JournalofMechanicalScienceandTechnology24(2010)145148/content/1738-494xDOI10.1007/s12206-009-1126-5OptimalinjectionmoldingconditionsconsideringthecoreshiftforaplasticbatterycasewiththinanddeepwallsDong-GyuAhn1,*,Dae-WonKim2andYeol-UiYoon31DepartmentofMechanicalEngineering,ChosunUniversity,Gwang-Ju,501-751,Korea2IncheonServiceDivision,KITECH,Incheon,406-840,Korea3Chun-BokInjectionMoldInc.,Gwang-Ju,1233-12,Korea(ManuscriptReceivedApril24,2009；RevisedSeptember21,2009；AcceptedOctober12,2009)-AbstractTheobjectiveofthispaperistoexaminetheinfluenceofinjectionmoldingparametersonthecoreshifttoobtaintheoptimalinjectionmoldingconditionsofaplasticbatterycasewiththinanddeepwallsusingnumericalanalysesandexperiments.Unlikeconventionalinjectionmoldinganalysis,theflexiblepartsofthemoldwererepresentedby3-Dtetrahedronmeshestoconsiderthecoreshiftinthenumericalanalysis.Thedesignofexperiments(DOE)wasusedtoestimatethepropermoldingconditionsthatminimizethecoreshiftandadominantparameter.TheresultsoftheDOEshowedthatthedominantparameteristheinjectionpressure,andthecoreshiftde-creaseswhentheinjectionpressuredecreases.Inaddition,itwasshownthattheinitialmoldtemperatureandtheinjectiontimehardlyaffectthecoreshift.Theresultsoftheexperimentsshowedthatproductswithoutwarpagearemanufacturedwhentheinjectionpressureisnearly32MPa.Comparingtheresultsoftheanalyseswiththoseoftheexperiments,optimalinjectionmoldingconditionsweredeter-mined.Inaddition,itwasshownthatthecoreshiftshouldbeconsideredtosimulatetheinjectionmoldingprocessofaplasticbatterycasewiththinanddeepwalls.Keywords:Optimalinjectionmoldingcondition；Coreshift；Thinanddeepwalls；Batterycase；Designofexperiments-1.IntroductionOneoftherecentconcernsintheautomotiveindustryisthereductionofoverallweighttoimprovefuelefficiencyandre-duceavehiclesenvironmentalimpact1-3.Theinjectionmoldingprocessofthinnerplasticcomponentsallowsconsid-erableweightsavingsontheautomotive,asignificantreduc-tioninproductioncost,andashortercycletime1.Variousstudiesareactivelyundertakeninanefforttodeveloptheinjec-tionmoldingprocessofthin-wallplasticparts1.Highinjec-tionpressureandhighinjectionspeedarenecessarytomanu-factureplasticpartswiththinanddeepwalls4.However,ahighinjectionpressurecangiverisetoacoreshift,whichisthespatialdeviationofthepositionofthecoreduringinjectionmolding5.Shepardetal.foundthatthemolddesignandinjectionmoldingconditionsstronglyaffectthecoreshiftofthemold6.Leoetal.reportedthatthedeflectionofweakplatesinthemoldcausesvariationsinthethicknessoftheproductandtheover-packingofmoldings7.Bakharevetal.reportedthatcoreshifteffectsininjectionmoldingcanbepre-dictedthroughmoldfillingsimulationcoupledwithanelasticanalysisoftheflexiblepartsofthemold4.Inthispaper,theinfluenceofinjectionmoldingparametersonthecoreshiftintheinjectionmoldingofaplasticbatterycasewiththinanddeepwallsisexaminedtoestimateanop-timalinjectionmoldingcondition.Thedesignofexperimentsisusedtoestimateapropermoldingconditionminimizingthecoreshift,aswellastodetermineadominantmoldingpa-rameter.Severalexperimentsarecarriedouttoobtainanop-timalinjectionpressure.Comparingtheresultsofnumericalanalyseswiththoseoftheexperiments,theoptimalinjectionmoldingconditionisacquired.Theeffectsofthecoreshiftonthequalityoftheproductarealsodiscussed.2.NumericalanalysisandexperimentsInordertosimulatethecoreshiftandinjectionmoldingcharacteristics,athree-dimensionalinjectionmoldinganalysiswasperformedusingthecommercialcodeMPIV6.1.Fig.1illustratestheanalysismodel.Thedimensionsofthebatterycaseare164.4mm(W)251.4mm(L)184.0mm(H).Themaximumandminimumthicknessesofthewallsare2.7mmand1.8mm,respectively.Thedepthofthewallsis168.7mm.TherunnersystemconsistsofaconicalspruewithaninitialThispaperwaspresentedattheICMDT2009,Jeju,Korea,June2009.ThispaperwasrecommendedforpublicationinrevisedformbyGuestEditorsSung-LimKo,KeiichiWatanuki.*Correspondingauthor.Tel.:+82622307043,Fax.:+82622307234E-mailaddress:smartmail.chosun.ac.krKSME&Springer2010146D.-G.Ahnetal./JournalofMechanicalScienceandTechnology24(2010)145148Table1.ParametersandtheirlevelsfortheDOE.LevelParameters123A:Initialmoldtemperature(oC)35.040.045.0B:injectiontime(seconds)C:Injectionpressure(MPa)40.035.030.0D:Holdingtime(seconds)2.03.24.4(a)(b)Fig.1.Designoftherunnersystemandcore:(a)Designoftherunnersystemandproduct,(b)Meshesfortheflexiblepartsofthecore.(a)(b)Fig.2.Moldforthemanufactureoftheplasticbatterycase:(a)Designofthemold,(b)Manufacturedmold.diameterof8mmandafinaldiameterof15mm,circularrunnerswithdiametersof8mm,andpin-pointgateswithdiametersof2mm.Inordertoconsiderthecoreshiftphe-nomenoninthenumericalanalysis,theflexiblepartsofthemoldandtheflowpartwererepresentedby376,674EAoftetrahedronmeshesand61,098EAofshellmeshes,respec-tively.Theinjectionmaterialwasapolypropyleneresin.ThemeltingtemperatureofPPwassetat230oC.Thedesignofexperiments(DOE)wasusedtoquantita-tivelyexaminetheinfluenceofinjectionmoldingparametersonthecoreshift.Table1showstheinjectionmoldingparame-tersandtheirlevelsfortheL9(34)orthogonalarray.Thesig-nal-to-noise(S/N)ratiowiththe-smaller-the-bettercharacteris-ticswascalculatedtoestimatetheproperconditionformini-mizingthecoreshift.ThecontributionratioofeachparameterisestimatedusingAnalysisofVariance(ANOVA)toobtainthedominantparameteraffectingthecoreshift.Severalexperimentswereperformedusinganinjectionmoldingmachinewith600tonsofclampingforce.Fig.2showsthedesignofthemold,aswellasthemanufacturedTable2.Orthogonalarrayandresultsofthenumericalanalyses.Deformationofflexibleparts(mm)#ofExpABCDF1F2F3F4F5F611170.140.080.23212220.180.090.070.140.080.23313330.130.090.050.110.080.1642170.130.080.18522310.120.090.050.110.080.1662370.140.080.21731320.110.090.050.100.080.1583270.140.080.2393370.150.080.22Fig.3.Resultsofthenumericalanalysis(Coreshift).mold,fortheexperiments.Thedimensionsofthemoldare750mm(W)700mm(L)870mm(H).Thedominantparameterwasvariedwithin10%oftheproposedconditionbytheDOEtodeterminetheoptimalcondition.Inordertoexaminetheinfluenceofcoreshiftonthesimulationoftheinjectionmoldingprocess,theresultsoftheexperimentsarecomparedtothoseofthenumericalanalyses.3.Resultsanddiscussion3.1ResultsofinjectionmoldinganalysisandDOEFig.3andTable2showtheresultsoftheinjectionmoldinganalysis.Theflexiblepartsofthecoreweredeformediniden-ticaldirectionsregardlessofthecombinationofinjectionmoldingconditions,asshowninFig.3.Fig.3andTable2showthatthedeformationsoftheF1,F5,andF6partsofthecoreweregreaterthan0.1mmandflexiblepartsofthecoredeformedsymmetrically.Inaddition,itwasnotedthattheshiftsoftheF2andF3partsarenegligibleincomparisonwiththoseofotherparts.Fromtheresultsoftheinjectionmoldinganalysis,theS/NandcontributionratioswerecalculatedfortheF1,F5,andF6partsofthecorewithrelativelylargecoreshifts.Fig.4andTable3showtheresultsoftheDOE.InFig.4,itcanbeseenthattheS/Nratiosoftheinjectionpressure,theholdingtime,theinjectiontime,andtheinitialmoldtempera-turearemaximizedwhentheirvaluesare30MPa,1.6seconds,4.4seconds,and40oC,respectively.TheS/NratiooftheinD.-G.Ahnetal./JournalofMechanicalScienceandTechnology24(2010)145148147Table3.Purecontributionratiosofeachinjectionmoldingparameterfordifferentflexibleparts.Contributionratio(%)FlexiblePartsInjectiontimeInjectionPressureHoldingtimeF14.156.515.8F40.099.20.2F63.373.36.3Mean2.576.37.4(a)(b)Fig.4.Variationofsignal-to-noiseratiosaccordingtothelevelsofinjectionmoldingparameters:(a)S/Nratiosofinjectionpressureandholdingtime,(b)S/Nratiosofinitialmoldtemperatureandinjectiontime.jectionpressurealsoincreasesremarkablywhentheinjectionpressuredecreases,asshowninFig.4.Table3showsthatthemeanvalueofthecontributionratiooftheinjectionpressureisnearly76.3%andthatthecontri-butionratiooftheinjectionpressureismarkedlyhigherthanthatoftheotherparameters.Fromtheseresults,itwasnotedthatthedominantparameter,whichmainlyaffectsthecoreshift,istheinjectionpressure.ThevariationinS/Nratioandthecontributionratiofortheinjectiontimeandtheinitialmoldtemperaturearenegligible,asshowninFig.4andTable3.Fromtheseresults,itwasnotedthattheinjectiontimeandtheinitialmoldtemperaturehardlyaffectthecoreshift.3.2ResultsoftheexperimentsUsingtheresultsoftheDOE,theexperimentalconditionsofinjectiontime,holdingtime,andinitialmoldtemperatureweresetat1.6seconds,4.4seconds,and40oC,respectively.Theinjectionpressurewasvariedintherangeof27-32MPa.Fig.5showstheresultsoftheinjectionmoldingexperi-Fig.5.Moldedproductfordifferentinjectionpressures:(a)Injectionpressure=30MPa(ConditionproposedbytheDOE),(b)Injectionpressure=32MPa(Optimalcondition).(a)(b)Fig.6.Comparisonoftheresultsofnumericalanalysesandthoseoftheexperiments:(a)Thicknessdistributionofchannels,(b)Post-deformationofthemoldedproduct.ments.Theshortshotdidnotoccurinallexperimentalcondi-tions,asshowninFig.5.However,thewarpagesofthewallsofthemoldedproductoccurredwhentheinjectionpressurewaslowerthan30MPa,asshowninFig.5(a).Thisresultsfromtheinsufficientholdingpressure.Thewarpagesoftheproductwallsdonotoccurwhentheinjectionpressureis32MPa,asshowninFig.5(b).Fromtheseresults,theoptimalinjectionmoldingconditionsofthetestedplasticbatterycasewiththinanddeepwallsweredeterminedasaninjectionpres-sureof32MPa,aninjectiontimeof1.6seconds,aholding148D.-G.Ahnetal./JournalofMechanicalScienceandTechnology24(2010)145148timeof4.4seconds,andaninitialmoldtemperatureof40oC.Theresultsoftheexperimentswerecomparedtothoseofthenumericalanalyses,asshowninFig.6.Fig.6(a)showsthatthedifferenceinwallthicknessbetweentheanalysesandtheexperimentsisreducedfrom-0.15mm0.18mmto-0.09mm0.07mmwhenthecoreshiftisconsideredinthenu-mericalanalysis.Inaddition,itwasnotedthatthethicknessvariationofthewallsisproperlypredictedbythecoreshiftanalysis.Fig.6(b)showsthatthecoreshiftappreciablyaffectsthepost-deformationpatternofthemoldedproduct,andthenumericalanalysisaccountingforthecoreshiftcanpredictthepost-deformationoftheproductwithin0.15mmofcomputa-tionalaccuracy.Theresultsofthenumericalanalysesshowthattheinjectionpressureisreducedfrom50.2MPato32.0MPawhenthecoreshiftisconsidered.Thisisduetoin-creasedcavityvolumeinducedbytheelasticdeformationofthecore.Basedontheaboveresults,itisnotedthatinjectionmoldinganalysisaccountingforthecoreshiftcanproperlysimulatetheinjectionmoldingprocessofthebatterycasewiththinanddeepwalls.4.ConclusionsTheinfluenceofinjectionmoldingparametersonthecoreshiftinthemoldingofaplasticbatterycasewiththinanddeepwallswasinvestigatedusingnumericalanalysisandtheex-periment.Theelasticdeformationofthecorewasconsideredtoreflectcoreshifteffectsonthenumericalanalysis.Throughnu-mericalanalysisandDOE,itwasshownthattheinjectionpressureisthedominantprocessparameteraffectingcoreshiftandthatthecoreshiftdecreaseswhentheinjectionpressuredecreases.Inaddi-tion,itwasnotedthattheinjectiontime,theholdingtime,andtheinitialmoldtemperaturehardlyaffectcoreshift.Itwasdemon-stratedthroughexperimentsthatthemoldedproductwithoutwar-pagescanbemanufacturedataninjectionpressureofapproxi-mately32MPa.Comparingtheresultsofthenumericalanalyseswiththoseoftheexperiments,theoptimalinjectionmoldingcon-ditionswereobtained.Inaddition,itwasshownthatthecoreshiftshouldbeconsideredtoaccuratelysimulatetheinjectionmoldingprocessofaplasticbatterycasewiththinanddeepwalls.References1R.Spina,Injectionmouldingofautomotivecomponent:comparisonbetweenhotrunnersystemsforacasestudy,J.Mater.Process.Technol.155-156(2004)1497-1504.2K.J.Kim,C.W.Sung,Y.N.Baik,Y.H.Lee,D.S.Bae,K.H.KimandS.T.Won,Hydroformingsimulationofhigh-strengthsteel