焊后矯形數(shù)值模擬及敏感性分析

AbstractItisnoticeablyeffectivewiththedigitalstraighteningtechnologyratherthanthemanualindicatedstraighteningtechnology.Thedigitalstraighteningtechnologynotonlyshortensthemanufacturingcycle,butalsoimprovesthemanufacturingqualityandprecision.Taking6005and6008aluminumalloyastheresearchsamplesinthispaper,incombinationwithstraighteningtestandsimulation,wedevelopedthenumericalsimulationmethodsforflamestraighteningandmechanicalstraighteningofthealuminumalloy.Weanalyzedtheinfluenceofstraighteningparameters(straighteningtemperature,heatingwidth,straighteningspeedandcoolingmethod)ondeformationinthispaper.Withtheincreaseofstraighteningtemperature,deformationincreasedgradually.Whenthetemperaturewasbelow250℃,onlyasmalldeformationoccurred.Whenthetemperaturewasover250℃,thedeformationincreasedrapidly.Westudiedtherelationshipbetweenthedeformationandthestrokeofpressingonthealuminumalloyinthispaper.Withtheincrementofpressure,thedeformationincreasedlinearly.Keywords:Aluminumalloy;Straightening;Numericalsimulation;SensitivityanalysisIntroductionStraighteningisoftenneededtoreducethedeformationafterweldingbecausetheweldingstressanddeformationproblemsofthethinaluminumalloyweldedcomponent,suchasunstablestructuresize,weldingexcessivepenetration,misalignmentandotherproblems,affectthestructureprecisionandthesubsequentassemblyprecision[1].Therearemanyreportsonthestraighteningprocesspredictionforaweldingstructure.Forexample,SongH[2],KHLo[3]andQSan-Huietal[4]describedthestraighteningprocessofthevitalstructureinrailwayfield,RongbaiFuetal[5]analyzedthecontrolandcorrectionofweldingdeformation,BABehrencesetal[6]developedananalytical3Dsimulationmodeltofindasuitableadjustmentofthelevelertoreachaflatsheetmetal.Also,theanalyticmodelhadbeenstudiedtopredicttheresidualstressesanddeformationformechanicalstraighteningandheatingstraightening[7-10].Itcannotmeettheweldingdesignandassemblyrequirementsofaluminumalloyparts,althoughtherigidfixationmethodisadoptedtocontroltheweldingdeformation.Sincethereisstilllargeweldingdeformationafterwelding,straighteningafterweldingisstillneeded[11-13].Straighteningmainlyincludescoldstraighteningandhotstraightening.Themechanicalstraighteningisakindofcoldstraightening.Hotstraighteningbasedonthedifferentheatsources,includesflameheating,inductionheatingandlaserheating.Inactualproduction,flameheatingstraighteningandmechanicalstraighteningaremainlyadopted[14-16].Inordertoestablishthenumericalsimulationmethodforflamestraighteningafterwelding,wetooka6005aluminumalloyplatewithathicknessof3mmastestingsample,didaflamestraighteningexperiment,andthedeformationwasmeasuredbythreecoordinatesmeasuringinstrument.Thisexperimentwasusedasabenchmarkof6005aluminumalloyflamestraighteningtovalidatethenumericalsimulationmethodofthe6005aluminumalloyflamestraightening,andtoexploretheinfluenceofparametersonthedeformation.Inordertoestablishthenumericalsimulationmethodofthemechanicalstraightening,wetookthethreepointsbendingtestasabenchmarktest,loadeddifferentpressdisplacementson6082aluminumalloyplates;measuredresidualdeformationaftertheexternalloadingwasreleased.Weestablishedtherelationshipamongspringback,pressurevalueandtheresidualdeformation,basedonthree-pointbendingtest.2.Experimentsandsimulationmethods2.1火焰矯形實(shí)驗(yàn)Figure1(a)showedtheflamestraighteningplates.Thesizeofthetestplatewas300mm×150mm×3mm,thelengthofthestraighteningflameheatingwas200mm.Duetotheflameheating,thermocouplecouldnotbespotweldedintheheatedsurface.Inordertomeasurethetemperatureoftheflamestraighteningprocess,thermocouplewasplacedonthebackoftheheating.Threethermocoupleswerespotwelded,asshowninthefigure.Thefirstthermocouplewasspotweldedinthecenteroftheheatingarea;theremainingtwothermocoupleswereplaced5mmand10mmawayfromheatingcenter,respectively.Figure1(b)wastheflamestraighteningfigure.Waterwasusedforcooling.Itcouldbeseenfromthediagram,thecleararchinheatingposition,wasshapedlikeasaddle.Afterflamestraightening,wemeasuredthedeformationusingthethreecoordinate’sfacility,andcalculateddeformationdistributionusingMATLABsoftware.圖1三點(diǎn)彎曲實(shí)驗(yàn)Thesizeofthethreepointbendingspecimenwas200mm×50mm×4mm,andthespanwas180mm.Thestrokeofpressingwassetto8mm,10mm,12mm,14mm,16mmand18mm,attherateof1mm/min,respectively.Thepressureheaddiameterwas20mm.Figure2（a）showedthethreepointbendingtestsite.Figure2(b)showedtherelationshipbetweenthepressureandstrokeat18mmofpressing.Itcouldbeseenfromthediagram,inthecaseoflesspressure,pressvalueincreasedwiththeincreaseofstrokeofpressing,characterizedbytheelasticbendingdeformation.Inthepressureof0.8KN,plasticdeformationbegantooccur.圖2有限元模型Figure3(a)showedtheFEAmodel.Duetothetemperaturegradientwashigherandthegridsattheflamestraighteningareaweredense,Awayfromtheheatingzone,thegridsizeincreasedgradually.Hexahedronunitsofthethree-dimensionalmodelwere15205,nodeswere19482.Figure3(b)showedtherelationshipbetweenthepressureandstrokeat18mmofpressing.Itcouldbeseenfromthediagram,inthecaseoflesspressure,pressvalueincreasedwiththeincreaseofstrokeofpressing,characterizedbytheelasticbendingdeformation.Inthepressureof0.8KN,plasticdeformationbegantooccur.Thesizeofthreepointbendingspecimenwas200mmx50mmx4mm.Thespanwas180mm.Thestrokeofpressingwas8mm,10mm,12mm,14mm,16mmand18mm,respectively,attherateof1mm/min.Thediameterofthepressheadis20mm.Figure3(c)wasthegeometricmodelofnumericalsimulationofthreepointbending.Thenumberof3Dunitsinthefigurewas2592;thenumberofnodeswas3597.Thegeometricmodelconsistedofthetabletandthepressurehead.Figure3(d)showedtheimposedconstraintforthenumericalsimulation.Weselectedallnodesonbothendsofthespan.ThedirectionofconstraintZdirectionwasshowninredarrows.ThedirectionsofconstraintsXandYwereshowningreenarrowandredarrow.Weselectedtheheadofsurfacedowntomechanicalstraighteningdisplacement.Theloadingrateofdisplacementwasconsistentwiththetestresults,andtheratewas1mm/min.ThePressingwasaddedtoacertainvalue,andthenreleaseddisplacement,sotheelasticdeformationoftabletwasrestored,namelyforthespringbackamount.Theplasticdeformationcouldnotberestored,whichwascalledthedeformation.圖33.Resultsanddiscussion3.1火焰矯形實(shí)驗(yàn)與模擬Figure4showedthestraighteningprocesstemperaturecurve.Theheatingtemperatureofthecenterwasthehighesttemperature,about369℃.Thetemperatureofthepositionof5mmawayfromtheheatingcenterpositionwasabout352℃,andthistemperaturewasclosetothecentertemperature.Andthehighesttemperaturewas93℃inthelocation10mmawayfromtheheatingcenter.Itcouldbeseenfromthetemperaturedistributionthat,theheatingradiusoftheflamestraighteningwasabout5mm.圖4Inflamestraighteningexperimentalsimulation,theknownconditionsweretheflamelength,gunmovementspeed,thedistancebetweencoolingwaterandheatinggun,butthewidthoftheactualflameheatingcouldnotbemeasured.Therefore,inthenumericalsimulation,thetemperaturecurvewasthesame,andwechosedifferentheatingwidths(8mm,10mm,12mm,14mmand20mm).And,weusedthemeasuringtemperaturecurveanddeformationintheprocessofvalue,todeterminetheactualheatingwidth.Figure5showedistemperaturesimulationcomparedwiththemeasuredvaluesunderdifferentheatingwidths.圖5Figure5TemperaturesimulationcomparedwiththemeasuredvaluesunderdifferentheatingwidthsItcouldbeseenfromthediagram,intheheatingstage,thetemperatureroserapidly,andthenslowlycooleddown.Thiswasmainlyduetoacertaindistancebetweenthenozzleandtheheatinggun.Whenwestartedtousewater,thesimulationvaluesandthemeasuredvalueswereconsistentandprovedthatwatercoefficientofthesimulationwasaccurate.Inthecenteroftheheating,theeffectofheatingwidthwasnotobviousonthepeaktemperature.And5mmawayfromthecenteroftheheating,theinfluenceofheatingwidthwasobviousonthetemperature.Thegreaterthewidthheated,thehigherthepeaktemperature.Whentheheatedwidthwas14mm,thesimulatedtemperaturecurvewasthemostconsistentwiththemeasuredvalues.圖63.2三點(diǎn)彎曲實(shí)驗(yàn)與模擬Figure7(a)showedthesamplesafterthethreepointbendingtest.Itcouldbeseenfromthediagram,thegreaterofthestrokeofpressing,themoreobviousofdeformationafterunloading.Figure7(b)showedtherelationshipforthedeformation,springbackandstrokeofpressing.Itcouldbeseenfromthegraph,withtheincrementofpress,thedeformationandthespringbackwerelinearlyincreased.Theincreasingrateofthespringbackwaslessthantheincreasingrateofthedeformation.Themainreasonisthat,foraluminumalloy,theplasticdeformationareawasbiggerwiththeincrementofpress.圖7Figure8showedacomparisonofthecalculateddeformationvalueandthemeasuredvalues.Itcouldbeseenfromthediagram,thesimulationvaluewasconsistentwiththeactuallymeasuredvalue.Themaximumerroris14%,andtheaverageerrorisabout4%.圖83.3焊后矯形參數(shù)化研究Weadoptedthemethodofnumericalsimulationanalysiswiththeconsiderationofflamestraighteningtemperature,heatingwidth,straighteningspeedandcoolingmedium(water,aircooled)effectonthedeformation.Figure9(a)showedtherelationshipbetweendeformationandtemperatureofflamestraightening.Straighteningspeed,coolingmediumandflameheatingwidthremainthesame.Flameheatingwidthandspeedofstraighteningweresetto14mmand25mm/s,respectively.Coolingmediumwaswater.Itcouldbeseenfromthegraph,withtheincreaseoftemperature,deformationincreasedgradually.Whenthetemperaturewasbelow250℃,theorthopedicafterdeformationwassmall.Whenthetemperaturewasmorethan250℃,thedeformationincreasedrapidly,mainlyduetoobvioussofteningofthe6005aluminumalloy.Thedecreasewascausedbytheresistancetodeformationability.Figure9(b)showedtherelationshipbetweenthedeformationandtheheatingwidth.Theheatingspeedofflameheatingtemperature,straighteningspeedandcoolingmediumremainedthesame.Flameheatingtemperatureandspeedofstraighteningweresetto250℃and25mm/s,respectively.Itcouldbeseenfromthegraph,withtheincreaseofheatingwidth,thedeformationincreasedgradually.Whentheheatedwidthislessthan12mm,thedeformationincreasewasrelativelyslow.Whentheheatedwidthwasmorethan12mm,thedeformationincreasedrapidly.Figure9(c)showedtherelationshipbetweenthedeformationandstraighteningspeed.Inthecalculation,heatingtemperature,heatingwidthandcoolingmediumremainedthesame.Flameheatingtemperatureandheatingwidthwassetto250℃and14mm,respectively.Itcouldbeseenfromthediagram,whenthestraighteningvelocitywassmall,thedeformationincreasedslightly,andthechangewasnotobvious.Whenstraighteningspeedwasmorethan20mm/s,thedeformationdecreasedwiththeincreaseofthestraighteningspeed.圖9Table1showedthedeformationcomparisonofthedifferentcoolingmediums.Theheatingtemperature,heatingwidthandstraighteningspeedremainedthesame.TheFlameheatingtemperature,heatingwidthandstraighteningspeedweresetto250℃,14mmand25mm/s,respectively.Itcouldbeseenfromthedatainthetable,theinfluenceofwatercoolingandaircoolingondeformationwasnotobvious.Thedeformationundertheaircoolingwasslightlybelowthedeformationofthewatercooling.Table1DeformationcomparisonofdifferentcoolingmediumsCoolingmediumDeformation，mmUXUYUZWatercooling-0.060-0.230.76Aircooling-0.0560.220.74Basedontheabovemechanicalstraighteningnumericalsimulationmethod,wesimulatedthreepointbendingtestfor6005aluminumalloy,establishedtherelationshipbetweendeformationandthestrokeofpressingof6005aluminumalloy.Itwascomparedwith6082aluminumalloy,asshowninfigure10.Itcouldbeseenfromthediagram,thechangetrendof6005aluminumalloydeformationandstrokeofpressingwasconsistentwiththatof6082.Asthestrokeofpressingincreases,bearingthesamequantityofcases,thedeformationof6005aluminumalloywashigherthanthatof6082aluminumalloy.Thereasonwasthattheelasticmodulusofthetwoaluminumalloyswasclose.However,theyieldstrengthof6082aluminumalloywasobviouslyhigherthanthatof6005aluminumalloy,whichmakestheelasticstrainof6005aluminumalloyislowerthanthatof6082.Moreplasticdeformationoccurredin6005,sotheresidualdeformationquantitywasbig.圖104ConclusionAnumericalsimulationmethodoftheflamestraighteningwasestablishedtoinvestigatetheinfluenceofflamestraighteningparametersonthedeformation.Withtheincreaseofstraighteningtemperature,deformationincreasedgradually.Whenthetemperaturewasbelow250℃,thedeformationchangewassmall.Whenthetemperaturewasover250℃,thedeformationincreasedrapidly.Withtheincreaseofheatingwidth,thedeformationincreasedgradually.Whentheheatedwidthwaslessthan12mm,thedeformationincreasewasrelativelyslow.Whentheheatedwidthwasmorethan12mm,thedeformationincreasedrapidly.Withtheincreaseofthestraighteningspeed,deformationdecreased;coolinghadnoobviouseffectondeformation.Anumericalsimulationmethodofthemechanicalstraighteningwasalsoestablished.Therelationshipbetweenthedeformationandstrokeofpressingfor6005aluminumalloyand6082aluminumalloywasstudied.Withtheincrementofpress,thedeformationincreasedlinearly.

References:HongyuanFang.Weldingstructural[M].Beijing:mechanicalindustrypress,2008:109-128.(InChinese).SongH,FuLH,PLWang.Theinfluenceofresidualstressforheavyrailhorizontalstraighteningregulationsafterstraightening[J]．MachineryDesignandManufacture,2010,11:100-102.(InChinese).KHLo,PMummery,DJButtle.Characterisationofresidualprincipalstressesandtheirimplicationsonfailureofrailwayrails[J].EngineeringFailureAnalysis,2010,17(6):1273-1284.QSan-Hui,LIXue-Mei,QGuang-Hui.DevelopmentandTechnicalInnovationofChina'sHigh-speedRailway[J].HighSpeedRailWayTechnology,2014,5(1):1-7.(InChinese).RongbaiFu.Thecontrolandcorrectionofweldingdeformation[M].Beijing:mechanicalindustrypress,2006:226-252.(InChinese).BABehrens,TENadi,RKrimm.Developmentofananalytical3D-simulationmodelofthelevellingprocess[J].JournalofMaterialsProcessingTechnology,2011,

211(6):1060-1068.MATSUOA,FUJITAM.Variousproblemsinlevelingtheplatesandsheetswithrollerleveling[J].Iron&SteelInstituteofJapanJournal,2006,19(2):327?330.(InJapanese).CUILi,HUXian-lei,GUO