原奥氏体晶界腐蚀

1RevealingPrior-AusteniteGrainBoundariesinHeatTreatedSteelsByGeorgeF.VanderVoortThemechanicalpropertiesofheat-treatedalloysteelsarestronglyinfluencedbythegrainsizeoftheparentaustenitephasebeforequenching.Butrevealingtheprior-austenitegrainboundaries(PgGBsorPAGBs)canbequitedifficultdependinguponthealloyanditsmicrostructure.Thereareanumberofwell-established(seeASTME112,forexample)proceduresthatareusedtodecoratethePgGBsduringaheat-treatmentcycle,e.g.,theMcQuaid-Ehncarburizingtest(Fig.1)andtheoxidationtest.Insomemedium-carbonsteels,ataspecificcoolingrate,proeutectoidferritewillprecipitateatthePgGBs,whileinhigh-carbonsteels(generallyhypereutectoidtoolsteels),proeutectoidcementitewillprecipitateonthePgGBsuponslowcoolingfromelevatedtemperatures.Theseconditionsareoftenseeninas-castoras-rolledsteels,asshowninFigures2and3.Butthesemethodscannotbeappliedtodeterminetheprior-austenitegrainsizeofasteelpartorcomponentthathasalreadybeenheattreated,asthesemethodswillproduceadifferentgrainsize.Forthisproblem–andthisisacommonsituationinfailureanalysis–onecanonlyuseanetchingtechniquetorevealthePgGBs.2HistoryofPrior-AusteniteGrainBoundaryEtchDevelopmentOneoftheearliestetchantstohavesomesuccessatrevealingPgGBsinsomesteelswasVilella’sreagent,publishedin1938.[1]Thisetchhashadlimitedsuccess,mainlywithtoolsteels.Subsequently,SchradermodifiedVilella’sreagent,butitalsohaslimitedvalue.In1949,MillerandDay[2]publisheda5%aqueousferricchloridereagentforlow-carbonmartensiticsteels.AqueousferricchlorideandHClsolutionshavealsobeensuggested.Nital,generallyinconcentrationsof2-10%(donotstoremorethan3%HNO3inethanolinatightlyclosedbottle,asitcanexplode),willrevealthegrainboundariesinonlyafewsteels–highlyalloyedtoolsteelsintheas-quenchedorlightlytemperedcondition,suchasD2andhigh-speedsteels.ThefirstreasonablysuccessfuletchantforPgGBswaspublishedin1955byBechetand3Beaujard[3]usingasaturatedaqueouspicric-acidsolution(ashadbeenusedinstudiesoftemperembrittlement)containing0.5%ofawettingagent,“Teepol”(sodiumalkylsulfonate),atroomtemperature.Thisetchanthasbeenthefoundationofmanysubsequentmodificationstoimproveitseffectiveness.Thewritertriedthisetch[4]onspecimensof8620,4140and5160intheas-quenchedconditionandaftertemperingat400,800and1200°Fusingsodiumtridecylbenzenesulfonateasthewettingagent.Itdidnotrevealgrainboundariesonanyofthe8620specimens.ItdidrevealthePgGBsonas-quenchedandtempered(400and800°F)specimensof4140and5160butdidnotrevealthemonanyspecimenstemperedat1200°F.Temperedmartensiteandtemperedbainitesbothrespondtothisetchbutonlyformediumtohigh-carbonsteelsandonlywhentemperedbelow~1050°F.Itiswellknownthatsaturatedaqueouspicricacidwithawettingagent(usedatroomtemperature)revealsPgGBsifphosphorusispresentinthegrainboundaries,andthisiseasierifthespecimenhasbeenheatedinthetemperembrittlementrange.[5]SegregationofSnorSbtothePgGBs,whichalsocausestemperembrittlement,doesnothelprevealthePgGBsusingthisetchinsteelsfreeofphosphorus.[6,7]PreeceandCarter[8]showedusingTEMthattherewasacleardifferenceinappearancebetweengrainboundariesthatweretemperembrittledduetoahighlocalphosphorousconcentrationcomparedtoanon-embrittledspecimenwithalowerlocalphosphorouscontent,eventhoughsaturatedaqueouspicricacidrevealsthePgGBsinbothcasesandtheboundarieslookedsimilarbylightmicroscopy.Studiesconductedinthistimeperiodexaminedtheeffectofavarietyofwettingagentsontheetchresponse.Nelson[9]conductedthemostextensivecomparisonusingfivewettingagents,includingthemostpopularone,sodiumtridecylbenzesulfonate,withseveraldifferent4etchantsincludingthesaturatedaqueouspicric-acidsolution.Withoutthewettingagentadded,saturatedaqueouspicricacidwasanexcellentgeneral-purposeetchantforsteels,butPgGBswerenotrevealed.Whenthiswettingagentwasadded,general-structureetchingwassuppressedandPgGBswererevealed.Noneoftheotherwettingagentstriedwereaseffective.Anumberofstudiesontheuseofwettingagentsinetchantshavebeenreviewed.[10]Theoriginaltridecylversionofthiswettingagenthasbranchedmolecularchains,whicharedifficulttomanufactureandhavepoorbiodegradability.Morerecentversionshavelinearchainsandarebiodegradable.Kilpatrick[14]evaluatedthedodecylversionofthiswettingagent,whichismoreeasilymade,readilybiodegradableandworksaswell.Consequently,thiswettingagentisthemostcommonlyusedtodayforrevealingPgGBs.Barraclough[11]reviewedetchantstriedby10differentauthorstorevealPgGBs.Heconcludedthatitwasnecessarytotemperembrittlespecimenstoobtainadequategrain-boundarydelineationtopermitmeasurementstobemadeofthegrainsize.HisworkconfirmedthatpicricacidwasthemostsuitableagentforrevealingPgGBsandthesolventusedwascritical.Alcoholsdidnotwork,butwaterorethergavegoodresults.Petroleumetherislessdangerousthanethylether,butbothareexplosivewhenheatedabove100°C,andstaticelectricitycancauseexplosions.Severalwettingagentsweretried,allweresuitable,buthepreferred“Teepol”(TeepolisaregisteredtrademarkoftheShellChemicalCo.ofHouston,Texas).Hefoundthattheaqueoussolutioncouldbeusedattemperaturesupto85°C,buthedidnotindicateifhighertemperaturesproducedanybenefitordetri