIron,Steel_HighStrength, Steel_LowStrength}

Figure5.Baselinedesigncostmodel.

HEEDS
Material Variable Values All

Abaqus/CAE
Abaqus Replay File

Execute Abaqus/CAE

SHERPA
Abaqus input deck

Response Values

Abaqus
Execute Abaqus Std

Displacement and Rotation Abaqus Portal Stress and Volume Abaqus *.odb file

Cost and Failure Index

MS Excel Output Portal

Execute MS Excel

MS Excel Input Portal

Figure6.ProcessflowchartforoptimizationofthechairlegutilizingHEEDSwithAbaqusandExcel. Page|3

DesigningforCostReduction

Optimized Design HEEDSfoundadesignthatreducedthecostby56% overthebaselinedesign,whilemeetingallofthe performancerequirements.Thoughthebaseline designwasnotconsideredtobeahighquality design,thisresultdemonstrates,nonetheless,the abilityofthisapproachtoidentifylowcostdesigns thatmeetperformancetargets.Theoptimized designhadthedesigncharacteristicsofTable3and thecostmodelresponsesshowninFigure7.Figure8 showstheshapeoftheoptimizeddesign.InFigure9, thepropertiesoftheoptimizeddesignaredisplayed inaparallelplotalongwiththeresultsfromother designsevaluatedduringthesearch.Itshouldbe notedthattheheightoftheribisatitsminimum value,indicatingthatthisribisnotneededinorder tomeettheperformancecriteriaasdefinedhere.
Table3.OptimizeddesignresultsfromFEAwithAbaqus.
DesignParameter Material RibLocation(mm) RibHeight(mm) RibWidth(mm) FloorThickness(mm) FrontWallThickness(mm) BackWallThickness(mm) Response Volume(mm3) MaxStress(MPa) MaxDisplacement(mm) MaxRotation(radians) Value Aluminum 247.6 4.0 20.676 7.95 6.43 4.2 Value 2.42E5 162.7 4.57 0.1

Figure7.Optimizeddesigncostmodel.

Figure8.Optimizeddesignshape.

Figure9.HEEDSparallelplotindicatingtheoptimizeddesignamongallotherdesignsevaluatedduringthesearch. Page|4

S-ar putea să vă placă și