焊接外文文献

ProbinginfluenceofweldingcurrentonweldqualityintwowiretandemsubmergedarcweldingofHSLAsteelD.V.Kiran1,B.Basu2,A.K.Shah2,S.Mishra3andA.De*1Twowiretandemsubmergedarcwelding(SAW-T)involvesapplicationoftwoelectrodewirestoformasingleweldpoolwithonewireleadingtheotheralongtheweldinterface.RealisationoftheinfluencesoftheleadandtrailwirecurrentsontheweldbeadmorphologyandtheweldstrengthareimportantinSAW-T.TheauthorsherepresentanexperimentalandcomputationalinvestigationonSAW-Tprocess.Theconceptofcentralcompositerotatabledesignisfollowedfordesigningtheexperimentaltrials.Thefinalweldwidthandthereinforcementheightshowedgreatersensitivitytotrailcurrentwhilethepenetrationwasinfluencedprimarilybyleadcurrent.Athree-dimensionalheattransferanalysisiscarriedoutusingfiniteelementmethod.Thecomputedweldprofileswerevalidatedwiththecorrespondingmeasuredresults.Thecomputedcoolingratesshowedadeclinewithhigherweldingcurrentand,thecorrespondingsampleweldsshowedreducedacicularferritepercentageandlowerweldstrength.Keywords:Tandemsubmergedarcwelding,Experimentalstudy,Weldgeometry,Mechanicalproperties,NumericalmodellingIntroductionTwowiretandemsubmergedarcwelding(SAW-T)iscapableofgreaterdepositionrateathigherweldspeedincomparisontotheconventionalsinglewireSAWprocess.1–2InSAW-T,twoelectrodewiresareindepen-dentlycontrolledwithusuallytheleadwireconnectedtoDCEPandthetrailwiretopulsedAC(withsimulta-neouscontrolofboththepositiveandthenegativepulses)toavoidarcblow.2–5Owingtothesimultaneousdepositionsfromtheleadandthetrailelectrodewiresintoasingleweldpool,SAW-Tcanprovidegreateramountofjointfillingatafasterrateandwithareducedoverallheatinput.2However,theoptimisationofprocessparametersinSAW-Tischallengingduetothepresenceofalargenumberofindependentprocessvariables.Theseincludeweldingspeed(v),leadwirecurrent(IL),timedurations(tzT,t{T)andthepositiveandnegativepeakcurrentpulses(IzT,I{T)intrailwire,interwiredistance(D),andelectrodeextensions(EL,ET)forboththewires.Althoughparametricstudiesarereportedforconven-tionalSAWprocess,6–10similarstudiesforSAW-Tarerare.11,12Inrecenttimes,empiricalmethodsbasedonfuzzyrule,13greyrelationalanalysis,14artificialneuralnetwork13,15andneurowaveletpacketapproach15,16arealsousedaimingatorganisingparametricstudiesingastungstenarc,13LaserGasmetalarc(GMA)hybrid14andGMA15,16weldingprocesses.Althougheachofthesestudieshastoinherentlyincludealargevolumeofexperiments,ofinterest,however,isthefactthathardlyanyconclusionismadetowardssolitaryinfluencesofsignificantweldingparameters,e.g.weldingcurrentonweldjointattributes.Inaway,thesestudieswereseeminglydirectedtowardsprobingthefeasibilityofthoseempiricalroutesinpredictingweldjointattributesbasedonexperimentalstudiesalonewithalargenumberofinputvariables.Incontrasttosuchempiricalroutes,numericalprocessmodelsarefoundmoresuitableforfundamentalunderstandingofheattransferandresul-tanttemperaturefieldinfusionwelding.17AlthoughsuchmodelsarereportedforconventionalsinglewireGMA18–22andSAW23–27processes,noefforttonumericallymodeltwowireSAW-Tisencounteredinopenliterature.SinceSAW-Tpromisesanenhanceddepositionrateandhence,higherproductivity,itiscertainlyimportanttoimprovethequantitativeknowl-edgethroughbothexperimentalandmodellingmeans.HeretheauthorspresentanefforttounderstandprimarilytheinfluenceofweldingcurrentontheweldjointattributesinatwowireSAW-Tprocess.Theoverallmotivationistocarryoutacontrolledsetofexperimentstorealisethesensitivityofweldbeaddimensionsandjointstrengthontheleadandtrailwirecurrentsandtodevelopareliableprocesssimulationmodelutilisingthemeasuredresultsandthecorrespond-ingweldingconditionsasreferences.Itisconceivedthataprocesssimulationmodel,basedonscientificprinci-ples,willnotremainspecifictomaterialandprocess1MechanicalEngineeringDepartment,IITBombay,Mumbai,Maharashtra400076,India2NavalMaterialsResearchLaboratory,Thane,India3DepartmentofMetallurgyandMaterialsScience,IITBombay,Mumbai,Maharashtra400076,India*Correspondingauthor,[e-mail]
2010InstituteofMaterials,MineralsandMiningPublishedbyManeyonbehalfoftheInstituteReceived17August2009;accepted28August2009DOI10.1179/136217109X12518083193432ScienceandTechnologyofWeldingandJoining2010VOL15NO2111aliketheempiricalmethodsandbeabletoreducethecostlyandtimeconsumingtrialanderrorexperimentsintheprocessofdesignforwelding.17Thus,acontrolledsetofbeadongrooveweldsamplesaremadeinaspecificgradeofHSLAsteelplate,whichisprimarilyusedinshipstructureshavinganeverdemandofgreaterrateofdepositionduringfabrication.Theexperimentalconditionsaredesignedbasedoncentralcompositerotatabledesignapproach,28whichfacilitatesareduc-tioninthetotalnumberofrequiredexperimentsespeciallywithalargenumberofindependentprocessvariables.Theinfluencesoftheleadandthetrailwirecurrentsonweldbeadmorphologyandjointstrengtharestudiedextensively.Athree-dimensional(3D)transientheattransferanalysisofthetwowiretandemsubmergedarcweldingprocessisperformedusingthecommercialfiniteelementmethodsoftwareSYSWELD.Thecomputedwelddimensionsarevalidatedwithcorrespondingexperimentalresults.Theinfluen