资源描述：
《Back Propagation neural network modeling for warpage prediction 》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、MaterialsandDesign32(2011)1844–1850ContentslistsavailableatScienceDirectMaterialsandDesignjournalhomepage:www.elsevier.com/locate/matdesBackPropagationneuralnetworkmodelingforwarpagepredictionandoptimizationofplasticproductsduringinjectionmoldingaa,⇑baaFeiYin,HuajieMao,LinHua,WeiG
2、uo,MaoshengShuaSchoolofMaterialsScienceandEngineering,WuhanUniversityofTechnology,Wuhan430070,ChinabSchoolofAutomobileEngineering,WuhanUniversityofTechnology,Wuhan430070,ChinaarticleinfoabstractArticlehistory:WarpageofplasticproductsisanimportantevaluationindexforPlasticInjectionM
3、olding(PIM).ABackReceived13September2010Propagation(BP)neural-networkmodelforwarpagepredictionandoptimizationofinjectedplasticpartsAccepted8December2010hasbeendevelopedbasedonkeyprocessvariablesincludingmoldtemperature,melttemperature,pack-Availableonline17December2010ingpressure,
4、packingtimeandcoolingtimeduringPIM.TheapproachusesaBPneuralnetworktrainedbytheinputandoutputdataobtainedfromtheFiniteElement(FE)simulationswhichareperformedonKeywords:Moldflowsoftwareplatform.Inaddition,akindofautomobileglovecompartmentcapwasutilizedinthisA.Polymersstudy.Trainedbyt
5、heresultsofFEsimulationsconductedbyorthogonalexperimentaldesignmethod,theC.MoldingpredictionsystemgotamathematicalequationmappingtherelationshipbetweentheprocessparameterF.Defectsvaluesandwarpagevalueoftheplastic.Ithasbeenprovedthatthepredictionsystemhastheabilitytopredictthewarpa
6、geoftheplasticwithinanerrorrangeof2%.Processparametershavebeenoptimizedwiththehelpofthepredictionsystem.Meanwhileenergyconsumptionandproductioncyclewerealsotakenintoconsideration.Theoptimizedwarpagevalueis1.58mm,whichisshortenedby32.99%com-paringtotheinitialwarpageresult2.358mm.An
7、dthecoolingtimehasbeendecreasedfrom20sto10s,whichwillgreatlyshortentheproductioncycle.Thefinalproductcansatisfywiththematchingrequire-mentsandfittheautomobileglovecompartmentwell.Ó2010ElsevierLtd.Allrightsreserved.1.Introductionpressure,packingtimeandcoolingtimeasthekeyprocessparam-
8、etersduringPIM.Andtheygottheoptim
