《Welded.Design.Theory.and.Practice，.Hick，.Wood.pdf》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。
Weldeddesign±theoryandpracticeJohnHicksCambridgeEngland PublishedbyAbingtonPublishingWoodheadPublishingLimited,AbingtonHall,Abington,CambridgeCB16AH,Englandwww.woodhead-publishing.comFirstpublished2000,AbingtonPublishing#WoodheadPublishingLtd,2000TheauthorhasassertedhismoralrightsAllrightsreserved.Nopartofthispublicationmaybereproducedortransmittedinanyformorbyanymeans,electronicormechanical,includingphotocopying,recording,oranyinformationstorageandretrievalsystem,withoutpermissioninwritingfromthepublisher.Whileagreatdealofcarehasbeentakentoprovideaccurateandcurrentinformationneithertheauthornorthepublisher,noranyoneelseassociatedwiththispublicationshallbeliableforanyloss,damageorliabilitydirectlyorindirectlycausedorallegedtobecausedbythisbook.BritishLibraryCataloguinginPublicationDataAcataloguerecordforthisbookisavailablefromtheBritishLibrary.ISBN1855735377CoverdesignbyTheColourStudioTypesetbyBookEnsLtd,Royston,HertsPrintedbyTJInternational,Cornwall,England ContentsPrefaceixIntroductionxii1Theengineer11.1Responsibilityoftheengineer11.2Achievementsoftheengineer31.3Theroleofwelding71.4Othermaterials91.5Theweldingengineeraspartoftheteam102Metals112.1Steels112.2Aluminiumalloys203Fabricationprocesses223.1Origins223.2Basicfeaturesofthecommonlyusedweldingprocesses253.3Cutting323.4Bending323.5Residualstressesanddistortion333.6Postweldheattreatment354Considerationsindesigningaweldedjoint364.1Jointsandwelds364.2Terminology394.3Weldpreparations424.4Dimensionaltolerances504.5Access52 viContents5Staticstrength545.1Buttwelds545.2Filletwelds556Fatiguecracking596.1Themechanism596.2Weldedjoints626.3Residualstresses676.4Thicknesseffect676.5Environmentaleffects686.6Calculatingthefatiguelifeofaweldeddetail687Brittlefracture757.1Conventionalapproachestodesignagainstbrittlefracture757.2Fracturetoughnesstestingandspecification777.3Fracturemechanicsandothertests798Structuraldesign828.1Structuralforms828.2Designphilosophies908.3Limitstatedesign959Offshorestructures969.1Theneedsofdeepwaterstructures969.2TheNorthSeaenvironment989.3Theresearch1019.4Platformdesignandconstruction1049.5Serviceexperience10510Managementsystems10610.1Basicrequirements10610.2Contractsandspecifications10610.3Formalmanagementsystems10810.4Weldedfabrication10911Weldquality11111.1Welddefects11111.2Qualitycontrol11911.3Weldedrepairs126 Contentsvii11.4Engineeringcriticalassessment12712Standards13112.1Whatwemeanbystandards13112.2Standardspecifications131References135Bibliography138Index139 PrefaceIhavewrittenthisbookforengineersofalldisciplines,andthisincludesthoseweldingengineerswhodonothaveabackgroundinmattersofengineeringdesign,aswellasforothersinallprofessionswhomayfindthissubjectofinterest.Asmightbeexpected,Ihavedrawnheavilyonmyownexperience.NotthatIdiscoveredanynewprinciplesormethodsbutbecauseIhadtheprivilegeoffirstlybeingassociatedwithresearchintothebehaviourofweldedjointsinserviceatitsmostactivetimeinthe1960sand1970sandsecondlywiththeapplicationofthatresearchinarangeofindustriesandparticularlyinstructuraldesignandfabricationwhichaccompaniedtheextensionofoilandgasproductionintodeeperwatersinthe1970s.TheresultsofthosedevelopmentsrapidlyspreadintootherfieldsofstructuralengineeringandIhopethatthisbookwillbeseeninpartasarecordofsomeoftheintenseactivitywhichwentoninthatperiod,whetheritwasinanalysingtestresultsinalaboratory,writingstandards,preparingaconceptualdesignorinstallingamanythousandtonnesubstructureontheoceanfloor.ThepositionfromwhichIwritethisbookisonewhere,afterbeingastructuralengineerforfiveyears,Ibecameaspecialistinweldeddesign.InthisroleIhaveformanyyearsworkedwithcolleagues,clientsandpupilswho,withoutexception,havebeenandareapleasuretoworkwith;theirmasteryoftheirowndisciplinesandtheresponsibilitieswhichtheycarrydwarfsmyownefforts.Ihavealsospent,Ibelieve,sufficientperiodsinotheroccupationsbothinsideandoutsidetheengineeringprofessiontogivemeanexternalperspectiveonmyspecialism.AsaresultIfeltthatitwouldbehelpfultowriteabooksettingoutthesubjectofweldeddesigninthecontextoftheoverallpictureofengineeringwithsomehistoricalback-ground.InpresentingthesubjectinthiswayIhopethatitwillencourageteachingstaffinuniversitiesandcollegestoseeweldedjointsandtheirbehaviourasanintegralpartofengineeringandthattheywillembedthesubjectintheircoursesinsteadoftreatingitasanadd-on.Itwillalsoservepractisingweldingandotherengineerswishingtoextendtheirknowledgeof xPrefacetheopportunitieswhichweldingoffersandtheconstraintsitimposesintheirownwork.Thesubjectofdesignforweldingrestsatanumberofinterfacesbetweenthemajorengineeringdisciplinesaswellasthescientificdisciplinesofphysics,chemistryandmetallurgy.Thispositionontheboundariesbetweentraditionalmainstreamsubjectsmayperhapsbethereasonwhyitreceivesrelativelylittleattentioninuniversityengineeringcoursesatundergraduatelevel.Myrecentdiscussionswithengineeringinstitutionsandacademicsrevealsasituation,bothintheUKandothercountries,inwhichtheappearanceorotherwiseofthesubjectinacurriculumseemstodependonwhetherornotthereisamemberoftheteachingstaffwhohasbothaparticularinterestinthesubjectandcanfindthetimeinthetimetable.Thisisnotanewposition;Ihavebeenteachinginspecialistcoursesondesignforweldingatallacademicandvocationallevelssince1965andlittleseemstohavechanged.MrRPNewman,formerlyDirectorofEducationatThe1WeldingInstitute,writingin1971,quotedareplytoaquestionnairesenttoindustry:Personnelenteringadrawingofficewithoutmuchexperienceofwelding,asmanydotoday(i.e.1971),canreachareasonablyseniorpositionandstillhaveonlya`stop-gap'knowledge,pickeduponageneralbasis.Thisisfundamentallywrongandisthecauseofmanyofourfabrication/designproblems.Therewasthen,andhasbeenintheinterveningyears,noshortageofbooksandtrainingcoursesonthesubjectofweldeddesignbutthematterneverseemstoenterorremaininmanypeople'sminds.InsayingthisIamnotcriticisingtheindividualengineerswhomayhavebeenledtobelievethatweldedjointdesignandmaterialselectionarematterswhichareeithernotpartofthedesigner'sroleor,iftheyare,theyrequirenoeducationinthesubjects.Indeed,suchwasmyownearlyexperienceinadesignofficeandIlookbackwithembarrassmentatmyfirstcalculationofthesuitabilityofweldedjointdesigninanindustryinwhichweldingwasnotcommonlyused.ItwasanexampleofbeingsoignorantthatIdidn'tknowthatIwasignorant.Thatfirstexperienceofaprematurefailurehasstayedwithmeandgivesmehumilitywhenassistingpeoplewhoareinasimilarpositiontoday.`There,butforthegraceofGod,goI'shouldbeonabanneraboveeveryspecialist'sdesk.Thereare,ofcoursemanyengineerswhohave,eitherbecausetheirworkrequireditorbecauseofaspecialinterest,becomecompetentinthesubject.Eitherway,thereisapointatwhichaspecialistinputisrequiredwhichwilldependuponthenature,noveltyandcomplexityofthejobsetagainsttheknowledgeandexperienceoftheengineer.Ihavetriedtoputintothisbookasmuchasisusefulandinformativewithoutincludingavastamountofjustificationanddetail;thatcanbe Prefacexifoundinthereferencedmorespecialistworks.However,Ihavetriedtokeepabalanceinthisbecauseiftoomanymattersarethesubjectofreferencesthereadermaybecomeexasperatedatcontinuallyhavingtoseekotherbooks,someofwhichwillbefoundonlyinspecialistlibraries.ForthemostpartIhaveavoidedreferencestostandardsandcodesofpracticeexceptinahistoricalcontext.Exceptionsarewhereastandardisanexampleofbasicdesigndataorwhereitrepresentsguidanceonanindustrywideagreedapproachtoananalyticalprocess.Ihaveadoptedthispositionbecauseacrosstheworldtherearesomanystandardsandtheyarecontinuallybeingamended.Inadditionstandardsdonotrepresentasourceoffundamentalknowledgealthough,unfortunately,someareoftenseeninthatlight.HoweverIrecognisetheirimportancetothepracticalbusinessofengineeringandIdevoteachaptertothem.Iacknowledgewithpleasurethosewhohavekindlyprovidedmewithspecialistcommentonsomepartsofthebook,namelyDrDavidWidgeryofESABGroup(UK)LtdonweldingprocessesandMrPaulBentleyonmetallurgy.NonethelessItakefullresponsibilityforwhatiswrittenhere.IamindebtedtoMrDonaldDixonCBEfortheillustrationoftheClevelandColossusNorthSeaplatformconceptwhichwasdesignedwhenhewasManagingDirectorofTheClevelandBridgeandEngineeringCoLtd.ForthephotographsofhistoricstructuresIamgratefultotheChambredeCommerceetd'IndustriedeNõÃmes,theIronbridgeGorgeMuseum,andPurcellMillerTrittonandPartners.IalsoampleasedtoacknowledgetheassistanceofTWI,inparticularMrRoySmith,ingivingmeaccesstotheirimmensephotographiccollection.JOHNHICKS IntroductionManyengineeringstudentsandpractisingengineersfindmaterialsandmetallurgycomplicatedsubjectswhich,perhapsamongstothers,arerapidlyforgottenwhenexaminationsarefinished.Thisputsthematadisadvantagewhentheyneedtoknowsomethingofthebehaviourofmaterialsforfurtherprofessionalqualificationsoreventheireverydaywork.Theresultofthispositionisthatengineeringdecisionsatthedesignstagewhichoughttotakeaccountofthepropertiesofamaterialcanbewrong,leadingtofailuresandevencatastrophes.ThisisclearlyillustratedinanextractfromTheDailyTelegraphon4September1999inanarticleofferingbackgroundtothepossiblecauseofafatalaircraftcrash.```Thereisnofaultinthedesignoftheaircraft,''the(manufacturer's)spokesmaninsisted.``Itisafeatureofthematerialwhichhasshownitdoesnottakethewearoveranumberofyears...'''Thisdismissalofthedesigner'sresponsibilityfortheperformanceofmaterialsisverydifferentinthecaseofconcreteinwhicheverycivilengineerappearstohavebeenschooledinitsconstituentrawmaterials,theirsource,storage,mixing,transportandpouringaswellasthestrength.ToemphasisethewiderresponsibilitywhichtheengineerhasIgivethebackgroundtosomeofthematerialsandthetechniqueswhichtheengineerusestodayandmakethepointthatmanyofthedesignmethodsanddataincommonusearebasedonapproximationsandhavelimitationstotheirvalidity.Anumberofso-calledruleshavebeenderivedonanempiricalbasis;theyarevalidonlywithincertainlimits.TheyarenottruelawssuchasthoseofNewtonianmechanicswhichcouldbeappliedinallterrestrialandsomeuniversalcircumstancesandwhosevalidityextendsevenbeyondthevisionoftheirauthorhimself;albeitNewton'slawshavebeenmodified,ifnotsuperseded,byEinstein'sevenmorefundamentallaws.Thetitleofthisbookreflectsthispositionforithastoberecognisedthatthereispreciouslittletheoryinweldedjointdesignbutalotofpractice.Thereappearinthisbookformulaeforthestrengthoffilletweldswhichlookverytheoreticalwhereasinfacttheyareempiricallyderivedfromlargenumbersoftests.Similarlytherearegraphsoffatiguelifewhichlook Introductionxiiimathematicallybasedbutarestatisticallyderivedlinesoftheprobabilityoffailureoftestspecimensfromhundredsoffatiguetests;subsequenttheoreticalworkinthefieldoffracturemechanicshasexplainedwhythegraphshavetheslopewhichtheydobutwearealongwayfrombeingabletopredictonsoundscientificormathematicalgroundsthefatiguelifeofaparticularitemasacommonplacedesignactivity.Carbonequivalentformulaeareattemptstoquantifytheweldabilityofsteelsinrespectofhardenabilityoftheheataffectedzoneandareexamplesoftheempiricalorarbitraryrulesorformulaesurroundingmuchofweldingdesignandfabrication.Anotherexample,notrestrictedtoweldingbyanymeans,isinfracturemechanicswhichuses,albeitinamathematicalcontext,the±3/2physicallymeaninglessunitNmm.Perhapsintheabsenceofanythingbetterweshouldregardthesedevicesasnoworsethananecessaryandrespectablemathematicalfudge±perhapsananalogyofthecosmologist'sblackhole.Alittlehistoryhelpsustoputthingsinperspectiveandoftenhelpsustounderstandconceptswhichotherwisearedifficulttograsp.Thehistoricalbackgroundtoparticularmattersisimportanttotheunderstandingoftheengineer'scontributiontosociety,thewayinwhichdevelopmentstakeplaceandthereasonswhyfailuresoccur.IhaveusedthehistoryofBritainasabackgroundbutthisdoesnotimplyanybeliefonmypartthathistoryelsewherehasnotbeenrelevant.OnonehanditisapracticalmatterbecauseIamnotwritingahistorybookandmyreferencestohistoryareforperspectiveonlyanditisconvenienttousethatwhichIknowbest.OntheotherhandthereisacertainrationaleinusingBritishhistoryinthatBritainwasthecountryinwhichthemodernindustrialrevolutionbegan,eventuallyspreadingthroughtheEuropeancontinentandelsewhereandweseethatarcweldingprocesseswerethesubjectofdevelopmentinanumberofcountriesinthelatenineteenthcentury.ThelastdecadeofthetwentiethcenturysawtheindustrialbasemoveawayfromtheUK,andfromotherEuropeancountries,mainlytocountrieswithlowerwages.ManyproductsdesignedinEuropeancountriesandNorthAmericaarenowmanufacturedinAsia.Howeverinsomeindustriestheoppositehashappenedwhen,forexample,carsdesignedinJapanhavebeenmanufacturedforsomeyearsintheUKandtheUSA.Amoregeneralmovementhasbeentomakeuseofmanufacturingcapacityandspecialistprocesseswherevertheyareavailable.ComponentsforsomeUSaircraftaremadeinAustralia,theUKandothercountries;majorcomponentsforsomeUKaircraftaremadeinKorea.Theseareonlyafewexamplesofageneraltrendinwhichmanufacturingaswellastradeisbecomingglobal.Thisdispersionofindustrialactivitymakesitimportantthatanadequateunderstandingoftherelevanttechnologyexistsacrosstheglobeandthismustincludeweldinganditsassociatedactivities. xivIntroductionNotallengineeringprojectshavebeensuccessfulifmeasuredbyconventionalcommercialobjectivesbutsomeofthosewhichhavenotmettheseobjectivesaresuperbachievementsinatechnicalsense.TheConcordeairlinerandtheChannelTunnelaretwowhichspringtomind.TheConcordeisinserviceonlybecauseitsearlydevelopmentcostswereunderwrittenbytheUKandFrenchgovernments.TheChannelTunnellinkingEnglandandFrancebyrailhashadtobere-financedanditspaybacktimerescheduledfarbeyondcustomaryperiodsforreturnsoninvestment.Further,howdoweratethespaceprogrammes?Theirpaybacktimemayrunintodecades,ifnotcenturies,ifatall.Ostensiblywithascientificpurpose,thesuccessofmanyspaceprojectsismoreoftenmeasurednotinscientificorevencommercialtermsbutintheirpoliticaleffect.Thescientificresultscouldoftenhavebeenacquiredbylessextravagantmeans.Indefenceequipment,effectivenessandreliabilityundercombatconditions,possiblyafterlengthyperiodsinstorage,aretheprimerequirementsherealthoughcostmustalsobetakenintoaccount.Therearemanyprojectswhichhavefailedtoachieveoperationalsuccessthroughlackofcommitment,poorperformance,orthroughpoliticalinterference.Ingeneraltheirhumanconsequenceshavenotbeenlasting.Moresadlytherearethosefailureswhichhavecauseddeathandinjury.Mostofsuchengineeringcatastropheshavetheiroriginsintheuseofirrelevantorinvalidmethodsofanalysis,incompleteinformationorthelackofunderstandingofmaterialbehaviour,and,sooften,lackofcommunication.Suchcatastrophesarerelativelyrare,althoughatragedyforthoseinvolved.Whatiswritteninthisbookshowsthataccumulatedknowledge,derivedovertheyearsfromresearchandpracticalexperienceinweldedstructures,hasbeenincorporatedintogeneraldesignpractice.ReaderswillnotnecessarilyfindhereinalltheanswersbutIhopethatitwillcausethemtoasktherightquestions.Theactivityofengineeringdesigncallsontheknowledgeofavarietyofengineeringdisciplinesmanyofwhichhaveastrongtheoretical,scientificandintellectualbackgroundleavenedwithsomeratherarbitraryadjustmentsandassumptions.Bringingthisknowledgetoausefulpurposebyusingmaterialsinaneffectiveandeconomicwayisoneoftheskillsoftheengineerwhichincludemakingdecisionsontheneedforandthepositioningofjoints,betheypermanentortemporary,betweensimilarordissimilarmaterialswhichisthemainthemeofthisbook.Howeverasinallwalksofengineeringtheweldingdesignermustbeawarethathavinglearnedhisstuffhecannotjustleanbackandproducedesignsbasedonthatknowledge.Theworldhasahabitofchangingarounduswhichleadsnotonlytotheneedforustorecognisetheneedtofaceuptodemandsfornewtechnologybutalsobeingawarethatsomeoftheoldproblemsrevisitus.WinstonChurchillisquotedashavingsaidthatthefurtherbackyoulookthefurtherforwardyoucansee. 1Theengineer1.1ResponsibilityoftheengineerAsweenterthethirdmillenniumannisdomini,mostoftheworld'spopulationcontinuesincreasinglytorelyonman-madeandcentralisedsystemsforproducinganddistributingfoodandmedicinesandforconvertingenergyintousableforms.Muchofthesesystemsreliesonthe,oftenunrecognised,workofengineers.Theengineer'sresponsibilitytosocietyrequiresthatnotonlydoeshekeepuptodatewiththeeverfasterchangingknowledgeandpracticesbutthatherecognisestheboundariesofhisownknowledge.Theengineerdevisesandmakesstructuresanddevicestoperformdutiesorachieveresults.Insodoingheemployshisknowledgeofthenaturalworldandthewayinwhichitworksasrevealedbyscientists,andheusestechniquesofpredictionandsimulationdevelopedbymathematicians.Hehastoknowwhichmaterialsareavailabletomeettherequirements,theirphysicalandchemicalcharacteristicsandhowtheycanbefashionedtoproduceanartefactandwhattreatmenttheymustbegiventoenablethemtosurvivetheenvironment.Themotivationandmethodsofworkingoftheengineerareverydifferentfromthoseofascientistormathematician.Ascientistmakesobservationsofthenaturalworld,offershypothesesastohowitworksandconductsexperimentstotestthevalidityofhishypothesis;thencehetriestoderiveanexplanationofthecomposition,structureormodeofoperationoftheobjectorthemechanism.Amathematicianstartsfromtheoppositepositionandevolvestheoreticalconceptsbymeansofwhichhemaytrytoexplainthebehaviourofthenaturalworld,ortheuniversewhateverthatmaybeheldtobe.Scientistsandthemathematiciansbothaimtoseekthetruthwithoutcompromiseandalthoughtheymaypublishresultsandconclusionsasevidenceoftheirfindingstheirworkcanneverbefinished.Incontrasttheengineerhastoachievearesultwithinaspecifiedtimeandcostandrarelyhastheresourcesorthetimetobeabletoidentifyandverifyeverypossible 2Weldeddesign±theoryandpracticepieceofinformationabouttheenvironmentinwhichtheartefacthastooperateortheresponseoftheartefacttothatenvironment.Hehastoworkwithinadegreeofuncertainty,expressedbytheprobabilitythattheartefactwilldowhatisexpectedofitatadefinedcostandforaspecifiedlife.Theengineer'scircumstanceisperhapssummarisedbestbytheoftquotedrequest:`Idon'twantitperfect,IwantitThursday!'Oncetheengineer'sworkiscompletehecannotgobackandchangeitwithoutdisproportionateconsequences;itisthereforalltoseeanduse.TheancientRomanswereparticularlydemandingoftheirbridgeengineers;theengineer'snamehadtobecarvedonastoneinthebridge,nottopraisetheengineerbuttoknowwhotoexecuteifthebridgeshouldcollapseinuse!Peopleplacetheirlivesinthehandsofengineerseverydaywhentheytravel,anactivityassociatedwithwhichisapredictableprobabilityofbeingkilledorinjuredbytheomissionsoftheirfellowdrivers,themistakesofprofessionaldriversandcaptainsorthefailingsoftheengineerswhodesigned,manufacturedandmaintainedthemodeoftransport.Theengineer'sroleistobeseennotonlyinthevehicleitself,whetherthatbeonland,seaorair,butalsointheroad,bridge,harbourorairport,andinthenavigationalaidswhichaboundandnowpermitapersontoknowtheirpositiontowithinafewmetresoverandabovealargepartoftheearth.Humanerrorisfrequentlyquotedasthereasonforacatastropheandusuallymeansanerroronthepartofadriver,amarinerorapilot.Othercausesareoftenlumpedunderthecatch-allcategoryofmechanicalfailureasifsucheventswerebeyondthehandofman;anaõÈveattribution,ifevertherewereone,forsomewheredownthelinepeoplewereinvolvedintheconception,design,manufactureandmaintenanceofthedevice.Itisthereforestillhumanerrorwhichcausedtheproblemevenifnotofthoseimmediatelyinvolved.Ifweneedtolabelthecauseofthecatastrophe,whatweshouldreallydoistoplaceitinoneof,say,fourcategories,allundertheheadingofhumanerror,whichwouldbefailureinspecification,design,operationormaintenance.An`ActofGod'sobelovedbyjudgesisaget-out.Itusuallymeansacircumstanceorsetofcircumstanceswhichadesigner,operatororlegislatoroughttohavebeenabletopredictandallowforbutchosetoignore.Ifthisseemsveryharshwehaveonlytolookatthenumberofliveslostinbulkcarriersatseainthepastyears.Therestillseemstobeacultureinseafaringwhichacceptsthatthereareunavoidablehazardsandwhicharereflectedinthenineteenthcenturyhymnline`...forthoseinperilonthesea'.Eventodaythereareculturesinsomecountrieswhichdonotseedeathorinjurybyman-madecircumstancesaspreventableorevenneedingprevention;conceptsofriskjustdonotexistinsomeplaces.Thatisnottosaythatanyactivitycanbefreeofhazards;weareexposedtohazardsthroughoutourlife.Whattheengineershouldbedoingistoconductactivitiesinsuchawaythattheprobabilityofnotsurvivingthathazardis Theengineer3knownandsetatanacceptedlevelforthegeneralpublic,leavingthosewhowishtoindulgeinhighriskactivitiestodosoontheirown.Weplaceourlivesinthehandsofengineersinmanymorewaysthantheseobviousones.Whenweusedomesticmachinessuchasmicrowaveovenswiththeirpotentiallyinjuriousradiation,dishwashersandwashingmachineswithapotentiallylethal240Vsuppliedtoamachinerunninginwaterintowhichtheoperatorcansafelyputhisorherhands.Patientsplacetheirlivesinthehandsofengineerswhentheysubmitthemselvestosurgeryrequiringthesubstitutionoftheirbodilyfunctionsbymachineswhichtemporarilytaketheplaceoftheirhearts,lungsandkidneys.Otherssurviveonpermanentreplacementsfortheirownbodilypartswithman-madeimplantsbetheyvalves,jointsorotherobjects.Aneminentheartsurgeonsaidontelevisionrecentlythathearttransplantsweresimple;althoughthiswasperhapsathrowawayremarkonehastoobservethatifitissimpleforhim,whichseemsunlikely,itisonlysobecauseofdevelopmentsinimmunology,onpost-operativecriticalcareandonanaesthesia(notjusttheoldfashionedgasbutthewholesubstitutionandmaintenanceofcompletecirculatoryandpulmonaryfunctions)whichenablesittobesoandwhichreliesoncomplexmachineryrequiringahighlevelofengineeringskillindesign,manufacturingandmaintenance.Weplaceourlivelihoodsinthehandsofengineerswhomakemachinerywhetheritbeforthefactoryortheoffice.Businessesandindividualsrelyontelecommunicationstocommunicatewithothersandforsomeitwouldseemthatlifewithouttelevisionandamobiletelephonewouldbeatbestmeaninglessandatworstintolerable.Werelyonanavailablesupplyofenergytoenableustouseallofthisequipment,tokeepourselveswarmandtocookourfood.ItistheengineerwhoconvertstheenergycontainedinandaroundtheEarthandtheSuntoproducethissupplyofusableenergytoaremarkablelevelofreliabilityandconsistencybeitintheformoffossilfuelsorelectricityderivedfromthemornuclearreactions.1.2AchievementsoftheengineerTheachievementsoftheengineerduringthesecondhalfofthetwentiethcenturyareperhapsmostpopularlyrecognisedinthedevelopmentofdigitalcomputersandotherelectronicallybasedequipmentthroughtheexploita-tionofthediscoveryofsemi-conductors,ortransistorsastheycametobeknown.ThesubsequentgrowthinthediversityoftheuseofcomputerscouldhardlyhavebeenexpectedtohavetakenplacehadwecontinuedtorelyonthethermionicvalveinventedbySirAlexanderFlemingin1904,letalonethenineteenthcenturymechanicalcalculatingengineofWilliamBabbage.Howeverletusnotforgetthatatthebeginningofthetwenty-first 4Weldeddesign±theoryandpracticecenturythevisualdisplaysofmostcomputersandtelecommunicationsequipmentstillrelyonthetechnologyofthermionicemission.Theliquidcrystalhasoccupiedasmallareaofapplicationandthelightemittingdiodehasyettoreachitsfullpotential.Theimpactofelectronicprocessinghasbeenfeltbothindomesticandinbusinesslifeacrosstheworldsothatalmosteverybodycanseetheeffectatfirsthand.Historicallymostotherengineeringachievementsprobablyhavehadalessimmediateandlesspersonalimpactthanthesemi-conductorbuthavebeenequallysignificanttothewayinwhichtradeandlifeingeneralwasconducted.AsfaraslifeintheBritishIsleswasconcernedthisprocessofacceleratingchangemadepossiblebytheengineermightperhapshavebegunwiththebuildingoftheroadsystem,centrallyheatedvillasandthesettingupofindustriesbytheRomansinthefirstfewyearsAD.Howevertheirwithdrawal400yearslaterwasaccompaniedbythecollapseofcivilisationinBritain.TheinvadingAnglesandSaxonsenslavedordrovetheindigenouspopulationintothenorthandwest;theyplunderedtheformerRomantownsandletthemfallintoruin,preferringtoliveinsmallself-containedsettlements.InothercountriestheRomansleftagreatervarietyoffeatures;notonlyroadsandvillasbutmightystructuressuchasthatmagnificentaqueduct,thePontduGardinthesouthofFrance(Fig.1.1).HundredsofyearsweretopassbeforenewtypesofstructureswereerectedandoftheseperhapsthegreatestwerethecathedralsbuiltbytheNormansinthenorthofFranceandinEngland.Themainstructureofthesecomprisedstonearchessupportedbyexternalbuttressesinbetween1.1ThePontduGard(photographbyBernardLiegeois). Theengineer5whichwereplacedtimberbeamssupportingtheroof.Exceptforthesebeamsallthematerialwasincompression.Themodernconceptofastructurewithseparatemembersintension,compressionandshearwhichwenowcallchords,braces,ties,webs,etc.appearsinexamplessuchasElyCathedralintheeastofEngland.Thecathedral'scentraltower,builtinthefourteenthcentury,isofanoctagonalplanformsupportedononlyeightarches.Thistoweritselfsupportsatimberframedstructurecalledthelantern(Fig.1.2).Howeverletusnotbelievethattheengineersofthosedayswerealwayssuccessful;thisoctagonaltowerandlanternatElyhadbeenbuilttoreplacetheNormantowerwhichcollapsedinabout1322.ExceptperhapsforthedrainingoftheFens,alsointheeastofEngland,whichwascommencedbytheDutchengineer,CorneliusVermuyden,underKingCharlesIin1630,nothingfurtherinthemodernsenseofaregionalornationalinfrastructurewasdevelopedinBritainuntilthebuildingofcanalsintheeighteenthcentury.Thesewereusedformovingbulkmaterialsneededtofeedtheburgeoningindustrialrevolutionandthemotivepowerwasprovidedbythehorse.Canalswerefollowedby,andtoagreatextentsupersededby,therailwaysofthenineteenthcenturypoweredbysteamwhichservedtocarrybothgoodsandpassengers,eventuallyinnumbers,speedandcomfortwhichtheroadscouldnotoffer.Alongsidethesecametheemergenceofthelargeoceangoingship,alsodrivenbysteam,toservetheinternationaltradeingoodsofalltypes.Thecontributionoftheinventorsanddevelopersofthesteamengine,initiallyusedtopumpwaterfrommines,wasthereforecentraltothegrowthoftransport.AmongstthemweacknowledgeSavory,Newcomen,Trevithick,WattandStephenson.Alongsidethesedevelopmentsnecessarilygrewtheindustriestobuildthemeansandtomaketheequipmentfortransportandwhichinturnprovidedamajorreasonfortheexistenceofatransportsystem,namelytheproductionofgoodsfordomesticand,increasingly,overseasconsumption.Todaysteamisstillamajormeansoftransferringenergyinbothfossilfiredandnuclearpowerstationsaswellasinlargeshipsusingturbines.Itsearlierroleinsmallerstationaryplantandinothertransportapplicationswastakenoverbytheinternalcombustionenginebothinitspistonandturbineforms.Subsequentlytheroleofthestationaryenginehasbeentakenoveralmostentirelybytheelectricmotor.Inthesecondhalfofthetwentiethcenturythefreightcarryingroleoftherailwaysbecamesubstantiallysubsumedbyroadvehiclesresultingfromthebuildingofmotorwaysandincreasingthecapacityofexistingmainroads(regardlessofthewiderissuesoftruecostandenvironmentaldamage).Onaworldwidebasisthedevelopmentandconstructionofevenlargershipsforthecheaplongdistancecarriageofbulkmaterialsandoflargeraircraftforprovidingcheaptravelforthemassesweretwootherachievements.Theirusebuiltupcomparativelyslowlyinthesecondhalfofthecenturybuttheiractual 6Weldeddesign±theoryandpractice1.2ThelanternofElyCathedral(photographbyJanetHicks,drawingsbycourtesyofPurcellMillerTrittonandPartners). Theengineer7developmenthadtakenplacenotinsmallincrementsbutinlargesteps.Themotivationfortheshipandaircraftchangeswasdifferentineachcase.AmajorincentiveforbuildinglargershipswastheclosureoftheSuezCanalin1956sothatoiltankersfromtheMiddleEastoilfieldshadtotravelaroundtheCapeofGoodHopetoreachEurope.Therestraintofthecanalonvesselsizethennolongerappliedandtheeconomyofscaleaffordedbylargetankersandbulkcarrierscompensatedfortheextradistance.ThedevelopmentofalargercivilaircraftwasaboldcommercialdecisionbytheBoeingCompany.Itsintroductionofthetype747intheearly1970simmediatelyincreasedthepassengerloadfromamaximumofaround150tosomethingapproaching400.InanotherdirectionofdevelopmentataroundthesametimeBritishAerospace(orrather,itspredecessors)andAeÂrospatialeofferedairlinepassengersthefirst,andsofartheonly,meansofsupersonictravel.Alongsidethesedevelopmentswerethechangesinenergyconversionbothtonuclearpoweraswellastolargerandmoreefficientfossil-fuelledpowergenerators.Inthelastthirdofthecenturyextractionofoilandgasfromdeeperoceansledtoveryrapidadvancementsinstructuralsteeldesignandinmaterialsandjoiningtechnologiesinthe1970s.Theseadvanceshavespunoffintowiderfieldsofstructuralengineeringinwhichphilosophiesofstructuraldesignaddressedmoreandmoreinaformalwaymattersofintegrityandeconomy.Insteelworkdesigngenerallymorerationalapproachestoprobabilitiesofoccurrencesofloadsandthevariabilityofmaterialpropertieswereconsideredandintroduced.Theserequiredacloserattentiontoquestionsofqualityinthesenseofconsistencyoftheproductandfreedomfromfeatureswhichmightrendertheproductunabletoperformitsfunction.1.3TheroleofweldingBearinginmindtheoverallsubjectofthisbookweoughttoconsiderifandhowweldinginfluencedthesedevelopments.Todothiswecouldpostulatea`whatif?'scenario:whatifweldinghadnotbeeninvented?Thisisnotanentirelysatisfactoryapproachsincehistoryshowsthatthemeansofteninfluencestheendandviceversa;industryoftenmaintainsandimprovesmethodswhichmightbecalledoldfashioned.Asanexample,machiningofmetalswas,manyyearsago,referredtobyaproponentofchemicaletchingasanarchaicprocessinwhichoneknocksbitsoffonepieceofmetalwithanotherpieceofmetal,notmuchofanadvanceonStoneAgeflintknapping.Perhapsthiswas,andstillis,true;nonethelessmachiningisstillwidelyusedandshapingofmetalsbychemicalmeansisstillaminorityprocess.Rivetsweregivenuphalfacenturyagobyalmostallindustriesexcepttheaircraftindustrywhichkeepsthembecausetheyhaven'tfoundamoresuitablewayofjoiningtheirchosenmaterials;theymakeaverygood 8Weldeddesign±theoryandpracticejobofit,claimingthebenefitoverweldingofastructurewithnaturalcrackstoppers.AsaconfirmationofitsintegrityamajorjointinaConcordefuselagewastakenapartafter20years'serviceandfoundtobecompletelysound.Solookingattheapplicationofweldingthereareanumberofaspectswhichwecouldlabelfeasibility,performanceandcosts.Itishardtoenvisagethecontainmentvesselofanuclearreactororamodernboilerdrumorheatexchangerbeingmadebyrivetinganymorethanwecouldconceiveofagasoroilpipelinebeingmadeotherthanbywelding.Ifweldinghadn'tbeenthereperhapsanothermethodwouldhavebeenused,orperhapsweldingwouldhavebeeninventedforthepurpose.Itdoesseemhighlylikelythatthelowcostsofmodernshipbuilding,operation,modificationandrepaircanbeattributedtothelowercostsofweldedfabricationoflargeplatestructuresoverrivetinginadditiontowhichistheweightsaving.Asearlyas1933theeditorofthefirsteditionofTheWeldingIndustrywrote`...thehullsofGermanpocketbattleshipsarebeingfabricatedentirelywithwelding±apracticewhichproducesaweightsavingof1000tonspership'.ThemotivationforthisattentiontoweightwasthatundertheTreatyofVersaillesaftertheFirstWorldWarGermanywasnotallowedtobuildwarshipsofover10000tons.Ayearlater,in1934,awriterinthesamejournalvisitedtheworksofAVRoeinManchester,forerunnerofAvrowholaterdesignedandbuiltmanyaircrafttypesincludingtheLancaster,Lincoln,ShackletonandVulcan.`Iwaspreparedtoseeaconsiderableamountofwelding,butthepitchofexcellencetowhichMessrsAVRoehavebroughtoxy-acetyleneweldinginthefabricationoffuselagesandwings,theirmanytypesofaircraftandthenumberofweldersthatwerebeingemployedsimultaneouslyinthiswork,gaveme,asaweldingengineer,greatpleasuretowitness.'Thewriterwasreferringtosteelframeswhichtodaywemightstillseeaseminentlyweldable.Howeverthescopeforweldinginairframeswastobehugelyreducedinonlyafewyearsbythechangeoverinthelater1930sfromfabriccoveredsteelframestoaluminiumalloymonocoquestructurescomprisingframes,skinandstringersforthefuselageandspars,ribsandskinforthewingsandtailsurfaces.ThisseriesofalloyswasunsuitableforarcweldingbutresistancespotweldingwasusedmuchlaterforattachingthelowerfuselageskinsoftheBoeing707airlinertotheframesandstringersaswerethoseoftheHandleyPageVictorandHeraldaircraft.Thematerialused,anAl±Zn±Mgalloy,wasamenabletospotweldingbutcontrolswereplacedonhardnesstoavoidstresscorrosioncracking.Itcannotbesaidthatwithoutweldingtheseaircraftwouldnothavebeenmade,itwasjustanothersuitablejoiningprocess.TheBristolT188experimentalsupersonicaircraftofthelate1950shadanairframemadeofTIGspot-weldedausteniticstainlesssteel.Thismaterialwaschosenforitsabilitytomaintainitsstrengthatthetemperaturesdevelopedbyaerodynamicfrictioninsupersonicflight,anditalsohappenedtobe Theengineer9weldable.ItwasnotasolutionwhichwaseventuallyadoptedfortheConcordeinwhicharivetedaluminiumalloystructureisusedbutwhosetemperatureismoderatedbycoolingitwiththeenginefuel.Apartfromtheseexamplesandtheweldedsteeltubularspaceframesformerlyusedinlightfixedwingaircraftandhelicopters,airframeshavebeenrivetedandcontinuetobeso.Incontrastmanyaircraftenginecomponentsaremadebyweldingbutgasturbinesalwayswereandsotheroleofweldinginthegrowthofaeroplanesizeandspeedisnotsospecific.Inroadvehiclebodyandwhitegoodsmanufacture,theweldingdevelopmentswhichhavesupportedhighproductionratesandaccuracyoffabricationhavebeenasmuchinthefieldoftooling,controlandroboticsasintheweldingprocessesthemselves.Inconstructionwork,economiesareachievedthroughtheuseofshop-weldedframesormemberswhichareboltedtogetheronsite;theextentoftheuseofweldingonsitevariesbetweencountries.Mechanicalhandlingandconstructionequipmenthaveundoubtedlybenefitedfromtheapplicationofwelding;manyofthemachinesinusetodaywouldbeverycumbersome,costlytomakeanddifficulttomaintainifweldedassemblieswerenotused.Rivetedroadandrailbridgesareamongstitemswhichareathingofthepasthavingbeensucceededbyweldedfabrications;apartfromtheweightsaving,thesimplicityoflineandlackoflapjointsmakesprotectionfromcorrosioneasierandsomemaysaythattheappearanceismorepleasing.Anexaminationofthehistoryofengineeringwillshowthatfewobjectsaredesignedfromscratch;mosttendtobestepdevelopmentsfromthepreviousitem.Motorcarsstartedoffbeingcalled`horselesscarriages'whichisexactlywhattheywere.Theywerehorsedrawncarriageswithanengineadded;theshaftsweretakenoffandsteeringeffectedbyatiller.Evennow`dashboard'remainsineverydayspeechrevealingitsoriginsintheboardwhichprotectedthedriverfromthemudandstonesthrownupbythehorse'shooves.Muchrecentsoftwareforpersonalcomputersreplicatesthephysicalfeaturesofoldermachineryinthe`buttons',whichdisplaysanextraordinarylevelofconservatism.Asimilarconservatismcanbeseenintheadoptionofnewjoiningprocesses.Thefirstweldedshipswerejustweldedversionsoftherivetedconstruction.Ithastakendecadesfordesignerstostopcopyingcastingsbyputtinglittlegussetsonweldeditems.Howeveritcanbeobservedthatonceanewmanufacturingtechniqueisadopted,andtheworkspractices,planningandcostingadjustedtosuit,itwilltendtobeusedexclusivelyeventhoughtheremaybeargumentsforusingthepreviousprocessesincertaincircumstances.1.4OthermaterialsHavingreflectedonthesepointsourthoughtsmustnotbetrammelledbyignoranceofotherjoiningprocessesorindeedbymaterialsotherthanthe 10Weldeddesign±theoryandpracticemetalswhichhavebeenthecustomarysubjectsofwelding.Thisbookconcentratesonarcweldingofmetalsbecausetheremustbealimittoitsscopeandalsobecausethatiswheretheauthor'sexperiencelies.Moreandmoreweseeothermetalsandnon-metalsbeingusedsuccessfullyinbothtraditionalandnovelcircumstancesandtheengineermustbeawareofalltherelevantoptions.1.5TheweldingengineeraspartoftheteamAsinmostotherprofessionstherearefewcircumstancestodaywhereonepersoncantakeallthecreditforaparticularachievementalthoughaleaderisessential.Mostengineeringprojectsrequirethecontributionsofavarietyofengineeringdisciplinesinateam.Oneofthemembersofthatteaminmanyproductsorprojectsistheweldingengineer.Theexecutionoftheresponsibilitiesoftheweldingengineertakesplaceattheinterfaceofanumberofconventionaltechnologies.Forcontributingtothedesignoftheweldedproducttheseincludestructuralandmechanicalengineering,materialprocessing,weldabilityandperformanceandcorrosionscience.Forthesettingupandoperationofweldingplanttheyincludeelectrical,mechanicalandproductionengineering,thephysicsandchemistryofgases.Inaddition,theweldingengineermustbefamiliarwiththegeneralmanagementofindustrialprocessesandpersonnelaswellasthehealthandsafetyaspectsoftheweldingoperationsandmaterials.Latetwentiethcenturypracticeinsomeareaswouldseemtorequirethatresponsibilityfortheworkbehiddeninafogofcontracts,sub-contractsandsub-sub-contractsadinfinitumthroughwhichareemployedconceptualdesigners,detaildesigners,shopdraughtsmen,quantitysurveyors,measure-mentengineers,approvalsengineers,specificationwriters,contractwriters,purchasingagencies,maincontractors,fabricators,sub-fabricatorsandinspectioncompanies.Allthesearesurroundedbyunderwritersandtheirwarrantysurveyorsandlossadjustersneededincaseofaninadequatejobbroughtaboutbyawardingcontractsonthebasisofpriceandnotontheabilitytodothework.Responsibilitiesbecomeblurredanditisimportantthatengineersofeachdisciplineareatleastawareof,ifnotfamiliarwith,theircolleagues'roles. 2Metals2.1Steels2.1.1TheoriginsofsteelThefirstironconstructionwhichmakesuseofstructuralengineeringprincipleswasabridgebuiltbyAbrahamDarbyin1779overagorgeknownasCoalbrookdalethroughwhichrunstheRiverSevernataplacenamedafterit,Ironbridge,inShropshireintheUK(Fig.2.1).ItwasinthisareathatDarby'sgrandfatherhad,in1709,firstsucceededinsmeltingironwithcokeratherthancharcoal,atechniquewhichmadepossiblethemassproductionofironatanaffordableprice.Thebridgeisintheformofframesassembledfromcastironbarsheldtogetherbywedges,atechniquecarriedoverfromtimberconstruction.CastironcontinuedtobeusedforbridgesintothenineteenthcenturyuntilRobertStephenson'sbridgeovertheRiverDeeatChestercollapsedunderatrainin1847killingfivepeople.Althoughthetensionloadsweretakenbywroughtironbarsthebridgefailedattheirattachmenttothecastiron.AtthetimeofthateventStephensonwasconstructingtheNewcastleHighLevelbridgeusingcastiron.Howeverhetookgreatcareindesigningthebowandstringgirdersrestingonfivestonepiers45mabovetheRiverTynesothatexcessivetensionwasavoided.Thespansareshort,themembersmassiveandparticularcarewastakenovertheircastingandtesting.Workcommencedonthebridgein1846andwascompletedinthreeyears;itstandstothisdaycarryingroadandrailtrafficonitstwodecks.NeverthelesspublicoutcryattheDeetragedycausedthedemiseofcastironforbridgebuilding;itsplacewastakenbywroughtiron,whichisalmostpureironandaveryductilematerial,exceptformembersincompressionsuchascolumns.Steelsdiscoveredthousandsofyearsagoacquiredwideusageforcutlery,toolsandweapons;aheattreatmentcomprisingquenchingandtemperingwasappliedasameansofadjustingthehardness,strengthandtoughnessofthesteel.Eventuallysteelsbecameoneofthemostcommongroupofmetals 12Weldeddesign±theoryandpractice2.1Ironbridge(photographbycourtesyoftheIronbridgeGorgeMuseum).ineverydayuseandinmanywaystheyarethemostmetallurgicallycomplex.Crudeiron,orpigironasitisknown,isusuallymadebysmeltingironorewithcokeandlimestone.Ithasahighcarboncontentwhichmakesitbrittleandsoitisconvertedtomildsteelbyremovingsomeormostofthecarbon.ThiswasfirstdoneonalargeindustrialscaleusingtheconverterinventedbyHenryBessemerwhoannouncedhisprocesstotheBritishAssociationin1856.SomesaythathebasedhisprocessonapatentofJamesNaysmithinwhichsteamwasblownthroughthemoltenirontoremovecarbon;othersheldthathebaseditonthe`pneumaticmethod',inventedtwoyearsearlierbyanAmerican,WilliamKelly.NeverthelessitwastheBessemerprocessthatbroughtaboutthefirstgreatexpansionoftheBritishandAmericansteelindustries,largelyowingtothemechanicalsuperiorityofBessemer'sconverter.Developmentsinindustrialsteelmakinginthelatterpartofthenineteenthcenturyandinthetwentiethcenturyleadtothepresentdaypositionwherewithfineadjustmentofthesteelcompositionandmicrostructureitispossibletoprovideawiderangeofweldablesteelshavingpropertiestosuittherangeofdutiesandenvironmentscalledupon.Thisbookdoesnotaimtoteachthehistoryandpracticeofironandsteelmaking;thatrepresentsafascinatingstudyinitsownrightandthereader2interestedinsuchmattersshouldreadworksbyauthorssuchasCottrell.Theabilityofsteeltohaveitspropertieschangedbyheattreatmentisa Metals13valuablefeaturebutitalsomakesthejoiningofitbyweldingparticularlycomplicated.Beforestudyingtheeffectsofthevariousweldingprocessesonsteelweoughttosee,inasimpleway,howironbehavesonitsown.2.1.2TheatomicstructureofironTheironatom,whichisgiventhesymbolFe,hasanatomicweightof56whichcompareswithaluminium,Al,at27,lead,Pb,at207andcarbon,C,at12.Inironatroomtemperaturetheatomsarearrangedinaregularpattern,orlattice,whichiscalledbodycentredcubicorbccforshort.Thesmallestrepeatablethreedimensionalpatternisthenacubewithanatomateachcornerplusoneinthemiddleofthecube.Ironinthisformiscalledferrite(Fig.2.2(a)).(a)(b)2.2(a)Bodycentredcubicstructure;(b)facecentredcubicstructure.oIfironisheatedto910C,almostwhitehot,thelayoutoftheatomsinthelatticechangesandtheyadoptapatterninwhichoneatomsitsinthemiddleofeachplanarsquareoftheoldbccpattern.Thisnewpatterniscalledfacecentredcubic,abbreviatedtofcc.Ironinthisformiscalledaustenite(Fig.2.2(b)).Whenatomsarepackedinoneoftheseregularpatternsthestructureisdescribedascrystalline.Individualcrystalscanbeseenunderamicroscopeasgrainsthesizeofwhichcanhaveastrongeffectonthemechanicalpropertiesofthesteel.Furthermoresomeimportantphysicaland 14Weldeddesign±theoryandpracticemetallurgicalchangescanbeinitiatedattheboundariesofthegrains.Thechangefromonelatticepatterntoanotherasthetemperaturechangesiscalledatransformation.Whenirontransformsfromferrite(bcc)toaustenite(fcc)theatomsbecomemorecloselypackedandthevolumeperatomofironchangeswhichgeneratesinternalstressesduringthetransformation.Althoughthefccpatternismorecloselypackedthespacesbetweentheatomsarelargerthaninthebccpatternwhich,weshallseelater,isimportantwhenalloyingelementsarepresent.2.1.3AlloyingelementsinsteelThepresenceofmorethanabout0.1%byweightofcarboninironformsthebasisofthemodernstructuralsteels.Carbonatomssitbetweentheironatomsandprovideastrengtheningeffectbyresistingrelativemovementsoftherowsofatomswhichwouldoccurwhenthematerialyields.Otheralloyingelementswithlargeratomsthancarboncanactuallytaketheplaceofsomeoftheironatomsandincreasethestrengthabovethatofthesimplecarbonsteel;therelativestrengtheningeffectofthesevariouselementsmaydifferwithtemperature.Commonalloyingelementsaremanganese,chromium,nickelandmolybdenum,whichmayinanycasehavebeenaddedforotherreasons,e.g.manganesetocombinewithsulphursopreventingembrittlement,chromiumtoimpartresistancetooxidationathightemperatures,nickeltoincreasehardness,andmolybdenumtopreventbrittleness.2.1.4HeattreatmentsWelearnedearlierthatalthoughtheironatomsinaustenitearemorecloselypackedthaninferritetherearelargerspacesbetweenthem.Aresultofthisisthatcarbonismoresolubleinaustenitethaninferritewhichmeansthatcarbonistakenintosolutionwhensteelisheatedtoatemperatureatwhichtheface-centredlatticeexists.Ifthissolutionisrapidlycooled,i.e.quenched,thecarbonisretainedinsolidsolutionandthesteeltransformsbyashearingmechanismtoastronghardmicrostructurecalledmartensite.Thehigherthecarboncontenttheloweristhecoolingratewhichwillcausethistransformationand,asacorollary,thehigherthecarboncontenttheharderwillbethemicrostructureforthesamecoolingrate.Thismartensiteisnotastoughasferriteandcanbemoresusceptibletosomeformsofcorrosionandcracking.WeshallseeinChapter11thatthisismostimportantinconsideringtheweldingofsteel.Thereadinessofasteeltoformahardmicrostructureisknownasitshardenabilitywhichisamostimportantconceptinwelding.Ifmartensiteisformedbyquenchingandisthenheatedtoanintermediatetemperature(tempered),althoughitissoftened,a Metals15proportionofitsstrengthisretainedwithasubstantialincreaseintoughnessandductility.Quenchingandtemperingareusedtoachievethedesiredbalancebetweenstrength,hardnessandtoughnessofsteelsforvariousapplications.Iftheausteniteiscooledslowlyinthefirstplacethecarboncannotremaininsolutionandsomeisprecipitatedasironcarbideamongsttheferritewithinametallurgicalstructurecalledpearlite.Theresultingstructurecanbeseenunderthemicroscopeasamixtureofferriteandpearlitegrains.Withtheadditionofotheralloyingelementsthesemechanismsbecomeextremelycomplicated,eachelementhavingitsowneffectonthetransformationand,inparticular,onthehardness.Toallowtheweldingengineertodesignweldingproceduresforarangeofsteelsinasimplewayformulaehavebeendevisedwhichenabletheeffectofthedifferentalloyingelementsonhardenabilitytobeallowedforintermsoftheirequivalenteffecttothatofcarbon.OnesuchcommonlyusedformulaistheIIWformulawhichgivesthecarbonequivalentofasteelinthecarbon±manganesefamilyas:MnCr+Mo+VNi+CuCeq=C+ÐÐ+ÐÐÐÐÐÐ+ÐÐÐÐ.[2.1]6515Thisrepresentspercentagequantitiesbyweightandwhatthisformulasaysineffectisthatweightforweightmanganesehasone-sixthofthehardeningeffectofcarbon,chromiumone-fifthandnickelone-fifteenth.Thisisaveryscientificlookingformulabutitwasderivedfromexperimentalobserva-tions,andperhapsonedaysomeonewillbeabletoshowthatitrepresentscertainfundamentalsintransformationmechanics.Atypicalmaximumfigureforthecarbonequivalentwhichcanbetoleratedusingconventionalarcweldingtechniqueswithoutriskinghighheataffectedzonehardnessandhydrogencrackingisabout0.45%.Somefabricationspecificationsputanupperlimitforheataffectedzonehardnessof350Hvtoavoidhydrogencrackingbutthisisveryarbitraryanddependsonarangeofcircumstances.Limitsarealsoplacedonhardnesstoavoidstresscorrosioncrackingwhichcanariseinsomeindustrialapplicationssuchaspipelinescarrying`sour'gas,i.e.gascontaininghydrogensulphide.Theheataffectedzonehardnesscanbelimitedbypreheatingwhichmakesthepartswarmorhotwhenweldingstartsandsoreducestherateatwhichtheheataffectedzonecoolsafterwelding.Preheattemperaturescanbebetween508and2008Cdependingonthehydrogencontentoftheweldingconsumable,thesteelcomposition,thethicknessandtheweldingheatinput.Forsomehardenablesteelsinthicksectionswhentheheataffectedzonehardnessremainshighevenwithpreheat,thelevelofretainedhydrogen,andsotheriskofcracking,canbereducedbypostheating,i.e.maintainingthepreheattemperatureforsomehoursafterwelding. 16Weldeddesign±theoryandpracticeSometimeslettingtheworkcooldownslowlyunderfireproofblanketsissufficient.Wherethecomposition,thicknessoraccessmakespreheatingimpracticableorineffectiveanausteniticweldingconsumablecanbeused.Thisabsorbshydrogeninsteadoflettingitconcentrateintheheataffectedzonebutthereisthedisadvantageinthatahardheataffectedzonestillremainswhichmaybesusceptibletostresscorrosioncracking;inadditiontheverydifferentchemicalcompositionsoftheparentandweldmetalsmaybeunsuitableincertainenvironments.2.1.5SteelsasengineeringmaterialsSteelsareusedextensivelyinengineeringproductsforanumberofreasons.Firstly,therawmaterialsareabundant±ironissecondonlytoaluminiuminoccurrenceintheearth'scrustbutaluminiumismuchmorecostlytoextractfromitsore;secondly,steelmakingprocessesarerelativelystraightforwardandforsometypesproductioncanbeaugmentedbyre-cyclingscrapsteel;thirdly,manysteelsarereadilyformedandfabricated.Theabilityofcarbonsteels±intheweldingcontextthismeansthosesteelswithfrom0.1%to0.3%carbon±tohavetheirpropertieschangedbyworkhardening,heattreatmentoralloyingisofimmensevalue.Perhapstheonlydownsidetothecarbonsteelsistheirpropensitytorustwhenexposedtoairandwater.Thestainlesssteelsarebasicallyironwith18±25%chromium,somealsowithnickel,andverylittlecarbon.Therearemanytypesofstainlesssteelandcaremustbeexercisedinspecifyingthemandindesigningweldingprocedurestoensurethatthechromiumdoesnotcombinewithcarbontoformchromiumcarbideundertheheatofwelding.Thiscombinationdepletesthechromiumlocaltotheweldandcanleadtolocallossofcorrosionresistance.Thiscanbeseeninsomeoldtablekniveswherethebladehasbeenweldedtothetangandshowsupasalineofpitsnearthebottomofthebladewhichissometimescalled`welddecay'.Toreducetheriskofthisdepletionofchromiumthelevelofcarboncanbereducedortitaniumorniobiumcanbeadded;thecarbonthencombineswiththetitaniumortheniobiuminpreferencetothechromium.Themostcommonlyknownmembersofthisfamilyaretheausteniticstainlesssteelsinwhichnickelisintroducedtokeeptheausteniticmicro-structureinplaceatroomtemperature.Theydonotrustorstainwhenusedfordomesticpurposessuchascooking,asdoesmildsteel,buttheyaresusceptibletosomeformsofcorrosion,forexamplewhenusedinanenvironmentcontainingchlorideionssuchaswatersystems.Theseausteniticstainlesssteelsareveryductilebutdonothavetheyieldpointcharacteristicofthecarbonsteelsandtheydonotexhibitastepchangeinfracturetoughnesswithtemperatureasdothecarbonsteels.Somevarietiesretaintheirstrengthtohighertemperaturesthanthecarbonsteels.Theferriticstainlesssteels Metals17containnonickelandsoarecheaper.Theyaresomewhatstrongerthanausteniticstainlesssteelsbutarenotsoreadilydeepdrawn.Proceduresfortheirweldingrequireparticularcaretoavoidinducingbrittleness.Thereisafurtherfamilyofthestainlesssteelsknownasduplexstainlesssteelswhichcontainamixtureofferriticandausteniticstructures.Theyarestrongerthantheausteniticstainlesssteels,andmoreresistanttostresscorrosioncrackingandarecommonlyusedinprocessplant.Metalsotherthanthesteelshavebetterpropertiesforcertainuses,e.g.copperandaluminiumhaveexceptionalthermalandelectricalconductivity.Usedextensivelyinaerospaceapplications,aluminiumandmagnesiumalloysareverylight;titaniumhasaparticularlygoodstrengthtoweightratiomaintainedtohighertemperaturesthanthealuminiumalloys.Nickelanditsalloys(somewithiron),includingsomeofthe`stainless'steels,canwithstandhightemperaturesandcorrosiveenvironmentsandareusedinfurnaces,gasturbinesandchemicalplant.Howevertheextractionofthesemetalsfromtheiroresrequirescomplicatedandcostlyprocessesbycomparisonwiththoseforironandtheyarenotaseasilyrecycled.Nootherseriesofalloyshastheallroundusefulnessandavailabilityofthecarbonsteels.Forstructuralusescarbon±manganesesteelshavealargelyunappre-ciatedfeatureintheirplasticbehaviour.Thisnotonlyfacilitatesasimplemethodoffabricationbycoldformingbutalsoofferstheopportunityofeconomicstructuraldesignthoughtheuseofthe`plastictheory'describedinChapter8.Whilstitmaynotbeafundamentaldrawbacktotheiruse,cognisancehastobetakenofthefracturetoughnesstransitionwithtemperatureincarbonsteels.2.1.6SteelqualityThecommercialeconomicsandpracticalityofmakingsteelsleadstoavarietyofqualitiesofsteel.Qualityasusedinthiscontextreferstofeatureswhichaffecttheweldabilityofthesteelthroughcompositionanduniformityofconsistencyandtheextenttowhichitisfreefromtypesofnon-metallicconstituents.Theordinarysteelmakingprocessesdeliveramixtureofsteelwithresiduesoftheprocesscomprisingnon-metallicslag.WhenthisiscastintoaningotthesteelsolidifiesfirstleavingacoreofmoltenslagwhicheventuallysolidifiesasthetemperatureoftheingotdropsasinFig.2.3.Obviouslythisslagisnotwantedandthetopoftheingotisburnedoff.Sincethesteelmakerdoesn'twanttodiscardanymoresteelthanhehasto,thiscuttingmayerronthesideofcaution,inthecostsense,sometimesleavingsomepiecesofslagstillhiddenintheingot.Whentheingotisfinallyforgedintoaslabandthenrolledthisslagwillbecomeeitherasinglelayerwithintheplate,alamination,ormaybreakupintosmall 18Weldeddesign±theoryandpractice2.3Formationofinclusionsinaplaterolledfromaningot.piecescalledinclusions.Forsomeusesofthesteeltheselaminationsorinclusionsmaybeofnosignificance.Forotherusessuchfeaturesmaybeundesirablebecausetheyrepresentpotentialweaknessesinthesteel,theymaygivedefectiveweldedjoints(Chapter11)ortheymayobscurethesteelorweldsonitfromeffectiveexaminationbyradiographyandultrasonics.Insomesteelmakingpracticesalloyingelementsmaybeaddedtothemolteningotbutiftheyarenotthoroughlymixedintheseelementsmaytendtostayinthecentreoftheingot,aplaterolledfromwhichwillhavealayeroftheseelementsconcentratedalongthemiddleoftheplatethickness.Suchsegregationmayalsooccurinsteelmadebythecontinuouscastingprocessinwhichinsteadofbeingpouredintoamouldtomakeaningotthesteelispassedthrougharectangularshapedapertureandprogressivelycooledasacontinuousbarorslab.Therearetechniquesformakingsteelmoreuniformbystirringbeforeitiscast;non-metallicsubstancescanbereducedbyre-meltingthesteelinavacuumorbyaddingelementswhichcombinewithnon-metallicinclusions,whicharemainlysulphides,tocausethemtohaveroundshapesratherthanremaininalamellarform.Suchtechniquesobviouslycostmoneyandthesteelmaker,asinothermatters,hastostrikeabalancebetweencostandperformance.Manyofthesesteelmakingimprovementswereintroducedinitiallyinthe1960sandwereextendedintheearly1970smainlyasaresultofthedemandsoftheNorthSeaoffshoreoilfieldsdevelopment.Asaresultthequalityofalargeproportionoftheworld'sstructuralsteelproductionimprovedmarkedlyandtheexpectationswerereflectedintheonshoreconstructionindustry.Otherdevelopmentsinsteelmakingpracticewereintroducedinthefollowingyearsaimedprincipallyatimprovingthestrengthwithoutdetractingfromtheweldabilityorconverselytoimproveweldabilitywithoutreducingthestrength.Thesedevelopmentswereinwhatwerecalledthethermomechanicaltreatmentofsteel.Basicallythiscomprisedthe Metals19rollingofthesteelthroughaseriesofstrictlycontrolledtemperaturerangeswhichmodifiedthegrainstructureinacontrolledway.Asaresultsteelsoffairlylowcarboncontent,`lean'steels,couldbemadewithastrengthwhichcouldformerlybeobtainedonlybyaddinglargeramountsofcarbon.Thesedevelopmentscreatedaconfidenceinthesupplyofconventionalstructuralsteelwhichbecamearelativelyconsistentandweldableproduct.Howeverthispositionwasnotuniversalandeveninthemid-1990ssteelwasstillbeingmadewithwhatwere,bythen,oldfashionedmethodsandwhoseconsistencydidnotalwaysmeetwhathadbecomecustomaryqualities.Certainlytheymetthestandardspecificationsincompositionbutthesestandardshadbeencompiledassumingthatmodernsteelmakingmethodsweretheonlyonesused.Inoneexampletheresultwasthatalthoughthesteelhadbeenanalysedbyconventionalsamplingmethodsanditscompositionshowntoconformtothestandard,thecompositionwasnotuniformthroughaplate.Virtuallyalltheironandcarbonwasontheoutsideoftheplatewiththede-oxidisingandalloyingelementsinabandinthemiddleplaneoftheplate.Anotherexamplehadbandsinwhichcarbonwasconcentratedwhichledtohydrogencrackingaftergascutting.Theconsequenceofthisisthatprecautionsstillhavetobetakenindesignandfabricationtopreventtheweaknessesofsteelfromdamagingtheintegrityoftheproduct.Themosteffectiveactionis,ofcourse,toensurethatthesteelspecificationrepresentswhatthejobneeds.Thequestionofcostorpriceisfrequentlyraisedbutalthoughsteelsofcertainspecificationgradesmaycostmoreitisnotbecausetheyareanydifferentfromtherunofthemillproduction,itisthatmoretesting,identificationanddocumentationisdemanded.2.1.7SteelspecificationsAnengineerwhowantssteelwhichcanbefabricatedinacertainwayandwhichwillperformtherequireddutyneedstoensurethathepreparesorcallsupaspecificationwhichwillmeethisrequirements.Moststandardstructuralsteelspecificationsrepresentwhatsteelmakerscanmakeandwanttosell;anythingbeyondthebasicproductandanyassurancelevelbeyondthatofthebasicstandardthenrequiresanappealto`options'inthestandard.Thesteel`grade'isonlyalabelforthecompositionofthesteelasseenbythesteelmakerandperhapstheweldingengineer.Itisnotanidentificationforthebenefitofthedesignerbecausethestrengthisinfluencedbythesubsequentprocessingsuchasrollingintoplatesorsections.Asaresultthesame`grade'ofsteelmayhavewidelydifferingstrengthsindifferentthicknessesbecauseinrollingsteelofthesamecompositiondowntosmallerthicknessesitsgrainstructureisalteredandthestrengthisincreased.Asanexampleatypicalstructuralsteelplate 20Weldeddesign±theoryandpractice2specificationcallsforaminimumyieldstrengthof235N/mmina16mm2thicknessbutonly175N/mmina200mmthickness,a25%differenceinstrength.Howeveritisnotunusualforsteelstohavepropertieswellinexcessofthespecifiedminimum,especiallyinthethinnerplates.Whilstthismaybesatisfactoryifstrengthistheonlydesigncriterion,suchsteelwillbeunsuitableforanystructurewhichreliesforitsperformanceonplastichingesorshakedown.Thesteelspecificationforthisapplicationmustshowthelimitsbetweenwhichtheyieldstrengthmustlie.Gradesmaybesub-dividedintosub-grades,sometimescalled`qualities'withdifferentfracturetoughnessproperties,usuallyexpressedasCharpytestresultsatvarioustemperatures.Further,standardspecificationsexisttoindicatethedegreeoffreedomfromlaminationsorinclusionsbyspecifyingtheareasofsuchfeatures,foundbyultrasonictesting,whichmaybeallowedinacertainareaoftheplate.2.1.8WeldmetalsWeldmetalisthemetalinaweldedjointwhichhasbeenmoltenintheweldingprocessandthensolidified.Itisusuallyamixtureofanyfillermetalandtheparentmetal,aswellasanyadditionsfromthefluxintheconsumables,andwillhaveanas-castmetallurgicalstructure.Thisstructurewillnotbeuniformbecauseitwillbedilutedwithmoreparentmetalinweldruns,orpasses,nearthefusionboundarythanawayfromit.Thiscaststructureandthethermalhistoryrequirestheconsumablemanufacturertodevisecompositionswhichwill,asfarasispossible,replicateormatchthepropertiesofthewroughtparentmetalbutinacastmetal.Thiscanmeanthatthecompositionoftheweldmetalcannotbethesameastheparentmetalwhichinsomeenvironmentscanpresentadifferentialcorrosionproblem.Aswellasstrengthanimportantpropertytodevelopintheweldmetalisductilityandnotchtoughness.Weldmetalscanbeobtainedtomatchthepropertiesofmostoftheparentmetalswithwhichtheyaretobeused.2.2AluminiumalloysAluminiumisthethirdmostcommonelementintheEarth'scrustaftersiliconandoxygen.Therangeofusesofaluminiumanditsalloysissurprisinglywideandincludescookingutensils,foodpackaging,beerkegs,heatexchangers,electricalcables,vehiclebodiesandshipandaircraftstructures.Purealuminiumissoft,resistanttomanyformsofcorrosion,agoodthermalandelectricalconductorandreadilywelded.Alloysofaluminiumvariouslywithzinc,magnesiumandcopperarestrongerandmoresuitableforstructuralpurposesthanthepuremetal.Ofthesealloys, Metals21threeseriesaresuitableforarcwelding;thosewithmagnesiumandsiliconandthosewithmagnesiumandzinccanbestrengthenedbyheattreatmentandthosewithmagnesiumandmanganesecanbestrengthenedbycoldworking.Weldingmayreducethestrengthintheregionoftheweldandinsomealloysthisstrengthisregainedbynaturalageing.Inothers,strengthcanberegainedbyaheattreatment,thefeasibilityofwhichwilldependonthesizeofthefabrication.Allowanceswhichhavetobemadeforthislossofstrengtharegivenindesignorapplicationstandards.Afourthseriesofalloys,aluminium±copperalloys,havegoodresistancetocrackpropagationandareusedmainlyforpartsofairframeswhichoperateusuallyintension.Insheetform,thisseriesisusuallycladwithathinlayerofpurealuminiumoneachsidetopreventgeneralcorrosion;ingreaterthicknesseswhichmaybemachinedtheyhavetobepaintedtoresistcorrosion.Thesealuminium±copperalloysareunsuitedtoarcweldingbuttherecentlydevelopedstirfrictionweldingprocessoffersaviableweldingmethod.Avaluablefeatureofaluminiumalloysistheirabilitytobeextrudedsothatcomplicatedsectionscanbeproducedwithsimpleandcheaptoolingwhichalsomakesshortrunsofasectioneconomical.ThereisaninternationalclassificationsystemforaluminiumalloyssummarisedinTable2.1.Thesystemusesgroupsoffourdigits,thefirstdigitgivingthemajorgroupingbasedontheprincipalalloyingelements;theotherdigitsrefertootherfeaturessuchascomposition.Additionalfiguresandlettersmaybeaddedtoindicateheattreatmentconditions.Thematerial3publishedbytheEuropeanAluminiumAssociationisanauthoritativesourceofknowledgeaboutaluminiumanditsalloys.3Table2.1SummaryofinternationalaluminiumalloyclassificationAlloygroupseriesMajoralloyingelementsPropertiesoruses1xxxNone99%Alcorrosionresistant2xxxCuHighstrength,aerospace3xxxMnSuitableforbrazing4xxxSiCastingsandfillerwire5xxxMgMediumstrength6xxxMg+SiHeattreatable7xxxZn+MgHighstrength,heattreatable8xxxOthere.g.Sc,Li,Fe 3Fabricationprocessses3.1OriginsThischapterdescribestheprincipalfeaturesoftheweldingprocessesappliedtothosematerialswhicharemostcommonlyusedinstructural,mechanicalandprocessplantengineeringnamelysteelsandaluminiumalloys.Tostartwithweneedtobeclearaboutwhatweldingisincontextofthisbook.Weldinghereisthejoiningoftwoormorepiecesofmetalsothatthepartstobejoinedmergewithoneanotherformingahomogeneouswholeacrosstheconnection.Thewordhomogeneousisusedguardedlyherebecausealthoughtotheeyeaweldmayappeartobehomogeneous,onamicroscopicscaleitmaycontainarangeofdifferentmetallurgicalstructuresandvariationsinthebasiccomposition.Itwillbeunderstoodthatthisdefinitionexcludessoldering,brazingandadhesivebondingbecausejointsmadewiththoseprocessesrelyforthebondonanintermediatelayerofasubstancetotallydifferentfromthatbeingjoined.Weldingametalrequirestheintroductionofenergywhichcanbeasheatdirectlyorinaformwhichwillconverttoheatwhereitisrequired.Theearliestweldingprocess,datingbackthousandsofyears,wasforgeweldingasappliedtowroughtironwherethepartstobejoinedareheatedinafiretoasoftstateandthenhammeredtogethersothatonemergeswiththeother.Thisisatraditionalblacksmith'sskillanditismostconvenientlyusedforjoiningthescarfedendsofbarsbutitwasusedinjoiningtheedgesofstripstomakegunbarrels(Chapter8).Themodernanalogueofthisweldingmethodisfrictionweldingwhichwillbereferredtolateron.Mostotherformsofweldinginvolvemeltingthepartswheretheyaretobejoinedsothattheyfusetogether.Thismeltingrequiresaheatsourcewhichcanbedirectedattheareaofthejointandmovedalongit.Suchsourcesaretheoxy-fuelgasflameandtheelectricarc.Theflameorthearccanbeusedtomeltthepartsonly(autogenouswelding)butitiscommontoaddfillermetalofthesamegeneralnatureasthemetalbeingjoined.Electricarcweldingemergedtowardstheendofthenineteenthcenturyandstill Fabricationprocessses23representsthebasisofalargeproportionofallweldingprocesses.Initially,in1881,anarcfromnon-consumablecarbonelectrodeswasusedbyAugustdeMeritensandwaspatentedbyBenardosandOlszewskiworkinginParis.Shortlyafterthat,in1888,aRussian,NGSlavianoff,usedaconsumablebaresteelrodasanelectrodeandheisgenerallyacceptedastheinventorofmetal-arcwelding.Barewireelectricarcweldingwasstillinindustrialusein1935andtheauthorsawitstillinusein1955foramateurcarbodyrestoration.TheSwede,OskarKjellberg,patentedtheuseoffusiblecoatingsonelectrodesinabout1910.Howeverweldingwasslowtobetakenupasanindustrialprocessinheavyindustryuntilthe1930swhenitbecameappliedonanindustrialscaletoships,buildingsandbridges.EventhentheadoptionofweldingwasnotwidelyaccepteduntiltheSecondWorldWargaveurgencytomanyapplications.Variationsonthearcweldingprocessblossomed,theindividualbareorcoveredrodbeingfollowedbycontinuouselectrodes,withandwithoutcoatings,whichofferedtheopportunityofmechanisation.Submergedarcweldingwasintroducedinthe1930sinboththeUSAandUSSRasanothermeansofcontinuousweldingwiththeaddedbenefitsofanenclosedarcandinwhichthefluxandwirecombinationcouldbevariedtosuittherequirementsofthework.Theprincipleofgasshieldedweldingwasproposedin1919withavarietyofgasesbeingconsidered.Inthe1930sattentionconcentratedontheinertgasesbutitwasnotuntil1940thatexperimentsbeganintheUSAusinghelium.Initiallydevelopedwithanon-consumabletungstenelectrodefortheweldingofaluminiumtheprinciplewastobeappliedtoacontinuousconsumableelectrodewirein1948.Thiseventuallyledtotheweldingofsteelsinthe1960sonaproductionbasisintheUSA,UKandUSSRbythedevelopmentoftechniquesforusingcarbondioxideasashieldinggasinplaceofthecostlyinertgases.Variationsonthistypeofweldingprocesscametobeusedintheformofwirewithacoreoffluxoralloyingmetalsandalsowireswithacoreofamaterialwhichgaveoffcarbondioxide,fluoridesormetalvapourstherebyavoidingtheneedforaseparategasshield.Intheearly1960sattentionwasturnedtotheuseofbeamsofenergyintheformofelectronsasaheatsourceforwelding.Theireffectiveuserequiredoperationinavacuumandequipmentandtechniquessoonfollowedwhichgavebenefitsinaccuracyandprecisionwithfreedomfromdistortionandwithmetallurgicalchangeslimitedtoanarrowbandoneithersideoftheweld.Waysofavoidingthedisadvantagesofinvacuoweldingbytechniquesusingpartialvacuumsarestillbeingdevelopedandnodoubtwillfindapplicationsinspecialisedmarkets.Theconstraintsofvacuumswereeventuallycircumscribedbytheadoptionofthelaserbeamasaheatsourcewiththeadditionalpropertiesofbeingabletobetransmittedaroundcornersandofbeingcapableofbeingsplit.Thelaserandelectronbeam 24Weldeddesign±theoryandpracticeprocessestodayexistascomplementarymethodseachbeingdevelopedfortheparticularfeatureswhichtheyoffer.Atthesametimeastheesoterichighenergydensitybeamprocesseswerebeingdevelopedattentionwasbeingpaidtothedevelopmentoffrictionwelding,afarmoremundaneandmechanicalbludgeonofaprocess.Oneofitsadvantagesisthatitdoesnotactuallymeltthemetalandsosomeofthemetallurgicaleffectsofarcweldingareavoided.Itrapidlygainedindustrialfavourasamassproductiontool,alsoinaversionknownasinertiawelding,inthemotorindustrybothinenginecomponentssuchasvalves,andtransmissionitemssuchasaxlecasings;today,variationsonthethemearestillbeinginventedandputtouse.Thelatestisfrictionstirweldingwhichamongstotheruseshasatlastofferedametallicjoiningprocesswithapotentialforweldingthealuminium±copperalloyscommonlyusedinairframesbecauseoftheirbenigncrackgrowthpropertiesandabsenceofstresscorrosioncrackingintheatmosphericenvironment.Anotherfamilyofweldingprocessesistheelectricalresistanceweldingprocesses;inthesethepartsareclampedtogetherbetweenelectrodeswhilstanelectriccurrentispassedthroughthem.Theelectricalresistanceofferedbytheinterfacebetweenthepartsconvertssomeoftheelectricalenergytoheatwhichmeltstheinterfaceandformsaweldnugget.Thisbasicprinciplefindsextensiveuseasspotweldinginsheetmetalfabricationincarbodies,whitegoodsandsimilarapplicationsandseamweldinginmorespecialisedfields.Trialsofresistancespotweldingoflargerthicknessesofstructuralsteels(*25mm)wereundertakeninFranceinthe1960sbutdidnotleadtoapracticalmethodoffabrication.Incontrastflashbuttwelding,anotherformofresistancewelding,wasextensivelyusedinarangeofthicknesseswhichamongstothersfoundapplicationinpipesandpipelines,particularlyintheformerSovietUnion.Thepartsareconnectedtoanelectricalpowersourceandbroughttogetherandpartedanumberoftimes,oneachoccasioncausinglocalarcingandmeltinguntilthewholeinterfaceisheatedatwhichpointthepartsareforcedtogethertomakethefinaljoint.Theprocessisalsousedforjoiningas-rolledlengthsofrailwaylines.On-sitejoiningofthelonglengthsoflinesomanufacturedcontinuestobeoneofthefewapplicationsofthethermitweldingprocess.Basicallyaninsituchemicalreactionbetweenaluminiumpowderandironoxide,itcastsapoolofmoltensteelinthejointwithouttherequirementforextraneouspowersupplies;itcanbeseenasanentertainmentbynightowlsincitiesallovertheworldwhichhavetramlines.Whilstmentioningthecastingofpoolsofmoltensteel,theelectroslagprocessisusedasameansofjoiningthicksectionsofstructuralsteelinonepassasin-linebutts,tee-buttsorcruciformjoints.Thiscanbefasterthanarcweldingandlessliabletogivedistortion;itcanbeperformedintheverticalpositiononlyalthoughitsapplicationcanbeextendedtootherpositionsby Fabricationprocessses25aversionknownasconsumableguidewelding.Variantsofthoseprocessesmentionedaboveandotherjoiningprocesseshavebeeninventedandeitherdiscardedalongthewayorlefttoserveasmallspecialisedmarket.Acynicmightseearcweldingasanextraordinarymeansbywhichtobejoiningmaterialsinthetwenty-firstcentury.Thematerialmanufacturerproducesametaltofinelimitsofcomposition,microstructureandproperties.Thenitissubjectedtoafiercearcsothatthemicrostructureandpropertiesofthemetaladjacenttotheweldarealteredbytherapidheatingandcooling.Theprocessgivesofftoxicfumeand,withtheopenarcprocesses,potentiallyinjuriousUVradiation.Theresultingjointiserraticinshape,pronetofatiguecracking,possiblydistortingthepartsandwithinternalstressesmuchlargerthananyprudentdesignerwouldthinkofusing.Arcweldinghasfollowedthepatternofotherinventionswhichseemtobequiteabominablebutwherethenewcomersneverseemtohavetherangeofapplicationsofthetraditionalones.Perhapsitisthatwegetusedtothem,andtheenergyneededbyhumanbeingstochangetheirhabitsandthemoney,timeandeffortinvestedinthetraditionalmethodspreventsordelaysothermeansfromemergingandthemselvesbeingdeveloped.Anotherexampleofsuchinventionsistheinternalcombustionpistonengineasusedinroadvehicles.Ithashundredsofmovingpartsbeingsentinonedirectiononemomentandreversedthenext,thousandsoftimesaminute,scrapingandhittingeachotherandwearingout.Itcan'tstartitself;itneedstobehandcrankedorturnedoverwithanelectricmotorwhichneedsahugebattery,muchlargerthanotherservicesrequire,andsoisjustdeadweightfortherestofthetime.Toallowtheenginetokeeprunningwhenittakesupthedriveithastohaveaslippingtransmission,eitherasolidfrictionorhydraulicclutch,whichwastesenergy.Theenginehassuchasmalleffectiveworkingspeedrangethatithastohaveatransmissionwhichhastobemanuallyormechanicallyreconfiguredinstepstokeeptheenginespeedwithintheworkingrange.Itsendsoutnoiseandtoxicgasesandparticlesandtheusedlubricatingoilispoisonousandenvironmentallydamagingunlessre-processed.ItsoundslikesomeEmmettcartoonmachine;wouldwereallystartfromthereifwehadtoinventanenginetoday?Nonethelesstakingthepragmaticviewwenowseehighlydevelopedarcweldingprocesseswhichcanmakereliablejointsgivingaperformanceconsistentwiththatoftheparentmetals.3.2Basicfeaturesofthecommonlyusedweldingprocesses3.2.1ManualmetalarcweldingThisprocessiswhatprobablycomestomostpeople'smindswhenarc 26Weldeddesign±theoryandpractice3.1Manualmetalarcweldingwithacoveredelectrode(photographbycourtesyofTWI).weldingismentioned.Thewelderholdsinaclamp,orholder,alengthofsteelwire,coatedwithafluxconsistingofminerals,calledaweldingelectrodeorrod;theholderisconnectedtoonepoleofanelectricitysupply.Themetalparttobeweldedisconnectedtotheotherpoleofthesupplyandasthewelderbringsthetipoftherodclosetoitanarcstartsbetweenthem(Fig.3.1).Thearcmeltsthepartlocallyaswellasmeltingofftheendoftherod.Themoltenendoftherodisprojectedacrossthearcinastreamofdropletsbymagneto-electricforces.Iftheweldermovestherodalongthesurfaceofthepartkeepingitsendthesamedistancefromthesurfacealineofmetalwillbedepositedwhichisfusedwiththemoltensurfaceofthepart,formingweldmetal,andwillcoolandsolidifyrapidlyasthearcmoveson.Thefluxcoatingoftheelectrodemeltsintheheatofthearcandvaporisessogivinganatmosphereinwhichthearcremainsstableandinwhichthemoltenmetalisprotectedfromtheairwhichcouldoxidiseit;thefluxalsotakespartinmetallurgicalrefiningactionsintheweldpool.Sometypesoffluxalsocontainironorotherelementswhichmeltintotheweldmetaltoproducetherequiredcompositionandproperties.Rodsformanualmetalarcweldingaremadeinavarietyofdiameterstypicallyfrom2.5mmto10mminlengthsrangingbetween200mmand450mm.Therearemanydifferenttypesofelectrodes,evenforthecarbon±manganesesteelfamily.Themaindifferencesbetweenthemlieinthefluxcoating.Therearethreemaingroupsofcoatingintheelectrodesusedinmostconventionalfabrications. Fabricationprocessses27.Rutilecoatingsincludeahighproportionoftitaniumoxide.Rodswiththistypeofcoatingarerelativelyeasytouseandmightbecalledgeneralpurposerodsforjobswhereclosecontrolofmechanicalpropertiesisnotrequired.Thesteelsonwhichtheyareusedshouldhavegoodweldability.Inpracticethismeansmildsteel..Basiccoatingscontainlime(calciumcarbonate)andfluorspar(calciumfluoride).Theyproduceweldmetalforworkwherehigherstrengththanmildsteelisrequiredandwherefracturetoughnesshastobecontrolled.Theyareusedwherethelevelofhydrogenhastobecontrolledasinthecaseofmorehardenablesteelstopreventheataffectedzonehydrogencracking.Rodswiththistypeofcoatingaremoredifficulttousethanthosewithrutilecoatings,thearcismoredifficulttocontrolandanevenweldsurfaceprofilemoredifficulttoproduce.Theneedforlowhydrogenlevelsmeansthattheymaybesoldinhermeticallysealedpacks;ifnot,theymustbebakedinanovenataspecifiedtemperatureandtimeandthenkeptinheatedcontainers,orquivers,untileachistakenforimmediateuse..Cellulosiccoatingshaveahighproportionofcombustibleorganicmaterialsinthemtoproduceafiercepenetratingarcandareoftenusedintherootruninpipelinewelding,`stovepipewelding'asitiscalled,andforthecappingrun.Thehighquantitiesofhydrogenwhicharereleasedfromthecoatingrequirethatprecautionsbetakentopreventhydrogencrackinginthesteelafterwelding.Rutileandbasiccoatedrodsmayhaveironpowderaddedtothecoating.Thisincreasesproductivitybyproducingmoreweldmetalforthesamesizeofcorewire.Thelargerweldpoolwhichiscreatedmeansthatironpowderrodscannotbeasreadilyusedinallpositionsastheplainrod.Coveredelectrodesarealsoavailableforweldingstainlesssteelsandnickelalloysbutareproportionatelylesspopularthanforcarbonsteels;muchoftheworkonthesealloysisdonewithgasshieldedwelding.Theelectricalpowersourceforthistypeofweldingcanbeatransformerworkingoffthemainsoranenginedrivengeneratorforsitework.ThesupplycanbeACorDCdependingonthetypeofrodandlocalpractice.3.2.2SubmergedarcweldingThisprocessusesacontinuousbarewireelectrodeandaseparatefluxaddedoverthejointseparatelyintheformofgranulesorpowder.Thearciscompletelyenclosedbythefluxsothatahighcurrentcanbeusedwithouttheriskofairentrainmentorseverespatterbutotherwisethefluxperformsthesamefunctionsasthefluxinmanualmetalarcwelding(Fig.3.2).Athighcurrentstheweldpoolhasadeeppenetrationintotheparentmetaland 28Weldeddesign±theoryandpractice3.2Submergedarcwelding(photographbycourtesyofTWI).thickersectionscanbeweldedwithoutedgepreparationthanwithmanualmetalarcwelding.Lowercurrentscanofcoursebeusedandwiththeabilitytovaryweldingspeedaswellasthefluxandwirecombinationstheweldingengineercanachieveanyrequiredweldedjointproperties.Theprocesshasthesafetybenefitoftherenotbeingacontinuouslyvisiblearc.Theprocessismostcommonlyusedinamechanisedsystemfeedingacontinuouslengthofwirefromacoilonatractorunitwhichcarriestheweldingheadalongthejointoronafixedheadwiththeworktraversedorrotatedunderit.Whenweldingsteelsaweldingheadmayfeedseveralwires,onebehindanother.BothACorDCcanbeusedandwithamulti-headunitDCandACmaybeusedonthedifferentwires;DContheleadingwirewillgivedeeppenetrationandAContheotherwireswillprovideahighweldmetaldepositionrate.Weldingcurrentsofupto1000Aperwirecanbeused.Manuallyoperatedversionsofsubmergedarcweldingareusedinwhichthecurrentlevelsarelimitedtosome400A.Thefluxesusedinsubmergedarcweldingofsteelscanbeclassifiedbytheirmethodofmanufactureandtheirchemicalcharacteristics.Theymaybemadebymeltingtheirconstituentstogetherandthengrindingthesolidifiedmixwhenithascooled,orbybondingtheconstituentstogether Fabricationprocessses29intogranularform.Thechemicalcharacteristicsrangefromtheacidtypescontainingmanganeseorcalciumsilicatestogetherwithsilicatothebasictypes,againcontainingcalciumsilicatesusuallywithalumina,butwithalowerproportionofsilicathantheacidtypes.Theacidfluxesareusedforgeneralpurposeworkwhereasthebasicfluxesareusedforweldsrequiringcontroloffracturetoughnessandforsteelsofhighhardenabilitytoavoidhydrogencracking.Thewireisusuallyofa0.1%carbonsteelwithamanganesecontentofbetween0.5%and2%witharelativelylowsiliconcontentaround0.2%.Asamechanicalprocess,submergedarcweldingiscapableofgreaterconsistencyandproductivitythanmanualweldingalthoughtobalancethistheprocessisnotsuitedtoareasofdifficultaccessandmulti-positionworkinsitu.3.2.3Gasshieldedwelding3.2.3.1ConsumableelectrodesHereabarewireelectrodeisused,aswithsubmergedarc,butagasisfedaroundthearcandtheweldpool(Fig.3.3).Asdoesthefluxinthemanualmetalarcandsubmergedarcprocessesthisgaspreventscontaminationofthewireandweldpoolbyairandprovidesanatmosphereinwhichastablearcwilloperate.Thegasusedisoneoftheinertgases,heliumorargon,fornon-ferrousmetalssuchasaluminium,titaniumandnickelalloys,whentheprocessiscalledmetalinertgas(MIG).Forcarbonsteelspurecarbondioxide(CO2)oramixtureofitwithargonisusedwhentheprocessiscalledmetalactivegas(MAG).Thefunctionsofthefluxintheotherprocesseshavetobeimplementedthroughtheuseofawirecontainingde-oxidisingelements,about1%manganeseand1%silicon.Thesecombinewiththe`active',i.e.theoxygen,partoftheshieldinggasandprotectthemoltensteelfromchemicalreactionswhichwouldcauseporosityintheweld.Forstainlesssteelsamixtureofargonandoxygenmaybeused.Therangeofcurrentswhichcanbeusedcoversthatofboththemanualmetalarcandthelowerrangesofthesubmergedarcprocesses.Thewireisfedfromacoiltoaweldingheadorgunwhichmaybehandheldormountedonamechanisedsystem.Thewiremaybesolidoritmayhaveacorecontainingafluxormetalpowderwhichgivestheabilitytovarytheweldmetalpropertiesbychoiceofthewire.Theneedforgasandwirefeedconduitsand,inthecaseofhighercurrents,coolingwatertubes,canmaketheprocessrathermorecumbersometousethanmanualmetalarcandrestrictsitsapplicationinsitework.Thevariationoftheprocess,selfshieldedwelding,inwhichthecoreisfilledwithachemicalwhichemitsshieldingvapoursonheatingeliminatestheneedforagassupplyandisused 30Weldeddesign±theoryandpractice3.3Gasshieldedwelding(photographbycourtesyofTWI).satisfactorilyonsite.Thesolidwiregasshieldedprocesshastheadvantageinproductionworkoverthefluxprocessesinthattheweldsdonotneedasmuchde-slagging,butsmall`islands'ofsilicatesmayremainontheweldsurfaceandhavetoberemovedifapaintsystemistobeapplied.Afluxprocesswithaselfreleasingslagwillhavetheadvantageoversolidwirewheretheweldhastobebrushed.DCisusedinoneoftwomodes.Atlowcurrentsthetransferofmetalfromthewiretotheweldpooltakesplaceaftershortcircuitsasthetipofthewireintermittentlytouchestheweldpool.Thisiscalleddiptransfer.Athighcurrentsthetransferisbyastreamofdropletspropelledacrossthearcandtermedspraytransfer.Thediptransfermodeisusedforsheetmetalwork,rootrunsandforpositionalwork,i.e.overheadorverticalwelds.Exceptwithrutilefluxcoredwires,thespraytransfermodeisunsuitedtopositionalweldingandisusedfordownhandfillingrunsinthickermaterialwherethegreaterdepositionratecanbeemployedwithadvantage.Awidercontrolofmetaltransfercanbeachievedbypulsing Fabricationprocessses31theweldingcurrentusingaspecialpurposepowersource.Thispermitsawiderrangeofconditionsforpositionalweldingbutcannotbeusedwithpurecarbondioxideasashieldinggas.Itisrestrictedtoweldingwithargon±CO2±oxygenmixtures.3.2.3.2Non-consumableelectrodesForthinsheetworkandprecisionweldingofcomponentstoclosetolerancesthetungsteninertgas(TIG)processcanbeused.Thearcisstruckbetweenatungstenelectrodeandtheworkpiecewithargonorheliumastheshieldinggas.Thetungstenelectrodeisnotconsumedandfillercanbeaddedtotheweldasawirealthoughmanyapplicationsemployajointdesigninwhichafillerisnotrequired(autogenouswelding)(Fig.3.4).ACisusedforaluminiumalloysandDCforferrousmaterials.TheTIGprocesscanbeusedmanuallyormechanised.Aprocesswithsimilarapplicationsatlowcurrentsisthemicroplasmaprocess.AjetofplasmaisproducedinatorchwhichlookssimilarexternallytoaTIGtorch.Itcanbeusedforveryfineworkonavarietyofmetals.Theplasmaprocessusedathighcurrents,e.g.400A,canbeusedforbuttwelding;themechanismhereisdifferentfromTIGandmicroplasma.Theplasmajetmeltsthroughthemetalandformsaholeintheshapeofakeyhole;asthetorchmovesalongthejointthemetalre-solidifiesbehindthekeyholesoastofusethetwoparts.Theprocessis3.4Tungsteninertgaswelding(photographbycourtesyofTWI). 32Weldeddesign±theoryandpracticeusedinamechanisedformforweldingstainlesssteelandaluminiumalloys,andisparticularlysuitedtopipeandtubularshapesinwhichthejointcanberotatedunderafixedweldinghead.3.3CuttingStructuralsteelsareusuallygascutalthoughlasercuttingisincreasinglyusedforplate.Ingascuttingaflameoffuelgassuchasacetyleneburninginoxygenheatstheareatobecut;astreamofoxygenistheninjectedaroundtheflamewhichactuallyburnsthesteelandejectstheoxideasdross.Thecuttingtorchmaybehandheldoritmaybemountedonamechanisedcarriage.Dependingonthethicknessthesteelhastobepre-heatedasforweldingtopreventahardheataffectedzonebeingformedonthecutedgewiththeattendantriskofcracking.Acuttingprocedurespecificationcanbepreparedandtestedinamanneranalogoustoaweldingprocedurespecification.Mechanisedcuttingispreferredasitcanproduceasmootheredgethanmanualcutting;theburnerscanbetraversedintwodirectionstocutshapesorholes.Numbersofcuttingheadscanbeusedsimultaneouslysothatmanycopiesofthesameshapecanbecut.Itgoeswithoutsayingthatcomputercontrolcanbeappliedasafirstphaseofacomputeraidedmanufacturingsystem.Thecuttingheadcanbesetatananglesothatabevellededgecanbecutasaweldedgepreparation.Two,oreventhree,headscanbemountedasshowninChapter4sothatadoublebevelwitharootfacecanbecutinonepass.Aproperlyadjustedgascutterwillleaveasmoothedgealthoughinclusionsorlaminationsinsteelplatescanblowoutgasesleavingalocalroughnessinthecut.Thecutmaycarryaglazeofsilicatesfromthesteelwhichmaypreventpaintadheringtothesurface.Forthisreasonitisusualtogrindorgritblastthesurfaceifitistobepainted.Thinsheetandplatemetals(<12mm)canbecutbyguillotineandholespunched.Inbothcasestheshearededgeisseverelycoldstrainedandinsomecarbonsteelsmayofferpoorfracturetoughnessifwelded;aweldingproceduretestshouldbeabletoclarifythispoint.Stainlesssteelsdonotcutwellbyburningandarefrequentlyplasmacut.Mechanisedversionsofplasmacuttingequipmentcanmakethecutunderwaterwhichgivesaverycleancutwithlittledistortion.Carbonandstainlesssteelscanbecutveryaccuratelywithlasersuptothicknessesof20mm.3.4BendingSteelscanbebentcold,i.e.atambienttemperature,althoughtheyhavetobeformulatedtobeabletodothisuniformlyincomplexshapessuchascarbodies.Thesteelsusedforcoldformedbuildingshapessuchashollowsections,purlins,railsandfloorbeamsaresimilartothecarbon±manganese Fabricationprocessses33steelsusedinrolledsections;perhapstheoldestexampleofacoldformedsectionisthewellknowncorrugatedirontobefoundallovertheworld.Thereisalimittothethicknessofsteelwhichcanbecoldformed,partlyowingtothemechanicalforcerequiredandalsobecauseaboveacertainthicknesstheplasticstrainingcanleaveamaterialwhichcanbe,particularlyonwelding,ofareducedfracturetoughness.Thisfracturetoughnesscanberecoveredbyaheattreatmentsimilartothermalstressrelieving.Thealternativeistobendorrollthesteelatahightemperaturesuchasthatusedforstressrelief.Thereisnosuitableintermediatetemperatureforformingbecausebrittlenesswillbeinduced.Stainlesssteelscanbeformedcoldbuttheyhavealargerspringbackthancarbonsteels.Aluminiumanditsalloyscanbefoldedorformedinvariouswaysandhaveaparticularadvantageinthattheycanbeextrudedthroughasimpledietogenerateanalmostinfinitevarietyofprofiles.3.5ResidualstressesanddistortionTheprogressionofthearcalongthejointrepresentsamovingheatsourceformingapoolofmoltenmetalaroundwhichisacomplicatedandchangingdistributionoftemperatureandstrainasthemetalisprogressivelyheated,melted,frozenandcooled.Inamulti-runjointthecomplexityiscompoundedbythepresenceoftheearlierruns.Onecanillustratetheformationofresidualstressesbyasimplifiedmodelofabuttweldwhichignorestheprogressiveaspectofwelding.Fig.3.5showstwoplatesandbetweenthemahotstripofmetalrepresentingtheweldarea.Iftheweldwerefreetocontractlengthwiseoncoolingitwouldendupbeingshorter3.5Originsofresidualstressesinweldedjoints.thanitshotlength.Bybeingfusedtotheplatesitisrestrainedbythemandsoisputintotension;converselytheplateisputintocompression.ThedistributionofstressismorecomplicatedthanthesimplemodelwouldshowandmeasurementssuggestthatitwouldbeofthegeneralformasinFig.3.6.InthedirectiontransversetotheweldthesimplemodelwouldnotproduceanyresidualstressexceptthatfromthePoissoneffect.Inpracticeofcourse 34Weldeddesign±theoryandpracticeStressparalleltoweldlengthStresstransversetoweldlength3.6Typicaldistributionofresidualstressesinabuttweldedplate.theprogressivenatureoftheweldmeansthatthestartiscoolandabletoreactforcesquitesoonaftertheheatsourcemovesaway;theresultofthisisthattransverseresidualstresssystemisestablishedalsoshowninFig.3.6.Residualstressesarethemainsourceofdistortion.Theyrepresentaself-equilibratingstresssystem,thatisasetofstresseswhicharereactedentirelywithintheiteminquestion.Inthesimplestcaseabarmadefromtwopiecesweldedendtoendwillshrinkfromitshotlengthwhenitcoolsdownastheweldandadjacentmetalcontract.Aweldonthesurfaceofaplatewillmakeitbendoutofitsplaneandanitemweldedtoabeamflangewilltendtobendthebeambythecontractionoftheweldandadjacentmetal.Evenunweldedrolledsectionscontainresidualstressesbecausethedifferentthicknessescoolatdifferentratesafterthesectionisrolled.TheeffectscanbeseenifanIsectionisslittomaketwoTsectionswhentheywillbowasaresultoftheresidualstresseswhichwerebalancedwithinthecompletesection.Whenacoverplateistobeweldedtoabeamflange(Fig.3.7)it3.7Reductionofdistortionbyweldingintermittentpadsinsteadofcontinuousplate.mustbeweldedsymmetricallyotherwisethebeamwillcurveintheplaneofthecoverplate.Evensoitmaycausethebeamtocurveintheotherplane.Thisdistortioncanbeminimisedbystitchwelding,thatisbymakingrelativelyshortlengthsofweldatintervalsandfillinginbetweenthem.Ifthecoverplatedoesnotreallyneedtobecontinuousthedistortioncanbeminimisedbycuttingitintoshorterlengths.Thisapproachisparticularlysuitablewhentheplateisreallyjustamountingblockforequipment. Fabricationprocessses35Residualstressesanddistortioncanaffecttheloadbearingcapacityofastructurebyreducingthebucklingstrengthofamember;thisisdealtwithinChapter8.3.6PostweldheattreatmentSteelfabricationsmaybesubjectedtopostweldheattreatment(PWHT).Thisusuallyreferstoatreatmentinanovenorwithexternalelectricalorchemicalheatinginwhichthefabricationisheatedtobetween5808and6208C,abrightredheat,forabout1hforeach25mmofthickness.Themostcommonreasonfordoingthisistorelaxtheresidualstressessetupbywelding.Thishastwopotentialbenefits:oneistostabilisethefabricationagainstdistortioninmachiningorserviceandtheotheristodecreasetheriskofbrittlefracture.Thelattereffectisalsoenhancedbytheeffectoftheheatingonthemicrostructurelocaltoanypotentialfractureinitiationsites.Theheattreatablealuminiumalloysmaybeheattreatedtoartificiallyagetheweldsandregainthestrengthoftheparentmetal. 4Considerationsindesigningaweldedjoint4.1JointsandweldsItisconvenienttodefineajointseparatelyfromaweld.Thejointisthemannerinwhichthepartsmeeteachother,e.g.buttjoint,lapjoint,Tjoint,cornerjoint,asshowninFig.4.1.Abuttjointiswheretwopartsbuttagainsteachotherendtoendoredgetoedge,whereasalapjointiswherethetwopartsoverlap.ATjointissocalledbecausetheparts,iftheyareofsimpleshapesuchasflatplate,meetintheformofaT.Acornerjointissocalled4.1Jointforms. Considerationsindesigningaweldedjoint374.2Basicweldtypes.becauseitformsanangle,orcorner,wherethetwopartsmeet.Weldsusedtomakethesejointswitharcweldingcanconvenientlybethoughtofasofonlytwomaintypes,thebuttweldandthefilletweld,asinFig.4.2.Thebuttweldjoinstwopiecesbyfusingtheircompletecrosssectionssocreatingamonolithicobjectwhereasthefilletweldconnectsthetwopartswithalineofweldmetalwithoutattemptingtocreateafullsectionjoint.Evensotheseweldnamesaresomewhatarbitrary±theweldmetaldoesn'tknowwhetheritisinabuttweldorafilletweld.Thebuttweldedjointhasapotentialforahigherperformancethanajointmadewithafilletweldbutitcanbemorecostlytomake.Thebuttweldiscapableofbeingexaminedforinternalsoundnesstogiveconfidencethattheweldwillperformasrequired.Thefilletweldgivesalowerperformance,cheaper,jointwhichwillstillprovidealoadcarryingconnectionbutwhichcannotbeexaminedforinternalsoundnessasreadilyasabuttweld.Thissuggeststhatingeneraloneshouldnotplaceashighalevelofconfidenceinthefilletweld'sperformanceasinthatofabuttweld;thismeansthatthesetwotypesofweldarenotjustdifferentinform,theyalsorepresenttwodifferentengineeringphilosophies.Requirementsforwelderqualificationsinmanyfabricationcontractswouldsuggestthatthemanualskillrequiredtomakeabuttweldisgreaterthanthatforafilletweldwhichplacesapremiumonthesupplyofweldersformanualbuttwelds.Itisdifficulttounderstandwhythisisheldtobeso;thetechniqueforachievingfullrootfusioninafilletweldisverydemandingandtherehavebeenoccasionswhenqualifiedbuttweldershavefailedafilletweldtestbecauseofthisveryfeature.Logicwouldalsoperhapssuggestthatiffilletweldscannotbenondestructivelyexaminedaseffectivelyasbuttweldsthenafilletwelderneedstobemoreskilledthanabuttwelder.Aswithmanyaspectsofweldingthisattitudehasprobablygrownoutoftraditionandnotoutoflogic.Thebasisforthiswouldbethatsincebuttweldsaregenerallyusedinhighintegrityapplicationsandfilletweldsareusedinlowintegrityapplicationsthenweldersqualifyingforfilletweldingneedtobelessskilledthanthoseforbuttwelds.Justtoconfusetheissue 38Weldeddesign±theoryandpractice4.3Tjointmadewithbuttandfilletwelds.furthertherecanbebuttweldswithpartialpenetration(NB:lackofpenetrationisadefect,seeChapter11)andtypesofjointmadewithacombinationofbuttandfilletwelds,asinFig.4.3.Whenwenominateweldingasthejoiningmethodwehavetochooseamaterialwhich,whenwelded,willperformasrequiredinservice.Thetypeofjointwecanuseisinfluenced,orevendefined,bythenatureoftheobjectofwhichitispart.Thechoiceoftypeofweldisthenlimitedtooneofthefewwhichwillsatisfythedemandsonthejointbothintermsofserviceperformanceandaccessibilityforweldingandinspection.Thelatterwilldependonthechosenweldingprocess;inpracticethechoiceisnarrowandmosthavetomakedowiththefewprocessesactuallyavailable.Takeasanexampleasimplejointconnectingtheedgesoftwosteelplatesofthesamecomposition,ofequalthicknessandinthesameplane.Forreasonsofstructuralperformancewemightoptforafullpenetrationweld.Theoreticallythechoiceofweldingprocessesisgreat,rangingfrommanualmetalarcweldingwithcoatedelectrodestoelectronbeamandlaser;wemightthinkalsoofelectroslag,diffusionbonding,frictionweldingandflashbuttwelding.Muchdependsontheindustryinwhichweareworking,itstraditions,itsexpectationsanditsmanufacturingsophisticationintermsofmaterials,dimensionaltolerances,surfacefinishandcleanliness.Wehavetorecognisetherestraintofcost,thesizeandshapeofthefabricationandwhetherornotitistobemassproducedoraoneoff.Unlesswehaveareallypressingcaseforahightechandexpensiveweldingprocesswewillendupwithoneofthemoremundaneprocesses.Thechoicewillbefurtherwhittleddowntothefacilitiesofthefabricator.Mostworkstillendsupbeingdonewithsimplearcwelding.Thereareanumberofotherfactorswhichwillinfluencethechoiceofthejointandweld.Amostimportantoneisthatoffeasibilityofinspection,fordespitethebestofintentionstheidealofon-lineprocesscontrolbasedonthequalitiesoftheweldbeingmadestillevadesmuchmechanisedweldingandofcoursehasnoroleinmanualwelding.Wethereforestillfindalotofworkbeinginspectedaftercompletionbyvariousmeansrangingfromvisualsurfaceexamination,assistedvisualexaminationsuchasmagneticparticleanddyepenetrant,toradiographyandultrasonics(methodsdescribedinChapter11)andtherelativelymoreesotericbutwellestablishedtechniquessuchaseddycurrentsandultrasonicimaging.Allofthesetechniquesaimtodiscoverphysicaldiscontinuitiesinthejointonamacroscalesuchasare Considerationsindesigningaweldedjoint39representedbylackoffusion,cracks,porosity,inclusionsandlaminations.Themethodsallrelyondetectingtheboundarybetweensolidmetalandcavitiesandtheirsuccesspresupposesthatnosuchcavitiesareintendedtobetheresuchasinpartialpenetrationbuttweldsandfilletweldedjoints.Thismeansthatiffullconfidenceinsuchinspectionisrequiredwehavetouseafullpenetrationbuttweld;inadditionwehavetobesurethattheinternalstructureofthesteeldoesnotitselfcontaincavitiesorinclusionsonamacroormicroscalesodistributedthattheywillconfusetheinspectiontechnique.Therearetechniquesforinternalexaminationoffilletweldsbuttheseareratherspecialisedandnotincommonuse.Allofthesetechniqueshavetheirindividualrequirementsforaccesswhichhavetobetakenintoaccountwhendesigningthejointandweld.Itisnotsurprisingthattherecanbeconflictbetweentheseconsiderationsandasinmanyotherwalksoflifethedesigneroftheweldedjointhastomakecompromises.Thenecessityandscopeforcompromiseisraisedinotherchaptersofthisbookfromwhichitwillbecomeapparentthatasinotherfieldsofengineeringdesignthereisnounique`correct'solutionalthoughtheremaybeabestormostexpedientsolutionforaparticularsetofcircumstances.Table4.1listsmanyoftheconsiderationsindesigningaweldedjoint.4.2TerminologyThissectionmightwellbeentitled`Communication'foritisaboutthemeansbywhichinstructionsareconveyedbetweenpeople.Spokenandwrittenlanguageisvitaltomosthumanendeavouranditsmasteryeludesmostofus.BecauseofthehistoryoftheBritishIslesoverthepastthreethousandyearstheEnglishlanguageisderivedfromthelanguagesoftheCelts,theRomans,theAngles,theSaxons,theVikingsandtheNormans,whowerethemselvesofVikingoriginbutovercenturieshadadoptedtheFrenchculture.ThisgivestheEnglishlanguagetheabilitytorepresentobjectsorideasinmanywaysandwithmorenuancesthanmany.Eveninthetwenty-firstcenturysomewordshavenottravelledfarfromthelocalityinwhichtheiroriginaluserssettled.Forconvenienceandfromfrequencyofuseeverytradeandprofessiondevelopsanarrowerinterpretationofsomeofthecommonwordsandinventssomeofitsown.TheabsorptionofthevariouslanguagesintotheBritishIslesoverthefirstmillenniumAD,anduptothetimeoftheNormans,producedaconglomeratelanguagewhichbecameknownasEnglish,mostfamouslyusedbyChaucerinthefourteenthcenturyandbyShakespeareinthesixteenthandseventeenthcenturies.TheirwritingsbecameavailabletomorethanthesmallgroupofeducatedpeoplethroughtheprintingpressinventedinthefifteenthcenturybyGutenberginGermany.Thesecondhalfofthe 40Weldeddesign±theoryandpracticeTable4.1.ConsiderationsindesigningaweldedjointFeatureExamplesofmattersforconsiderationServiceperformancestaticstrengthductilityfatiguelifecorrosionresistanceMaterialweldabilityas-weldedstrength"ductility"fracturetoughnesschemicalcompositionsusceptibilitytocrackingWeldingconsumablesmatchingparentmetalpropertiesWeldingprocessweldtypeaccessmaterialsizeofcomponentcostshoporsiteDistortionweldpreparationsingleordoublesidedweldheatinputweldrunsequenceAccessforweldingpositionconfigurationreachobstructionshoporsiteAccessforinspectionandNDTasforweldingCostbuildingandplantcharges±e.g.interestoncapital,depreciation,leasingcharges,maintenanceconsumables,materials,energypayrollcostsoverheadstaxesWeldqualitystandardNDTjointandweldconfigurationmethodsPositionofjointinfabricationservicestresssizeofcomponentshoporsiteworktransportaccessforwelding/inspection Considerationsindesigningaweldedjoint41secondmillenniumADsawthislanguagemovedtootherpartsoftheworldbyBritonswhosettledinothercontinents.Wordsusedinallwalksoflifemayremainunchangedovercenturiesinonecountrywhilstchanginginanother;forexample,someusageofEnglishwordscommonintheUSAisnowseenasoldfashionedintheUK.TheirmeaningswerecommoninBritainintheseventeenthandeighteenthcenturieswhentheearlysettlerscrossedtheAtlanticbuthavesincepassedintodisuseintheircountryoforigin.AsimilarpositionexistswithFrenchasspokeninCanada.AnexampleintheEnglishusedintheUSAisa`chapter'inthesenseofagroupofpeoplebelongingtoalargerorganisationwhichintheUKisnowcalledabranchexceptinsomechurcheswhicharemoreresistanttochangethanmost.InAustralia`manchester'isawordappliedtocottongoodsbecauseatthetimeoftheBritishsettlementofAustraliaManchesterwasthecityatthecentreofthecottontradeinBritainandtheworld;thewordhassincepassedintodisuseintheUK.WeshouldnotbesurprisedtheniftheEnglishlanguageterminologyusedinweldingandweldedjointscanvaryevenwithinonecountryandthetermsusedforthesamethingmaydifferevenbetweenindustriesinacountryorbetweendifferentgroupsofpeopleinthesameindustry.Weldingrodisatermwellrecognisedontheshopfloorandweldingelectrodeislesscolloquialwhilsttheformalwrittentermmightbecoveredelectrodewhichfewinindustrywoulduseinspeech.Forthesakeofclarityinconveyinginstructionsmostcountriesestablishaformalterminologybyissuingstandardvocabularies.Alongsidetheseareinternationaldictionarieswhichoffertheequivalentwordsinanumberoflanguages.TheterminologygiveninthischapterisbasedonthatcommonlyusedintheUKwhichispublishedinBS499.Inengineeringmuchoftheinstructionisconveyedintheformofdrawingsinwhich,forsimplicity,symbolsareusedinplaceoftext.Thishelpstoavoidambiguityandininternationaltradealsoavoidsthepotentialproblemsassociatedwithhavingtotranslatetext.Nonethelesstherecomesapointwhenasymbolicrepresentationmaybecometoofussyandconfusedatwhichtimethedraughtsmanmayresorttodetailscrapviews.Therearestandardsatalllevelsgivingsymbolsforuseondrawingsrelatingtothewrittenterms;theinternationallevelisrepresentedbyISO2553.Figures4.1±4.4showthebasicjointsandweldterminology.Thereareafewmoretermsincommonusewhichareneededtodefineaweld.Therearenational,4±7regionalandinternationalstandardswhichgiveterminology. 42Weldeddesign±theoryandpracticeFaceToeRootMitreConvexConcaveDeeppenFaceThroat|||<Ð>|RootleglengthFILLETWELD4.4Commonlyusedweldterminology.4.3Weldpreparations4.3.1In-linebuttjointsWiththeweldingconditionsforrutileandbasiclowhydrogenelectrodesusedformostmanualmetalarcweldingthereisverylittlepenetrationatall.Itfollowsthenthatwhenabuttweldistobemadebetweentheedgesoftheplatetheyhavetobebevelledsothattheweldmetalcanbeplacedinthejointandfusedwiththeparentmetal.Cellulosiccoatedelectrodesgiveamorewidelypenetratingarc(Fig.4.5)andareusedforrootrunsinsomestructuralsteelworkapplicationsbutmorecommonlyinpipelinecircumfer-entialweldsmadeonsiteusingatechniquecalledstovepipewelding.Theseelectrodesreleaseahigherlevelofhydrogenthantheothertwotypesandtheweldingprocedureshavetobedesignedtorecognisethissoastoavoidheataffectedzonecracking.WithmechanisedMAGorsubmergedarcweldingequipmentafullpenetrationbuttweldcanbemadefromonesidewithoutedgebevelsiftheweldingcurrentishighenough.Howeverthisrequiresedgeshavingaclosefitallthewayalongthejointandthattheweldingconditionsarepreviously Considerationsindesigningaweldedjoint434.5Penetrationofweldbeadon10mmsteelplate;4mmcoveredelectrodesat167A.Coatingtypesfromlefttoright:cellulosic,basic,rutile.provenandcontrolledduringtheweldingofthejoint.Ifthesemattersarenotattendedtothearcwilleitherfailtopenetratethethicknessofthematerialleavingalackofpenetrationdefectoritwillblowthroughinacuttingactionwhichwillnotleaveafullyfusedjoint.Ifoccasionallengthsoflackofpenetrationcanbetoleratedthenthismethodcanbeusedwiththeweldingconditionssetdeliberatelytoofferlackofpenetrationratherthanblowthrough;wherefullpenetrationisessentialtherootoftheweldcanbegroundoutorgougedfromtheoppositeside(backgouged)andaweldmadeonthatside.Forrelativelythickmaterialsamiddlepathcanbefollowedwheretheedgeisbevelledoversomeofthedepthofthejointandthefirstrunisdesignedtopenetratetherootfaceandsubsequentrunsaremadeinthepreparation.Theweldcanbemadewitharunmadesuccessivelyfromeachsideofthejointwiththesecondrunmadeovertheas-weldedrootofthefirstrun.Indoingthisthereisariskofsporadiclengthsoflackofpenetrationorslaginclusions.Aswithsinglesidedweldsifasoundrootinthefinishedweldisessentialthentherootofthefirstrunmustbegroundorbackgougedleavingacleangrooveforthesecondweld,Fig.4.6(a).Therootfacemustbekepttoaminimumdepthotherwisealargeamountofmetalislefttobegougedoutwhichisnotonlycostlybutresultsintheintroductionofagreatdealofheatandtheriskofexcessivedistortion.Whenselectingaweldpreparationdistortionisoneofthefactorstobetakenintoaccount.FirstsideSecondsideBackgouge4.6(a)Backgouginginaweldingsequence. 44Weldeddesign±theoryandpracticeWithhighweldingcurrentstheweldpoolislargeandsurfacetensioneffectsarerelativelylesspronounced;theweldmetalcanthenrunoutofthejointormayflowaheadofandunderthearcpreventingitsstrikingtheparentmetalandcreatingalackoffusiondefect.Asaresulthighweldingcurrentsareusedonlyintheflatorthehorizontalverticalposition.Anotherpointisthatwithhighheatinputstheweldmetalisvirtuallyascastwithperhapsalargegrainsizewhichmayhavepoorpropertiessuchasfracturetoughness;whenusinganumberofsmaller,lowerheatinputruns(orpasses)eachrunheattreatsthepreviousrunandimprovestheproperties.Forthesereasonssomeapplicationspecificationsrestricttheheatinputto,forexample,5kJ/mm.Aswellastakinglesstime,fewlargerunshaveanotheradvantageoveranumberofsmallonesinthatangulardistortioncanbeless.Acompromisearisesheredependingonthedemandsofthespecification.Thechoiceoftheweldpreparationisbasedontheconfigurationofthejoint,theaccessforweldingandinspectionandthecostorthetypeofcuttingequipmentavailable.Thesimplesttypeofedgepreparationisthesinglebevel,asshowninFig.4.6(b).Thiscanbemadeverycheaplybygascutting.Therootfaceisleftbecausethearcwouldmeltasharpedge,orfeatheredge,makingaconsistentwelddifficult.Inadditionanywanderinthecuttingline,Fig.4.6(c),doesnotchangethepositionoftheedgesomaintainingtheconsistentrootgaprequiredforasoundroot.Abouttheonlyjointforwhichthefeatheredgeissuitableiswherethebuttweldismadeontoabackingstriporbarwherethepositionoftheplateedgeisnotsoimportantandwheretheedgemaybefusedintothebackinginanycase.Asmaterialthicknessincreasesitmaybedesirabletochangefromasinglesidedbeveltoatwosidedbevel,Fig.4.6(d),fortworeasons.Firstly,thevolumeofweldmetalandtherebythecostisreduced,Fig.4.6(e)secondly,theheatinputandthermalhistoryismorebalancedthroughthethickness,leadingtolowerlevelsofdistortion.TominimisedistortiontheAngleofbevelIncludedangleRootfaceRootgap4.6cont(b)NomenclatureforVweldpreparation. Considerationsindesigningaweldedjoint45GascuttingnozzletrackvariationsVariationincut(c)Rootfacestaysinthesameposition1/3±2/3preparationEqualpreparations(d)(e)4.6cont(c)Toleranceongascuttingofbeveledge;(d)doubleVpreparations;(e)relativevolumeofweldmetalinweldpreparations;(f)singlepasscuttingofaplateedgewithdoublebevel.preparationisnotmadesymmetrical±thesizeofthepreparationonthefirstsidetobeweldedislessthanthatonthesecondside.Therootfaceservesthesamepurposeaswiththesinglesidedpreparation.Thedoublesidedbevelpreparationcanbegascutquitequicklyandcheaplyinonepasswithathreeburnercuttinghead,Fig.4.6(f).Morecomplicatededgepreparationsareusedtoreduceweldmetalvolumeanddistortion.Thesearebasedoncuttingacurvedrebateintheedgewhichcanbeeitherfromonesideonlygivinganin-linebuttweldand 46Weldeddesign±theoryandpracticeAngleofbevelIncudedangleRootradiusRootgapRootface(g)(h)4.6cont(g)NomenclatureforsingleUweldpreparation;(h)limitationofaccesstoMIG/MAGnozzleiinJpreparation.iscalledaUpreparation,Fig.4.6(g),orfrombothsidesgivingansymmetricaldoubleUpreparation.Wherethispreparationisononeplateonly,asforexampleforaTjoint,itiscalledaJpreparation.Thesetypesofpreparationarenotreadilygascut,exceptbygougingtheassembledjoint,andthemostcommonmethodofmakingthemisbymachining.Thisimmediatelyaddsanothercostandarequirementformachinetoolsotherthanagascuttingtable.TheUandJpreparationsintroduceanotherconsiderationwhichisthatofaccessfortheweldingarc,Fig.4.6(h).Asthethicknessofthematerialincreases,thecostanddistortionbenefitsoftheUandJpreparationsincreasebuttherecomesadepthofthepreparationbeyondwhichthemanuallyheldgasshieldedweldingtorch,orgun,istoobigforthespaceinwhichithastobeoperated.Thearccannotbedirectedatthecorrectangletothesidewallsortherootandthereisariskoflackoffusiondefects.Inthesecircumstancesamanualmetalarcelectrodecanstill Considerationsindesigningaweldedjoint47BackingbarBackingstripCeramictapeorbar4.7Typesofbackingmaterials.beusedsatisfactorilyandtosomeextentashieldedcoredwiregunwhichhasasmallerdiameterthanonewithaseparategasshield.Makingabuttweldfromonesideonlyasinatuberequiresconsiderableskill.Themakingofsuchaweldcanbeeasedbytheuseofbackingstripsorbackingbars.Fig.4.7showsthreetypes.Thebackingbar,whichmaybeofcopperorothergoodheatconductor,isnotfusedintothejointandisremovedafterwelding.Thisdeviceisusedextensivelyinautomaticpipeweldingmachineswherethebackingbarorback-upbarisattachedtothepoweroperatedclampwhichkeepsthepipeendsinalignmentduringwelding.Thebackingstripismadeofthesamemetalasthatbeingjoinedandremainsinposition;itisthereforenotsuitableforpipescarryingfluidsandalsohasarelativelylowfatigueperformance(seeChapter6).Theceramicbacking,intheformofabaroratape,isusedforfineworkwhereasmoothinternalprofileisrequiredandisremovedafterwelding.Theceramicorplainbackingcanbemadeasasprungringwhichcanbeslidintothetubeandwillbeaclosefittotheweldroot.Abuttweldonabackingcanbeusedtoprovideassemblytolerances.If,forexample,theendsofaplateortubecannotbepositionedaccurately,abuttweldonabackingwillineffectoffera`slidingjoint'.Onaplateitmaybeconvenienttotackweldthebackingstriptooneorothersideofthestrip,asinFig.4.8(a).Tubeendscanbemachinedtomakeaspigotjoint,Fig.4.8(b).Onheavywalltubessuchaspilesabackingstripcanbeattachedtothestab-inguide,showninFig.4.8(c). 48Weldeddesign±theoryandpracticeBackingstripStab-inguide4.8Typesofweldbacking.Machiningedgepreparationsonflatplatesrequiresaplanerormillingmachinewhichmayrepresentanexpensivepieceofcapitalequipmentwhereasthemachiningofpreparationsontubesorcircularbarsrequiresalathewhichmaybereadilyavailableinmostmachineshops.Thechoiceofedgepreparationmaythendependnotonlyontheweldingrequirementsbutontheshapeofthepartsbeingjoined.Preparationsmadeeitherbygascuttingorbymachiningwillrequireattentionbeforeweldingcommences.Gascutedgesmustbecleanedofscaleanddrossandanyareasoferraticcutdressedsmooth.Machinededgesmustbecleanedofoil.Inbothcasesifthepartshavebeenexposedtoweathertheymayneedtobecleanedofrustoranypaintandpreservativecoatings.Whatevermethodisusedtoformanedgepreparationtheopportunitymustbetakentoexaminethecutedgeforevidenceoflaminations,inclusionsorotherdefectsinthematerialwhichcouldcreatedefectsintheweldorappearaswelddefectsinanynondestructiveexaminationoftheweld.Theprocessofgascuttingitselfwill Considerationsindesigningaweldedjoint494.9Tbuttweldonabackingstrip.sometimesrevealthepresenceofinclusionsorlaminationsinplate.Theheatfromthecuttingmayexpandthegasesininclusionswhichthendisturbthecuttinggasstreamcreatinganirregularcut;eventhesoundofthishappeningwillpointoutapotentialproblemtotheexperiencedoperator.4.3.2TjointsTheedgepreparationmaybeintheformofasingleordoublebevelorJpreparation.Fig.4.9showsaTjointbetweena50mmanda30mmplate.Thisisabuttweldonabackingstripwherethepatternoftheweldrunsshowthatithasbeenmadeintheverticaluppositionwiththeweldingrodbeingweaved,i.e.movedfromsidetoside.Otherfeaturestonotearethatthebackingstriptackweldhasbeenbrokenbytherotationoftheupstandplatecausedbythethermalshrinkageacrosstheweld.Aswithin-linebuttsthematterofangulardistortioncanbeaddressedbyusingdifferentialedgepreparations,weldingsequenceandrunsize.Alongthemiddleplaneofthemainplateisalineofsegregationorinclusions.Inahighlyrestrainedjoint,forexampleifthisdetailwerea30mmstiffeningringinsidea50mmcylinder,precautionsmayhavetobetakentoavoidlamellartearingundertheweld.Suchprecautionsmayincludeusingalowstrengthweldmetalorusingforthecylinderaplatewithaguaranteedthroughthicknessductility.Inflatplates,distortionislikelytoariseinthecontinuousplatebutthereislittlethatcanbedoneaboutthis. 50Weldeddesign±theoryandpractice4.3.3CornerjointsWesawinChapter2thatsteelplatecanbeafarfromhomogeneousmetal.Itmaycontainlayersofsegregatedconstituentsonamicroscopicscale,inclusionsofnon-metallicsubstancesonaslightlylargerscaleandlaminationsonanevenlargerscale.Thesefeaturescancausedefectivejointsiftheyarepresentandarenottakenintoaccount.ThedesignofacornerjointinsteelrequiresmorecarefulconsiderationthanaTjointtoavoidpotentialdefectsfromlaminations,inclusionsorlamellartearing;theproblematthedesignstageistoknowhowfarourprecautionsneedtogobecausewemaynotknowthenatureoftheparticularsteelthatistobeused.Toexaminethepositionletustakeaworstcasescenarioinwhichthesteelissegregatedandfulloflamellarandlaminarinclusions.Ifwemakeasimplecornerjointwiththepreparationononeside,asinFig.4.10(a),therearehighresidualstressesactingonthesurfaceofplateAwhichcanresultinde-laminationifthereareinclusionsclosetothesurfaceorlamellartearing.Therearetwomeansofreducingtheriskoftheseoccurrences.Oneistospecifycleansteel,i.e.a`Zquality'plate;thiscanbeexpensiveorcausedeliveryproblemsonavailability.Theotheristoputpartoralloftheedgepreparationontheoppositesideofthejoint,asinFig.4.10(b).ThiswillreducethethroughthicknessstressontheplateAandsotheriskofdefects.Howeverwithsomepoorqualityplatethereismoreofapossibilityoffindingsegregatedconstituentsalongthecentrelineofthesurface,whichmayresultinhydrogencrackingatthetoeoftheweld.Itmaythereforebewisetoextendthebevelbeyondthecentreofthethickness.Thisprecautionmayalsobewiseinafilletweldedlapjointwheretheweldtoemaylandonalineofsegregation,asinFig.4.10(c).4.4DimensionaltolerancesAswithallengineeringworkitisnecessarytodefineatoleranceonthedimensionsoftheweldpreparations.Thishastoincludenotonlyanallowanceforunavoidablesmallvariationsintheshapeoftheedgepreparationbutinthefitupbetweenthematingparts.Suchtolerancesmustallowforlinearorangularmismatchacrossthejointandtheweldingproceduremustbedesignedtocopewiththepermittedtolerances.Acertaindegreeofmismatchcausedbylocalvariationsinfitupbetweenpartscanbereducedorevenedoutbyclampsordogs.Notethatthisisnotthefunctionofaweldingjigwhichisdesignedtoholdinplace,andtowithinthefit-uptolerance,partswhichthemselvesaremadetowithintolerance.Whereclosetolerancescannotbeheld,forexampleintheassemblyonsiteoflargesub-assemblies,thewelddetailitselfmaybedesignedtoacceptmismatches.Ifarootgapcannotbeheldtoaconsistencynecessarytoofferassuranceofa Considerationsindesigningaweldedjoint514.10(a)Jointshowingpossiblelocationoflamellartearinplateofpoorthrough-thicknessductility;(b)edgepreparationreducesriskoflamellartearingandtoepositionavoidssegregatedareas;(c)filletweldtoelandingoncentrelinesegregationshowingpossiblecrack.soundweldthenadevicesuchasatemporaryorpermanentbackingasinFig.4.8maybeconsidered.Withthisapproachadeliberatelylargerootgapisleftsothatfullrootfusionisobtainedallalongthejoint.Toaccommodatelargerpotentialgapsbetweenmemberssuchasinafinalclosingjoint,apuppiecemaybecutonsitetothelengthforthefinalclosure.Suchinsertsmustbelongenoughtoallowfortwofullbuttweldstobemadewithoutinterferingwitheachotherandmustbethesubjectofadrawingsoastoprovideanopportunityforauthorisationandanas-builtrecord.Unauthorisedmake-uppiecescancauseproblems,asinFig.4.11,whichshowsanexampleinarooftrusswhichfailedatthetoeofafilletweldshortlyaftererection.Thediagnosisofthiswasnotstraightforward;tothenakedeyethe`fracturesurface'showedaseriesoflinesredolentofafatiguefracture±notverylikelyinarooftrussjustbeingerected.Closerexaminationshowedthatthesewereinfactsawmarks!Anunauthorisedmake-uppiecehadbeeninsertedtomakeupthespan,noweldpreparationhadbeenmadeontheabuttingendsandtheweldbeadhadbeendressedflush.ThetoeofthefilletweldbetweentheTsectionandthemountingplatehappenedtohavebeenplacedjustatthe`joint'. 52Weldeddesign±theoryandpracticeFracture(f)4.11Unauthorisedmake-uppieceinaTsection.Therehavebeenanumberofengineeringcatastrophescausedbyyardorsite`fixes'intendedtoovercomemismatchordistortionandwhichhavedisastrouslyreducedtheintegrityofthestructure.Inallsuchcircumstancesitisvitaltoensurethatthejointdetailisconfirmedbyanengineerasbeingconsistentwiththebasisofthestructuraldesignoftheconstruction.DistortionisdiscussedinChapter11.4.5AccessThechoiceofweldingpreparationmayhavetorecogniseaccesstothejoint.Accessinweldingmeansallowingtheweldertoseeandreachthejointwiththeweldingrodorgunwhilststillbeingabletoseethearcandmanipulateitalongthejoint.Fig.4.6(h)showshowthedetailoftheedgepreparationcanaffectaccesswithdifferenttypesofweldingequipmentbutthepresenceof Considerationsindesigningaweldedjoint53Clash,noaccesstoweldrootClearaccess(a)(b)4.12(a)Weldpreparationshowingneedforaccess;(b)accesstotheweldrootontheflange.adjacentpartsmayalsointerferewithaccess;examplesareshowninFig.4.12.Withmechanisedweldingequipmentthemovementoftheweldingheadandassociatedmachinerymustnotbeobstructedandtheoperatorneedstoseethearcoritsposition. 5Staticstrength5.1ButtweldsButtweldsincarbon±manganesesteels,madebyarcweldingwithconsumablesgivingweldmetalmatchingtheparentmetalstrength,areasstrongasorstrongerthanthesteelitself.Inveryhighstrengthsteelsitmaynotbepossibleorfeasibletoproduceaweldmetalofmatchingstrengthandsoaweldmetaloflowerstrengththantheparentmetalmayhavetobeaccepted.Formostpurposesthen,abuttweldincommonstructuralsteelsdoesnothavetobetakenintoaccountwhencalculatingthestaticstrengthofastructureinthesensethatitsstrengthmaybeconsideredtobeidenticaltotheparentmetal.Howeverifductility,asmuchasstrength,isanecessaryproperty,suchasintheplasticdesignofabeamtocolumnjoint(Chapter8),itisnecessarytoensurethatthewholejoint,comprisingtheparentmetal,itsheataffectedzoneandtheweldmetal,canoffertherequiredrotationofasectionuptothespecifiedlimitstate.Thisrequiresattentiontotheductilityoftheweldmetalandtheductilityofthecolumnflangematerialinallaxesaswellastothedesignoftheweldpreparationsandweldingproceduressoastoavoiddefectssuchaslackofpenetrationandlackoffusion.Thesemattershaveaspecialsignificanceinthecaseofearthquakeresistantdesignswhichdemandextremesofductilityatthebeamtocolumnjoints.Thedifficultiesinherentinachievingsuchpropertieshavebeenrecognisedandproposalshavebeenmadefortheuseofabeamdesigndetailwhichinducestheplastichingetooccuratapositioninthebeamawayfromthewelded8beamtocolumnjoint.Toallowhighductilitytobeexploitedparticularattentionmustbepaidtothenotchtoughnessofboththeparentandweldmetals.Thismaybeparticularlyimportantinthecaseofundermatchingweldmetalswhereplasticstrainmayoccurintheweldmetalwhilsttheparentmetalremainsinanelasticstate(seeChapter2).Forweldedjointsinaluminiumalloys,whosestrengthmaybereducedbytheheatofwelding,theapplicationstandardsgivestrengthsorfactorswhichhavetobeusedtoaccountforthis. Staticstrength555.2FilletweldsFilletweldedjointscansupportloadsbydevelopingstresseswhichareexpressedfordesignpurposesintermsofastressacrosstheweldthroatarea.Thisapproachpostulatesfailurebyshearacrosstheweldthroatalthoughfracturedfilletweldsoftenexhibitdifferentfracturepositionsalongtheirlengthrangingfromtheassumedthroatfracturetodetachmentoftheweldlegfromtheparentplate.Usingthissimpleapproachfordesignpurposesthenotionalstressinafilletweldisderivedfromtheappliedloadactingontheweldthroat;thisstressiscomparedwiththeallowableshearstressintheweldmetalwhich,instructuralsteels,isoftentakenashalfoftheyieldstress.Theweldthroatstressiscalculatedfromtheformula:P/tl[5.1]wherePistheappliedloadandtistheweldthroatsizeandltheweldlength,9,10asinFig.5.1.Morecomplexformulaehavebeenderivedfromtestsand11aresetoutinsomestandards.Theseformulaearebasedontestsonarangeoffilletweldconfigurations;theallowabletensilestressorresistance,dependingonwhichapproachtostructuraldesignisbeingused,issetasthePLt5.1Filletweld.throatstressorthestressalongthefusionboundaryinanempiricalformula.StressesinorthogonaldirectionsusedintheseformulaearelabelledasinFig.5.2.Testsonfilletweldsinmildandhighyieldsteelswithnominallymatchingweldmetalfoundthenormalstress,s//,tohavenomeasurableeffectonthestrengthoftheweld.ThistypeofstressismostcommoninthewebtoflangeweldinanIbeaminbending.Fordesignpurposesinstructuralsteelsitwasfoundthatthethreeotherstressescouldberelatedtoanallowablestressbyaformulaofthetype 56Weldeddesign±theoryandpractice_____________22bHs|+3(t|+t//)4sc[5.2]ands|4sc[5.3]wheresccanbetheallowabletensilestressorlimitstatestress.Thisisusedasabasisforfilletwelddesigninanumberofstandardsinwhichvaluesforbaretypicallyintheregionof0.8±0.9dependingonthestrengthoftheparentmetal.5.2Stressnotationforfilletweld.Theseformulñlookveryscientificandhaveapassingsimilaritytoother,moreformallyderivedequivalentstressformulae,butreallytheyarejustempiricalrelationshipsdesignedtofittestresults.Onemaydebatewhetherthereisanypointinpursuingacomplicatedcalculationforatypeofweldwhoserootfusion,andthereforethroatsize,cannotbeconfirmed,andwhoseresultsareinanycaseverysimilartothesimplemethod.Ifconcernforstructuralintegritycarriesaneedtocalculateastresswiththisapparentdegreeofdetailsurelyabuttweldshouldbeused;itsintegritycanbeverifiedbyconventionalnondestructiveexaminationandnostrengthcalculationisneeded.Underload,filletweldsaremoreductileinshearalongtheirlengththanacrossitbutthisisrarelytakenintoaccountinengineeringdesign.Inlongfilletwelds,asinlongrivetedorboltedjoints,thedifferentialstraincausestheendstotakemoreoftheinitialloadthanthecentre.Noallowanceappearstobemadeforthisincommonlyuseddesignprocedures.Thesemethodsallrequirethattheweldbeloadedsothatthereisnobendingofthefilletweldedjointaboutitslongitudinalaxis(Fig.5.3).Thisofcoursecouldonlyoccurwithasinglesidedweld.Somesourcessaythatforsuchsinglesidedwelds`eccentricityeffectsshallbetakenintoconsideration'.Theseeffectsarenotdescribedandtheirreferencecancauseunnecessaryconcern,forexampleinweldsbetweenhollowsectionsandendplates.Analyseshavebeenpublishedwhichtakeintoaccountthemomentinducedbythis Staticstrength575.3Loadingonafilletweldedjoint.`eccentricity'buttheycannotbejustifiedfordesignpurposesastheresultstheyproducearenotsignificantlyclosertotheobservedstrengthsthanthoseofsimplermethods.Thedesignthroatsize,onwhichthestrengthofafilletweldiscustomarilyheldtodepend,isbasedonthesuppositionthatthepartsareincontactandtherootisfullyfused.Anylackoffitbetweenthepartsorlackofrootfusionwill,ofcourse,produceaweldthroatshortofthenominalsize.Bythenatureofweldingsuchfeaturesarenotnormallydetectableafterfabricationandthereareanumberofwaysofdealingwiththismatterdependingonthecriticalityoftheitem.Ablanketapproachmaybeadoptedinwhichitisassumedthatalljointshaveacertainlackoffit/lackoffusionofacertainamountwhichisthenaddedtothenominalthroatsize.Amoredisciplinedapproachwillbetoenforcefit-up,ortolerances 58Weldeddesign±theoryandpractice5.4Examplesofpartialpenetrationbuttwelds.thereon,andcontrolofweldingsothatthenotionalthroatisachievedinallcases.Ifthematterofjointstrengthisparticularlycriticalthenperhapsoneshouldconsiderthatafilletweldisinappropriateandabuttweld,whosequalitycanbeverifiedbynon-destructiveexamination,shouldbeused.Partialpenetrationbuttwelds(Fig.5.4)sufferfromthesamelimitationsinthissenseasdofilletwelds.Someproductstandardsrequireanassumptionthattheactualthroatislessbyacertainamountthanthatwhichtheedgepreparationallowsortheweldingprocedurespecifies.Whereanin-linejointisincompressionsuchasacolumninabuilding,andwherethereisnobending,itmaybepermissibleforthefullcrosssectionofthemembertobeusedevenwithonlyapartialpenetrationbuttweld.Thedivisionofweldsintofilletsandpartialpenetrationbuttsisanarbitrarydevicefortheconvenienceofcalculation,sincetheybothrepresentaweldwhoseintegritycannotbeestablishedasreadilyasthebuttweld.Asafinalpointonthesubjectoffilletweldsboththedesignerandfabricatormustbeawarethattherearetwoconventionsforstatingthesizeofafilletweld.Oneistoquotethethroatthicknessandtheothertheleglength.Inthepastdifferentcountrieshavecustomarilyusedoneorotherofthesetwoconventions.Thishascausedproblemswherethefabricatoranddesignerareindifferentcountriesusingdifferentfilletweldsizingconventions.Forexampleaweldmadetoadrawingcallingfora12mmthroatfilletweldbutinterpretedbythefabricatorasa12mmlegwillhaveonlyan8mmthroat.Forstaticloadingtheweldwillhaveonlytwo-thirdsofitsdesignstrengthandinfatigueonethirdorlessofthedesignlife.UseoftheISO2553symbolsovercomesthispotentialproblembyrequiringthedrawingtostatebytheletter`a'or`z'againsttheweldsizethatitiseitherthethroatthicknessortheleglength. 6Fatiguecracking6.1ThemechanismFatigue,inthesenseofthewordusedinengineering,isamisnomerforwhatisaverystraightforwardmechanism.Whatisbeingspokenofhereisperhapsmoreproperlycalledfatiguecracking;eventhisisanirrationaltermforthestepbystepgrowthofacrackunderasuccessionofloadsofamagnitudelessthanthatwhich,inasingleapplication,wouldnotcausecompletefractureorevenyielding.Theinitiationandgrowthofthecrack,beingafunctionoftheapplicationofasequenceofloads,takestimeduringmostofwhichthecrackisinvisibletotheuntrainedeye.Toearlyobservers,thefracturefromsuchacrackevokedideasofaphenomenonofdegradationovertimewhichwassupportedinthemindsoftheproponentsofthisinterpretationbythesmoothfracturefaceredolentofafractureinabrittlematerial;thisledthemtopostulatethatthematerialhadsufferedaninstantaneousfracture.Aphrasecommonlyused,unfortunatelyeveninsomequarterstoday,wasthatthematerialhad`becomecrystalline'withage;ofcoursethemetalsinquestionhadalwaysbeen`crystalline'aswesawinChapter2butthiswasnotwhattheproponentsofthisobservationmeant.Theterm`fatigue'wasadoptedtoreflectthelossoffacultiesfromtirednesswhenviewedintermsofhumanexperience.Thisnomenclaturecanthereforebeseentohavederivedfromalackofobservationandanalysis,circumstancesmorewidelyassociatedwithreligiousbeliefswheretheapparentlyincomprehensibleisgiveneithermysticalattributeswhichcanneverbefathomedornamedafterhumanconditionswhichareananalogyoftheperceivedphenomenon.Whatissosurprisingisthattheseviewswereheldbysomanywhoshouldhaveknownbettersolongintothetwentiethcentury.Probablythefirstpublishedworkonthesubjectwhichtodayisstillcalled`metalfatigue'or`fatiguecracking'wasin1843byProfessorWJMRankineinthecontextofsomebridgegirders.Thenin1871AWoÈhler,ChiefEngineeroftheRoyalLowerSilesianRailway,publishedresultsof 60Weldeddesign±theoryandpracticeMaximumstressStressMeanstressMinimumstressStresscycleTime6.1SNcurveandnomenclature.12someexperimentalworkwhicharosefromtheneedtosolvetheproblemofbrokenwagonaxles.Hesetupatestprogrammesimulatingtheloadingoftheaxlesbytestingroundsteelbarsunderrotatingbending.Hediscoveredthatthetimetofailureofthespecimens,asexpressedbythenumberofrotations,wasafunctionoftheload;heplottedhisresultsonagraphofload,orstress,againstnumbersofrotationstofailure,apresentationwhichbecameknownastheWoÈhler.Perhapsmoreprosaically,itistodaycalledanSNcurve(Fig.6.1),agraphshowingtherelationshipofthefluctuatingstressinamaterialagainstthenumberofrepeats,orcycles,ofthatstresstofailure.Thisfigurealsoshowsthenomenclatureofloadorstresshistories.Theindustrythatprobablydidmoreworkontheoccurrenceandunderstandingoffatiguecrackinginthefirsthalfofthetwentiethcenturywastheaircraftindustry.Thesubjectwasbroughttothepublic'sattention13byNevilShute'snovelNoHighway,inwhichanengineer,MrHoney,predictsthatthetailplaneoftheReindeertypeofaircraftinwhichhewastravellingisliabletofailasaresultoffatiguecrackingafteronlyafewmorehours'flying.Theauthor,whosefullnamewasNevilShuteNorway,was Fatiguecracking61wellplacedtowriteaboutthissubjectashehadworkedattheRoyalAircraftEstablishmentatFarnborough.HeadoptsacceptableartisticlicencebyhavingMrHoneyarriveatanexactcalculatedlifeinhourswhereasinpracticethephenomenonissetaboutwithuncertaintiesderivingfromloadhistory,stress,environment,materialandmanufacturingdetailwhichleadustodaytoquoteprobabilitiesoffailureinacertaintimeornumberofcycles.Theaircraftindustrywasnotaloneinhavingfatiguecrackingtocontendwithbuttheconsequencesofcrackingwereusuallypublicandmorbidlyfinalcomparedwithmanyotherindustries.Inadditiontheefficiencyofanaircraftstructure,drivenbytheneedtominimiseweight,meantthatstresseswereproportionatelyhigherinadditiontowhichtheearlyaluminiumalloysweremorenotchsensitivethanthecommonsteelsusedinotherproducts.AvastamountoftestdatawasacquiredovertheyearsandvariouspresentationsofthatdatadevisedinadditiontotheSNcurve;oneinfluencewhichwastakenintoaccountwasthemeanstress,orexpressedinanotherway,thestressratio,inastresscycle.Inmanyapplicationsthesequenceofstressisnotasimplerepetitionofidenticalstressrangesandmuchworkwasdoneonpredictingfatiguelifeundervariableload,orstress,patterns.Thisledtotheconceptofcumulativedamageinwhichasimplesummationofthenumberofcyclesatvaryingstressrangeswasusedasameasureoftheproportionofthefatiguelifeusedup.Variousmeansofcountingthestresscycleswereinventedtotakeaccountofthecomplexityofmanyloadhistories.Inthe1950sthedesignphilosophywasthatof`safelife'atwhichmajorstructuralcomponentssuchaswingsparswouldbechanged.Becauseoftheuncertaintiesreferredtoabovethissafelifehadtobelessthanthedesignortestlifebyalargefactor,typicallyfive.Thedesignlifewasbasedonanotionalflighthistorywithloadcyclesbasedonassumedaircraftweights,flighttimes,gustfrequenciesandintensities.Naturallythiswasintendedtoprovideaconservativedesign.Itbecameapparentthatbymeasuringtheloadsactuallyexperiencedbyanindividualaircraftthedesignerscouldremoveoneoftheuncertaintiesandcalculatetheactualfatiguedamagewhichthataircrafthadsustainedduringaperiodofoperation.Recognisingthis,thewingsparsoftheVickersViscountairlinersofanumberofoperatorswerefittedwithelectro-mechanicalstrainrangecounters.Fromtherecordsproducedbythesecounters,thetimeatwhichasparchangewasnecessarywouldbepredicted.Thiswouldfrequentlyhavebeenafteralongerlifethanthenominalsafelifesothatacostbenefitaccruedtotheoperatorinmakingthesemeasurements.Converselyanaircraftexperiencinggreaterthanusualdamagecouldbeinspectedearlierthanotherwisemightbethecase.Furthermoreknowledgeoftheactualdamagehistoriesgaveoperatorstheopportunityofplanningroutesandoperatingprocedurestominimisefatiguedamage. 62Weldeddesign±theoryandpractice6.2WeldedjointsBearinginmindthatarcweldingasameansoffabricatingsteelstructureswasbeingadoptedonlyslowlyinthe1930sitissurprisingtofindthatfatiguetestsonbuttweldsinstructuralsteelplateswerereportedasearlyas1939byProfessorWilsonattheUniversityofIllinoisintheUSA.IntheUK,theInstitutionofCivilEngineerssetupacommitteein1937toreviewthedesignpracticesforsteelgirderbridgesbutitsworkwasinterruptedbytheSecondWorldWaranditsreportwasfinallypublishedin1949.Theamountofinterestexpandedgreatlyworld-wideaftertheendoftheSecond14WorldWar;DrTRGurneyhascomprehensivelyreviewedthesignificantworldliteratureofthatandsubsequentperiods.TheBritishStandardforBridges,BS153,thencarriedcertain,somewhatarbitrary,requirementsonreversalsofstressunderthepassageofaliveloadandthecivils'committeecouldfindnologicbehindtheiradoption.Thecommitteeexaminedresearchworkaroundtheworld;GermanandAmericanstandardswereconsideredtobetooconservativeinsomerespects,andunsafeinothers.ArisingfromthisreviewarevisionofBS153waspublishedin1958reflectingcurrentknowledgeofthefatiguebehaviourofweldedsteelstructures.Thiswasfairlylimitedbycomparisonwithtoday'sequivalentstandards.Therewereonlytwoclassesofweldedjoint,nodatawasgivenforlivesgreater6than2x10,forloadcyclesgivingcompressionasthenumericallymaximumstress,forhightensilesteelsorcumulativedamage.By1959itbecameapparentthatthenewclausecouldbeimprovedinthelightofnewresearchresults.MorecomprehensiveruleswerealreadyinexistenceinGermanyandintheUSSRenablingthosecountriestodesigntohigherstresseswhilstusinginferiorsteels.TheInstitutionofCivilEngineersCommitteeproposedthatexistingdatabecollatedandfurtherresearchperformedwhereitwasseentobenecessary.Thistaskwasconductedby15GurneyattheBritishWeldingResearchAssociation(laterTWI)andasaresultafurtherrevisedfatigueclausewasintroducedintoBS153in1962.Inthe1960sinstallationandreplacementofindustrialplantproceededapaceinanumberofindustriesandweldedfabricationswereusedforthefirsttimetoreplacemanyoftheoldcast,forgedandrivetedconstructions.Arashoffatiguefailuresoccurredasdesignersalmostcopiedformershapeswithoutunderstandingtheeffectofweldedjointsonthefatiguebehaviouroftheirmachines.Someofthesefailureshadmajoreffectsontheirmanufacturers,someofwhomwereevendrivenoutofbusinessbytheconsequentiallosses.Theseproblemsaffectedshipandrailwaylocomotivedieselengines,minewindingheadgear,conveyorbeltrollers,dieselrailcarbogies,coalscreens,earthmovingequipment,offshoredrillingvessels,chemicalplantmixers,overheadtravellingcranesandahostofotheritems.Butletusnotbetoohardonthesedesigners,foritwasonlyinthemid- Fatiguecracking631950sthataViscountaircraftcrashedatManchesterbecauseaboltholdingpartofthewingflapsbrokeasaresultoffatiguecracking.Theboltheadwasnotsquarelyseatedandthefluctuatingbendingstresssetupintheshankbytheaerodynamicbuffetingontheflaphadbeensufficienttocauseacracktogrow.ThiswasonlyafewyearsafterthelossinserviceoftwoCometaircraft,thefirstjetpoweredairliner,owingtofracturesinthepressurecabinfromfatiguecracks.Despitetheknowledgeandexperienceintheaircraftindustry,ithadnotbeenappreciatedthattheproofpressuretestingappliedtoaCometstructuraltestfuselagepriortofatiguetestinghadtheeffectofincreasingthefatiguelifesothepotentiallifeoftheaircraftinservicewasinfactshorterthanhadbeenmeasuredinthetest.This,coupledwiththeuseofanaluminiumalloyofratherpoorcrackresistanceproperties,resultedintragiclossoflifeinnotonebuttwocrashes.Howthecrackstartsinanapparentlyhomogeneoussmoothmetalisperhapslessthanobvioustothecasualobserverbutthoseinvolvedinresearchanddiagnosisoffailuresobservedthatwherefatiguecracksdidappeartheytendedtodosoatsharpchangesofsectionoratholes.ThisofcoursereflectedWoÈhler'sobservationsthatthefatiguelifewasafunctionofthestress.Foratimeitwasthoughtthattheshapeofaweldcapwassufficienttocreateastressconcentrationsevereenoughtostartthefatiguecrackbuttestsshowedthattestspecimensmachinedoutofasolidpieceofsteelwithacrosssectionintheshapeofabuttweldhadamuchlongerfatiguelifethantheactualweldswhichtheywereintendedtoreproduce.It16wasnotuntilSignesandcolleaguesattheBritishWeldingResearchAssociation(laterknownasTWI)inthe1960sexaminedthemicroscopicdetailofthetoesofarcweldsinsteelthatthereasonbecameclear.Theyobservedthatalongthetoeoftheweldthereweresmallirregularitiesratherlikecracks;inrealitytheyweretinysurfacecavitiesfilledwithslagandaveragingabout0.1mmindepth(Fig.6.2).Fracturemechanicsconceptswereusedtopredicttheprogressionofafatiguecrackfromsuchaweldtoeandtheresultswerefoundtocorrespondwiththosemeasuredinfatiguetests.Thisworkledtoanunderstandingofothercharacteristicsofweldedjointsunderfatigueloading.Itexplainedwhy,incontrasttoplainunweldedspecimens,therewasnoinitiationphaseinthe17life.Maddoxmadegrowthratemeasurementsinplainsteelsofvariousstrengthandobservedthatforcertainstressintensityrangestherewaslittledifferenceinthecrackgrowthrates.Theabsenceofaninitiationphaseandthesimilarityofcrackgrowthrateshelpedtoexplainwhytherewaslittledifferencebetweenthefatiguebehaviourofweldedmildandhighstrengthsteels.Thisworkconfirmedtheneedtoapproachfatigueinweldedjointsverydifferentlyfromthatinplainsteels.Inplainsteelsthereisaleveloffluctuatingstressbelowwhichfatiguecrackingwouldnotoccur,thefatiguelimit,whichisatastressamplitudeabouthalfofthetensilestrengthofthesteel. 64Weldeddesign±theoryandpractice6.2Slagintrusionatthetoeofaweld(photographbycourtesyofTWI).Forweldedjointsthereisananalogouslimit,definedbythenon-propagatingcracksizerepresentedbytheweldtoeintrusions,atamuch2lowerstressrange,typically20N/mmforallsteelstrengths.18Intheearly1970sGurneyandMaddoxre-analysedtheavailablefatiguedataforweldedjointsandexplainedhowsomeofthepreviouslyderiveddesigndatacouldberationalised.Usingstatisticalanalysessupportedbyfracturemechanicstheywereabletoshowthatsomeoftheapparentlywidescatterreportedintestdatawasnotactuallyrandomscatterbutwastheeffectofsuperimposingdifferentstatisticalpopulationseachwithitsown,butmuchless,scatter.ThefatiguetestingprogrammesetupintheUKinthe1970sassociatedwiththeoffshoreindustrydescribedinChapter9paidparticularattentionbothtothemanufacturingandsettingupoftestspecimensandthemeasurementofthestressinthespecimen.Itwasfoundthatthescatterintheresultsofthisprogrammewasmuchlessthanhadbeencustomarilyaccepted.Itshowedthatwhathadbeenthoughttohavebeennaturalscatterinprevioustestresultshadbeencausedinpartbyspecimentestingtechniques.TheworkofGurneyandMaddoxwaseventuallyincorporatedintheUKDepartmentofEnergyGuidanceontheDesignofOffshoreInstallations,whichisnolongercurrentasformaldesignguidance,andlaterinanewBritishStandardforbridges,BS5400:Part10`Codeofpracticeforfatigue'firstissuedin1980andwhichreplacedBS153:Parts3Band4:1972.Howevertherewasseentobeaneedforthedatatobepublishedwithoutbeingattachedtoorconstrainedinitsapplicationbybeingpartofaproductstandard.Accordinglyin1993BS7608`Codeofpracticeforfatiguedesignandassessmentofsteelstructures'waspublishedwhichcanbeapplicabletoanyproductorsituation. Fatiguecracking65Crack6.3Stressconcentrationataholeatatransversebuttweld.Thestressrangeputintothefatiguelifecalculationwillbethenominalstressinthepartsjoinedbytheweld.TheeffectofthewelddetailsonthelocalstresshasbeenallowedforintheseSNcurves.Howevertherewillbesituationswhereaweldedjointcoincideswithafeatureofthepartwhichcreatesitsownstressconcentration.Anexampleisshownhere(Fig.6.3)inwhichtheweldcomesuptotheedgeofaholeorwheretheholewasdrilledthroughthecompletedweld.TheelasticstressconcentrationintensioncausedbytotheholeattheweldtoewillbethreeandsothenominalstressrangeinthepartwillhavetobemultipliedbythreebeforeenteringtheSNcurve.Stressconcentrationsforallsortsofshapescanbefoundinthe19compendiumoriginallycompiledbyPeterson.Theeffectonthefatiguelifecanbequitesevere,beingafunctionofthethirdorevenfifthpowerofthestressrange.TheshapeofthejointitselfcanalsointroducestressconcentrationsandmostofthefatiguedesigndatausedintheworldgivesseparateSNcurvesfordifferentjointtypes.Inthiswaythedesignerdoesnothavetoanalysethelocalstressdistributionaroundtheweld.Itisimportantinusingthedatatodefinewherethecrackingislikelytooccur.Afilletweldedjointmaycrackattheweldtoeorthroughthethroatdependingonthestressinthetoeandtheparentmetal.Thejointhastobecheckedforbothtypesofcracklocation(Fig.6.4).Amorecomplicatedsituationwherethestressattheweldedjointismagnifiedbytheconfigurationofthepartsisinthecaseofnodaltubularjoints.Fatiguecracking,aswithanyweldedjoint,willstartwherethelocalrangeofstressintensityishighest.ThemodesofloadingillustratedforthetubularTjointhere(Fig.6.5)causebendingstressesaswellasaxialstressesinthetubewall;thehigheststressatthetoeoftheweldwillbeatapositiononitscircumferencedependingonthedirectionoftheload.Thisstresscan 66Weldeddesign±theoryandpractice6.4Potentialpositionsoffatiguecracksinafilletweldedjoint.6.5Deflectionsinthechordwallofatubularjoint.becalculatedusingathreedimensionalfiniteelementanalysis.Ifthisistoocostlyandtimeconsuming,anapproximatemethodcanbeusedtoderive20thehotspotstress;thiscanthenbeusedwiththerespectivepublishedSNcurvetoestimatethefatiguelifeofthejointwhich,asmightbeexpected,isverysimilartotheSNcurveforin-linebuttwelds.Thehotspotstressisthe Fatiguecracking67higheststresswhichwouldbedevelopedatapointaroundthetoeoftheweldwithouttakingintoaccountthestressconcentratingeffectoftheweldprofile.Thehotspotstresscanbefoundbymeasuringthestressesonatestspecimenandextrapolatingtotheweldtoeorbyusingoneofanumberofempiricalformulaeknownasparametricformulaewhichhavebeenpublishedbyvariousresearchworkers.Thefatiguecrackingbehaviourofweldedaluminiumalloysisanalogous21tothatofweldedsteels.DesignSNcurveshavebeenpublishedwhichshowthattherelativefatiguebehaviourofweldedaluminiumdetailsisanalogoustothatofsteels.Iftheappliedstressrangesarecomparedforthesamedetailatthesamelifetheywillbefoundtobeapproximatelyproportionaltotheelasticmoduliofthetwomaterials.6.3ResidualstressesResidualstresses(Chapter11)havetheeffectofplacingmuchoftheweldinanareaundertensilestresswhich,insteels,canbeconsideredequaltotheyieldstress.Anyappliedtensilestresswillyieldthemateriallocallywiththeeffectthatregardlessofthestressratiooftheappliedloadingtheactualstressinthematerialadjacenttotheweldwillvaryfromyieldstresstensiondownwards.WithnoeffectofmeanstressthereisthenonlyoneSNcurveforthedetail,whichforoncemakeslifeeasierforthedesigner!6.4ThicknesseffectItwasknownthatweldedjointsinlargerthicknessesofsteelappearedtohaveshorterfatiguelivesforthesamestresshistorythanthejointsinthinnermaterials.Thiswasoriginallyputdowntothegreaterextentoftheresidualstressfieldinathickersection.Later,fracturemechanicsconsiderationsweretoshowthatthisbehaviourcouldbeexpectedfor15reasonsrelatingtothecracksizerelativetothethickness.Theneedsoftheoffshoreindustryfordesigndataforthickersteelsbecamepressingintheearly1970s.Theexistingfatigueruleshadbeenbasedonlaboratorytests,themajorityofwhich,becauseoftestingmachinecapacitywereonjointsinsteelofaround12mminthickness.Manyofthethennewgenerationofoffshoreplatformshadtubularswithwallthicknessesof50mmoreven75mm.WorkconductedintheUKOSRP(Chapter9)andassociatedprogrammeshadshownthatthiswasnotastraightforwardmatter.Ondetailedreview,someofthetubularjointtestresultsrevealedthatanapparentthicknesseffectcouldarisefromthewayinwhichanextrapolationmethodwasusedtocalculatethehotspotstressattheweldtoe.Withthisspuriouseffecteliminated,thetruethicknesseffectswereidentifiedandincorporatedinvariouscurrentstandardsandcodesof 68Weldeddesign±theoryandpracticepracticeasadjustmentswhichhavetobemadetobemadetothestandardSNcurves.6.5EnvironmentaleffectsTherateatwhichafatiguecrackwillgrowinsteelsintermsofmm/cycleisaffectedbythelocalenvironment.Anaqueousenvironmentwillincreasetherateofcrackgrowthoverthatindryairorvacuum.Aseawaterenvironmentisofpracticalinterestinthedesignofmarineandoffshorestructures.Seawaterisacomplexmixtureofsubstancesandtestshaveshownthatitseffectonfatiguecrackgrowthinsteelisnotasstrongasasimplesalinesolution.Thedesigndataforsteeloffshorestructuresmakesallowancesfortheeffectofseawaterongrowthrateandalsothewayinwhichitreducesthelowerthresholdstress.Thecathodiccorrosionprotectionsystemsusedonmarineandoffshorestructurescaninhibittheeffectofseawateroncrackgrowthrateandinmanycircumstancesrestoreittotherateinair.Otherenvironmentscanaffectfatiguelifeandspecificattentionhastobepaidtothematerialanditsworkingenvironment.6.6CalculatingthefatiguelifeofaweldeddetailFirstly,weneedtohaveinformationonthestresshistoryactingonthedetail;thismaybeobtainedfrommeasurementsinserviceorfromcalculatedstressesfromtheloadhistory.Inas-weldedjointswedonothavetoworryaboutthemeanstress;allthatisneededisthestressrange.Thismaybeofaconstantamplitude,whichistosaythatthesamestressrangeisrepeatedtimeandtimeagain.Itdoesnotmatterwhetheritisrepeatedquicklyorslowly.Ontheotherhandeachrangemaybedifferentinwhichcaseitissaidtobeofvariableamplitude.ThenextstageistofindtheSNcurvefortheparticularwelddetail.Fordesignpurposesvariousstandardsgroupwelddetailsintocategorieswithasimilarfatiguebehaviour.Thetypeofdetailhastobeidentifiedbythetypeofweldandalsobythedirectionofthestresswithrespecttotheweld;Table6.1showsatypicalcategorisation.ForexamplethestressmaybeacrossabuttweldoralongitandtherewillbeadifferentSNcurveforeachofthesesituations.Ifthestressisatanotheranglethematterismorecomplicatedbutthiswillbeignoredforthisexample.Insomestandardstheweldmayappearinoneoftwocategoriesdependingonhowlongitis,howclosetoaplateedgeitisorwhetherthememberonwhichtheweldismadewillbeinbendingordirectstress.Itisalsonecessaryforfilletandpartialpenetrationweldstodefineatwhichlocationthecrackwilleventuallyoccur±rootortoe±sincethelivesmaybedifferent(Fig6.4).Fromthenatureoftheweld,thedirectionofthestressandthelocation Table6.1AnexampleoffatiguedesigncategoriesofweldedjointsDescriptionofdetailClassExplanatorycommentsExamplesshowingcracksitesTransversebuttwelds(a)Weld-machinedflushandprovedfreefromCDefectsignificanceassessedbyfracturesignificantdefectsbyNDTmechanics(b)As-weldedconditionwithgoodprofileDWeldblendssmoothlywithparentmaterial(c)As-weldedconditionotherthan(b)EAppliestoweldswith`peaky'profile(d)ButtweldmadeonabackingstripwithouttackweldsFThecracklocationisattherootoftheweld Table6.1continuedDescriptionofdetailClassExplanatorycommentsExamplesshowingcracksites Weldedattachmentstoastressedmember,buttorfilletweld(a)Attachmentwithinthewidthofthemember,notFThecrackwillstartinthememberatthetoecloserthan10mmtoedgeofstressedmemberoftheweld(b)As(a)butonorwithin10mmofedgeofstressedGmemberWeldsinotherlocations,suchasattubularnodaljointBadicCalculatethehighestlocalstressactingatSNrightanglestothedirectionoftheweld.ThecurveTdiagramisanexampleonly.Dependingontheforaxial/bendingratiothecrackingmaystartatwelds.differentplacesaroundthejoint.UseSNWithhotcurveforapplication.AsanapproximationspotclassDcanbeusedwiththecalculatedlocalstressstress Table6.1continuedDescriptionofdetailClassExplanatorycommentsExamplesshowingcracksitesTypicaldesignSNcurvesshowingweldedjointclasses. Fatiguecracking73ofpotentialcracksweselectanSNcurve.Ontheverticalaxiswefindthestressrangeandreadingacrosstothecurve(actuallyastraightlineinmostlog±logpresentations)wereadoffonthehorizontalaxisthelifeinnumbersofstresscycles.TheseSNcurvesaretheresultofstatisticallyreducingscatteredtestdatatoasingleline.Thislinemaybethemeanofthetestdataoritmayincorporateanotherlevelofconfidence.Themeanlinewillgivethelifeatwhichhalfthenumberofweldsofasimilartypecanbeexpectedtohavecracked.Thismaynotbethoughtsuitableandamoreconservativeline,themeanminustwostandarddeviations,willbeoneinwhichonly2.5%oftheweldswillhavecrackedatthatlife.Thislevelofconfidenceiscommonlyusedasabasisforpracticaldesign.Whentheconsecutivestressrangesarenotthesame,adevicecalledthecumulativedamageruleisused.Thisrulewasproposedin1924byPalmgrenandrestatedin1945byMinerunderwhosenameitismorecommonlyknown.ItisverysimpleandsaysthatatanystressrangeSwhenthenumberofcyclesofstresstofailure(thefatiguelife)isNthenanylessernumberofcycles,n,ofthesamestressrangewilluseupafractionofthisfatiguelifeequalton/N.Thisfractioniscalledthe`fatiguedamage';whenthisdamagereachesonetheweldhascracked,orfailed.SoifthestresshistorycomprisesstressrangesS1,S2,S3forn1,n2,n3,cyclesrespectivelytheamountofthefatiguelifeusedup,thedamage,isn1n2n3Ð+Ð+Ð[6.1]N1N2N3whichisshownindiagramaticforminFig.6.6.Thewholelifeisusedupwhenthisisequalto1.Thisisnotanexactcalculationandvaluesintestshaverangedfromlessthan1to3ormore.Likeotherformulaeinthisbook,itstartedlifeasanempiricaldeductionandhasbeenshownbyfracturemechanicstohavesomebasisinmaterialbehaviour.Inmuchofengineering,theaccuracyofthestressfiguresaswellasthestresshistoryareuncertainandtobeconservativesomedesignauthoritiesplaceafactoronthisdamageoronthecalculatedlife.InsomecomplicatedstresshistoriesStressrange6.6SNdataforcumulativedamagecalculation. 74Weldeddesign±theoryandpracticeitcanbeverydifficulttodecidewhatconstitutesastressrangeandtherearemethodsofdividingupthestressfluctuationswhichtrytoconservethe15intentofthecumulativedamagerule;Gurneyreviewsthesemethods. 7Brittlefracture7.1ConventionalapproachestodesignagainstbrittlefractureAbrittlefractureinametalisaresultofcrackpropagationacrosscrystallographicplanesandisfrequentlyassociatedwithlittleplasticdeformation.Thepropagationofacleavagecrack,asitisknown,requiresmuchlessenergythandoesaductilecrackandsocanoccuratanappliedstressmuchlowerthanthatatwhichfailurewouldnormallybeexpected.Inengineeringmaterials,suchafractureusuallystartsfromsomenotchsuchasafatiguecrack,aweldingcrackorlackoffusion±inotherwordsahighlylocalisedstressconcentration.Theexplanationofthemetallurgicalmechanismsandinfluencessurroundingbrittlefractureareverycompli-catedandthereaderwhowishestoknowmoreshouldconsultreferences22suchasHoneycombe.Oneoftheprincipalreasonswhythesubjectofbrittlefractureoccupiesakeyplaceinthedesignofsteelfabricationsisthattheferriticsteelschangetheirfracturebehaviourwithtemperature,frombeingnotchbrittleatlowertemperaturestobeingnotchductileathighertemperatures.Whatismore,thetemperatureatwhichthischangetakesplacedependsonthechemicalcompositionandmetallurgicalstructureofthesteel.Thisismorethananacademicdistinctionbecausethistransitionfrombrittletoductilebehaviourtakesplaceclosetothetemperatureatwhichmanysteelfabricationsoperate.Thephenomenonisassociatedparticularlywithweldedfabrica-tionsbecausetheenergyrequiredtopropagateabrittlefractureislowwhichmeansthatthestressrequiredtostartthecrackcanbesuppliedjustbytheresidualstressesfromweldingwithoutthenecessityofanexternallyappliedstress.Weldsmaysupplyastressconcentrationintheformofacrack.Furthermoreweldingcandamagethefracturetoughnessofthesteel,andinthepastsomeweldmetalsthemselveshadverypoorfracturetoughness.Abrittlefracturecanbedrivenbythestrainenergylockedupinthemetalandmaynotneedanexternalloadorforcetostartit.Brittlefractureisafastmovingunstablefracturewhichhasbeenknowntosevercompletesections 76Weldeddesign±theoryandpracticeofweldedbridges,ships,pressurevesselsandpipelines.Thespeedoftheprogressionofthecrackfronthasbeencalculatedasabouthalfthespeedofsoundinthesteel.Insomecasesthecrackhasbeenarrestedbytheexhaustionofthestrainenergyorbyitsrunningintoaregionofhighfracturetoughness.Thebasisoftheapproachtodesignandfabricationtopreventbrittlefractureoccurringthenliesinappropriatematerialselectionandweldingproceduredevelopment.Inalimitednumberofapplicationsstepsaretakenindesigntointroducedeviceswhichwillarrestarunningcrack.Forexampleinpipelinesandothervesselsthelongitudinalweldsinadjacentpipelengthsareoffsettoavoidpresentingacontinuouspathofsimilarpropertiesalongwhichafracturecouldrun.Asanalternativetothis,aringofthickermaterialorhighertoughnessmaterialmaybeinsertedatintervalswhichlocallyreducesthestresssufficientlytoarrestacrack.Ithastoberecognisedthatitrequiresamaterialofmuchhigherfracturetoughnesstostopacrackthanwouldhavebeennecessarytohavepreventeditstartinginthefirstplace.Weshouldrecognisethattheconsequencesofservicecanalsoleadtocircumstanceswhereabrittlefracturemayoccurinafabricationwhichwasinitiallysound.Forexamplefatigueorcorrosioncracksmaygrowtoacriticalsizeduringthelifeofthefabrication;irradiationinnuclearplantcanreducethefracturetoughnessofsteels.Materialsotherthanferriticsteelsneedtohavedefinedfracturetoughnessbuttheydonotexhibitasignificantchangeofthatpropertywithtemperatureandsothequestionofmaterialselectionhasonelessdimension.Weshallseelateroninthischapterhowthesteelcanbetestedtoclassifyitssuitabilityforuseinanyparticularcircumstancebutfirstweneedtoconsiderthefactorsthathaveabearingontherequirementsforfracturetoughness.Foranygivenqualityoffabricationtheseare:.thickness.appliedstress.fracturetoughness.Thecriterionofappliedstressreferredtohereisnotaquestionofsmalldifferencesincalculatedstressinamemberbutwhetherornottherearelargeareasofhighstressconcentrationandconstraint.Examplesoftheseareasarethenodesintubularjointswheretherearelargelocalbendingstresses,causedbyincompatibilityofdeformations,andthestressconcentrationsinherentinopenings,nozzlesandbranchesinpipelines,pipeworkandpressurevessels.Greaterthicknessisafeaturewhichengenderstri-axialstresssystemswhichfavourplanestrainconditions.Inadditionthickermaterialwillcontainmorewidelyspreadresidualstresssystemsthanthinnermaterial.Foranycombinationofthicknessandstresswecanthenchoosethelevelofparentmetalfracturetoughnesswhich Brittlefracture77researchandexperiencehasshowntobeappropriate.Perhapsitisnotunexpectedthattheappropriatechoicewillbesetdowninastandardspecificationfortheproductorapplicationwhichwehaveinmindandwhichitselfwillrefertoarangeofsteelspecificationsinanotherstandard.Theapplicationwillalsoperhapsintroduceasabasisofselectionothercriteriawhichhavenotbeenmentionedsofarsuchasthatofrisk,representedbythehazards,theirconsequencesandthelikelihoodoftheiroccurrence.7.2FracturetoughnesstestingandspecificationIncidentsofbrittlefractureinrivetedstructureswerereportedinthelatenineteenthcenturyandofweldedstructuresinthe1930sbutnocoherentapproachtoinvestigatingthereasonsemerged.Eventuallyitwasthefractureofthehullsofmorethanonefifthofthenearlyfivethousand`LibertyShips'builtintheSecondWorldWarwhichledtoworkon23,24categorisingweldedsteelsbytheirpropensitytobrittlefracture.TheLibertyShipwasatypeofmerchantship,virtuallymassproducedbyweldingintheUSAinresponsetotheneedtokeeptheUKandtheUSSRsuppliedwithfuel,arms,foodandothernecessitiesinthefaceoftheGermanattempttoblockadetheNorthAtlanticandothersearoutesusingsubmarines.InvestigationsintheUSAandtheUKconcludedthatincidentsofbrittlefractureintheseshipsweremorelikelytohaveoccurredwheretheCharpytestenergyofthesteelwaslessthan15ftlb(20J).Eventodaymoststructuralsteelspecificationsusethismeasureoffracturetoughness,evenifnotthesamenumericalvalue.IntheCharpytestanotchedbarofthesteelisstruckbyapendulum(Fig.7.1).TheenergyabsorbedbythebendingandEnergyTransitiontemperaturerange7.1Charpytestspecimenandtypicalresults. 78Weldeddesign±theoryandpracticefracturingofthebarisameasureofthefracturetoughnessofthesteel.Thesetestsaredoneonanumberofsamplesatdifferenttemperaturesandtheenergyabsorbedisfoundtovarywiththetemperature.Thechangeofenergyoccursoverarangeoftemperaturecalledthetransitiontemperaturerange.Theenergymeasuredisnotafundamentalmeasurementwhichcanbemathematicallyrelatedtoquantitiessuchasstressintensityalthoughcertainempiricalrelationshipshavebeenderived.Howeverasaresultofexperience,certainminimumvaluesofCharpytestenergyhavebeenfoundwhichgivefreedomfrombrittlefractureinconventionallyfabricatedconstructions.Thefullline,2,isthecurvegivenbythesetofresultsmarkedX.Thehigherresultstotherightareonwhatiscalledtheuppershelfalthoughtheminimumvaluesrequiredbymanyspecificationswilloftenbefoundinthetransitionrangebutabovethelowershelffigures.Thesteelmakercanproducesteelswithdifferentlevelsoffracturetoughnessanddifferenttransitiontemperaturerangesasinlines1and3.Withinthecarbon±manganesesteelsthisisachievedbyacombinationofmetallurgy,mechanicalworkingandheattreatments.Generallythefinerthegrainsizeofthesteelandthefewerthenon-metallicinclusionsthehigherwillbethefracturetoughness.Thispropertyintheparentmaterialdeterminesthelowesttemperatureatwhichafabricationcanbeused,providedthatitisnotoverriddenbytheweldandheataffectedzoneproperties.Theminimumtemperatureatwhichitispracticaltousecarbon±manganesesteelfabricationsisaround±408C.Thealloysteelscontainingaround9%nickelaresuitabledowntoaround±1908C.Belowthattemperatureausteniticsteelsoraluminiumalloyscanbeused.Althoughtheyexhibitnosharptransitiontemperatureeffecttheirfracturetoughnessstillhastobecontrolled.ThetemperatureatwhichtheminimumCharpyenergyisspecifiedisnotnecessarilytheminimumtemperatureatwhichthefabricationcanbeusedsafely.TheCharpytestspecimenisofastandardsize,55610610mm,regardlessofthethicknessofthesteelfromwhichitistaken.(Therearesub-standardsizesformaterialsthinnerthan10mm.)TheeffectsofthicknesswhichwehavespokenaboutmeanthatasthethicknessgoesupwehavetouseasteelwiththeminimumrequiredCharpyenergyatlowertesttemperatures.Forexample,atypicaloffshoreplatformspecificationmightrequireacertainminimumenergylevelatcertainCharpytesttemperatures,dependingontheminimumservicetemperatures,foras-welded,i.e.notpostweldheattreated,fabricationsinregionsofhighstress.Theseregionswouldnormallybethenodaljointsintubularstructures.Forotherpartsofthestructurelessdemandingpropertiesmightberequired.Thesetemperaturesapplytocarbon±manganesesteelsofallstrengthsandadifferentCharpyenergyisrequiredofeachgradeofsteel.TypicallytheminimumenergyrequiredisequivalentnumericallyinJoulestoonetenthofthehighest Brittlefracture792minimumyieldstrengthofthatgradeofsteelinN/mm.ThisisnecessarybecausetheenergyrequiredtobendaCharpytestspecimenpriortofractureinahigheryieldstrengthsteelwillbegreaterthanthatrequiredforalowerstrengthsteel.Otherproductssuchasbuildingsandbridgeshavetheirownrequire-mentswhichareusuallylessdemandingthanthoseforoffshoreconstruc-tion;theyrecognisethestressconcentrations,serviceconditions,theconsequencesoffailureandthecustomarylevelsofcontrolintherespectiveindustries.Theserequirementsareexpressedinvariousways.InsomeproductstherequiredCharpytesttemperatureforthesteelisrelatedtoarangeofthicknesses.InothersthethicknessofthesteelrequiresacertainsteelgradewithoutdirectreferencetoaCharpyvalueorsomeothermeasureofnotchtoughness.InsummarytheCharpytesthasanumberoflimitations.Aswellasbeingconductedonathicknessofmaterialnotnecessarilyrepresentativeofthestructureinquestion,thetestmeasuresboththeenergyabsorbedinbendingandthenfracturingthespecimen;further,itiscarriedoutatahighloadingrateunrelatedtomostserviceconditions.Thesefeaturesmeanthatitisnotabasicmeasureoftheabilityofamaterialtosurviveandtheresultscannotbeinterpretedinaquantitativeway.Nonethelessithasthebenefitofusingeasilymadeandrepeatablespecimensandthetestitselfissimpleandquick.Itisthereforeavaluablequalitycontroltool.7.3FracturemechanicsandothertestsWhereamorediscriminatingtestthantheCharpytestoronegivingresultswhichcanbeappliedtotheassessmentofdefectsisrequired,afracturemechanicstestcanbeused.SuchtestscanuseaspecimenfromthefullthicknessofthematerialunderstudyandwithacrackstartingnotchwhichismorerepresentativeofactualwelddefectsthantheratherbluntnotchoftheCharpyspecimen.Thestateofstressaroundthetipofasharpcrackcanbedescribedbyaquantityknownasthestressintensity,K1.Inafullyelasticmaterialthisquantitymayreachacriticalvalueatwhichfractureoccurs,K1C.Wecanmeasurethisbycarryingoutafracturemechanicstestwhichentailsbendingacrackedspecimenandmeasuringtheloadatwhichfractureoccurs.BycalculatingthestressatthecracktipatfracturethevalueofK1Ccanbecalculated.Thiscanbeusedtomakeanassessmentofthesignificanceandacceptabilityofwelddefectsorfatiguecracksifthestressesareelastic.Howeverwehaveseenthatstructuralsteelsarefarfrombeingelasticwhentheyreachyieldpoint.Thestressdistributionaroundacrackorotherwelddefectisextremelycomplicatedespeciallywhenplasticitycomesintoplay.Intheweldedjointtheresidualstressesareasignificantpartofthe 80Weldeddesign±theoryandpracticeoverallstresspattern.Toinvestigatethissituationatestonanactualweldinarealisticthicknessofplatewasneeded.Thisledinthe1960stothedevelopmentoftheWellswideplatetestandthecrackopeningdisplacement(COD)testattheBritishWeldingResearchAssociation23(BWRA,latertobecomeTWI).Weldswithartificialdefectsweremadeinatestplatetogiveaspecimenabout1msquarewhichitselfwasweldedintocastendpiecesthroughwhichloadsfromhydrauliccapsuleswereapplied,eventuallyupto4000tonnes.TheCODtestinvolvedtakingasampleofthesimilartypeofweldasacouponandcuttingintoitanotch.Thisspecimenwasthenputinbendingwhilsttheopeningofthenotchwasmeasureduntilfractureoccurred.Thetestwasfurtherdevelopedwithmorerefinedmeasuringtechniquesandtheextensionofthenotchbyfatiguecracking(Fig.7.2).Thisgavethefinestpossiblenotchandonewhichcouldbeproducedconsistently.Evenatlowappliedstressesthecracktipactuallystretchesplasticallyandthiscanbemeasuredasthecracktipopeningdisplacement,d(CTOD).Thevalueofthisasmeasuredatfractureisusedinassessmentofthesignificanceofcracksorotherfeatures,particularlyinweldedjoints.Inpreparingthespecimenthenotchisfirstsawnandthengrownbyfatiguecrackingtoproducethefinestpossibleandmostconsistentcracktip.Theopeningofthenotchismeasuredbyanelectricaldisplacementgaugeandtheactualtipopeningiscalculatedonthebasisofthecrackandbargeometry.Theuseofafatiguecrackednotchnotonlyensuresthatthefinestcrackisproducedbutitcanbeplacedwithinthecrosssectionofaweldedjointsoastosamplequitenarrowregionsofaparticularmicrostructureintheweldmetalortheheataffectedzone.InFig.7.3a`K'preparationhasbeenusedtogiveaheataffectedzonestraightacrossthesectionsothatthefracturewillalwaysbewithinthesamemicrostructureasClipgaugeFatiguecrack7.2Specimensetupforcracktipopeningdisplacementtest. Brittlefracture817.3CTODtestspecimenforabuttweldshowingthetipofthenotchinaweldselected.7.4Abrittlefracture(photographbycourtesyofTWI).itmovesintothespecimen.Fig.7.4showsabrittlefracturesurfacewiththetypicalchevronpattern`pointing'totheoriginatingcrack. 8Structuraldesign8.1Structuralforms8.1.1SteelframesSteelbuildingframesrangefromsimplesingle-storeybuildingstovastmulti-storeyskyscrapers.Theyhavebolted,rivetedorweldedjointsandattachments.Oneoftheearliestofthelargeironbuildingframeswasthe25CrystalPalacebuiltinLondonfortheGreatExhibitionof1851.ThiswasdesignedbyJosephPaxton,whowasnotanengineer.Hewasinitiallyagardener,becomingheadgardeneratChatsworth,theseatoftheDukeofDevonshirewhosegroundswerelaidoutbyCapabilityBrown.PaxtoneventuallybecameadirectoroftheMidlandRailway.TheBuildingCommitteeoftheGreatExhibitionincludedengineersoftheeminenceofBrunelandStephenson.TheyacceptedPaxton'sdesigninpreferencetotheCommittee'sowndesignwhich,likemostcommitteeoutputs,wasthelowestofcommondenominators.SincePaxtonhadnoengineering,knowledgethedetaildesignandcalculationsforhisconceptwereperformedbythecontractors,FoxHenderson&CoofSmethwick.FoxwaslatertofoundthefirmofconsultingengineersthatbecameFreemanFoxandPartnersinthetwentiethcenturyandthatwasresponsibleforsomeofthelargebridgesintheworldtoday.SampleelementsoftheCrystalPalacestructureweretestedandsurvivedfourtimesthedesignloadbeforefracture.Thestructurereliedforitslateralstabilityentirelyontherigidconnectionbetweenverticalironcolumnsandhorizontalbeams.Inthisitdifferedfromallpreviousironconstructionsinwhichthisportalbracinghadbeenachievedeitherbyarchedgirdersorspandrelbrackets.Inthismanneritreflectedthebasisoffuturebeamandcolumnstructuraldesignwhichhasbeenusedformostbuildingframessince. Structuraldesign838.1.2BoxsectionsInthiscontextwearespeakingofbuilt-uprectangularboxsectionsandnotrolledhollowsections,whicharecoveredinthenextsection.Boxesareaveryefficientsectionforlongbridges.Theyarerelativelyeasytobuildandpaintandtheinteriorscanbeusedforaccessforinspectionandrepairaswellasforcarryingservices.Beingineffectlargeandrelativelythinplatestructuresdistortionhastobecontrolledandparticularattentionhastobepaidinthedesigntostructuralstabilitytopreventprematurebuckling.ThisinstabilityledtoadisasterwhenadiaphragminoneofthelengthsoftheboxsectionofthenewbridgeforamotorwayatMilfordHavennearBristolcollapsedasitwasbeingrolledoutoverasupport.Thisincidentfollowedcloselyonthecollapseofanotherboxgirderbridgeduringerection,theYarraBridgeinMelbourne,Australia.Thiscomprisedtwoparallelboxescurvedinplan.Inattemptingtorectifyamismatchinelevationsomeflangeboltsweretakenoutofoneboxbutthisallowedthetopplatetobuckleandthestructurecollapsedandfell.ThecausesofthetwocollapsesweredifferentbutbothbeingboxgirdersdesignedbyUKconsultingengineers,anenquiry,theMerrisonEnquiry,wassetupintheUKtoexaminethewholematterofweldedboxgirderdesign.Anoutcomeofthisenquirywasthatthedesignpracticeswerechangedtotakeaccountoftheeffectonstabilityofresidualstressesanddimensionaltolerancesinlargethinpanels.Residualstressesanddistortionaretwosidesofthesamecoinandboth26canaffecttheabilityofaplatetocarryacompressiveload.Aplateincompressionwillsupportaloaduptoapointwhereitbeginstobuckle.Thestressatwhichbucklingstartsinaperfectlyflatplateisafunctionofitsthickness,thewidthandlengthbetweenmembersboundingtheplateand27theproportionsoftheboundarymembersthemselves.Whentheplatebucklesitcannolongersupporttheloadwhichisthentakenbytheboundarymemberswhichthemselvesmaybeunabletosupporttheload.Fig8.1showsasimplepanelincompression.Thetheoreticalbucklingstressisgivenbyanequationoftheform:2Etsb=KÐÐб2().[8.1]1±ubKhasvaluesdependingontheratioofthelengthofthesidesandthefixityoftheedges.Iftheplateisweldedontotheboundarymemberstherewillbetensileresidualstressesalongitsedgeswhichwillbebalancedbycompressivestressinthecentreoftheplate.Theresultwillbethattheappliedloadrequiredtocausetheplatetobucklewillbelessthanforaplatewithoutresidualstresses.Further,ifweldinghascausedtheplatetodistortoutofitsplaneitwillbuckleearlierthanwouldaperfectlyflatplate.Theseeffectsaretakenintoaccountinsettingdesignstressesforweldedplateand 84Weldeddesign±theoryandpractice8.1Platebuckling.boxstructures.Clearlyitisimportanttostructuralperformancethattheresidualstressesanddistortionarekeptassmallaspossiblebycarefuldesignofthestructure,theweldingproceduresandtheplanningofweldingsequences.8.1.3Tubularmembers8.1.3.1EarlyexamplesInasurprisinglyshorttimeafterironandsteelbegantobeusedasastructuralengineeringmaterial,tubeswereadoptedasastructuralforminsomeverylargestructures.AmongsttheearliestexamplesoflargescaletubularsteelstructureswerethreerailwaybridgesintheBritishIsles.In1848,RobertStephensonbuilttheBritanniaBridgetocarrytherailway Structuraldesign85acrosstheMenaiStraitsbetweenNorthWalesandtheislandofAnglesey.Thisbridgeisarectangularboxsection,anotherformoftube,orhollowsectionaswemightcallittoday.TheTamarBridge,openedin1859,wastheworkofIsambardKingdomBrunel,wellknownforhisotherengineeringworks,andcarriesthemainlinefromLondontotheWestofEnglandacrosstheRiverTamaratSaltashnearthesouthwestcoastofEngland;itisatwospanbridgeinwhicheachspanhasacurvedovalsectiontubeasatopchord.TheForthBridge,Fig.8.2,theworkofSirJohnFowlerandSirBenjaminBaker,carriesthetwotracksofthemainEastCoastrailwaylinebetweenLondonandthenortheastofScotlandacrosstheFirthofForthinScotland;itsconstructionwasstartedin1882anditwasopenedin1890.Itsportstubularmembersonagrandscaleandamongstotherthingsitisnotablethatasacontributiontostructuralintegritytherivetholeswerereamed.Thesestructuresdonotstrictlycomewithinthescopeofthisbookbecausetheyarenotweldedbuttheydoillustratethatthetubewasastructuralformwhosepropertieswereappreciatedbysomeofthegreatestengineersofthepast.Onamuchsmallerscalethanthesegrandbridgesofthenineteenthcentury,steeltubebegantobeusedforbicycleandmotorcycleframesinthenineteenthcenturyandformanyyearsthetubeswerejoinedmainlybybrazedsocketjointsalthoughweldinghassincetakenoveronmotorcycleframesandsomecycleframes.ThefirstweldedproductionmotorcycleframesweremadewithMAGweldinginthe1960sandsufferedearlyfatiguecracking.Thedesignershadnotrealisedhowgoodwasthefatigueperformanceoftheoldbrazedsocketjointwhichhastheotherbenefitsofbeingselfjigging,easytopaintandeasytocleaninusebecauseofthe8.2TheForthrailwaybridge. 86Weldeddesign±theoryandpracticesmoothnessofthebrazedsocket.Thelowtemperatureofthebrazingprocessalsoallowedalloysteelstobeusedwithoutlossoftheirstrength.Tubewasusedformajorcomponentsofmanyoftheearlyaeroplanefuselageandwingstructures,evenuntilthe1940s,inairframecomponentssuchasthefuselageoftheHawkerHurricane,firstflownin1935,andthewingsparsoftheVickersWellingtonwhichfirstflewin1936.TheWellingtonsparwasofaluminiumalloytubewhichatspanwisewingjointswasconnectedbyserratedplatesclampedinplacebytransversebolts,adetailwhichtodaywouldraiseconcernsaboutfatigueperformance.Afterthatperiodtheonlymajoritemsinaircraftmadeoftubehavebeenenginemountingsandsomelightaircraftfuselagesandhelicoptertailboomsmainlyconstructedofweldedsteeltubes.Suchstructureswereoriginallymadebygasweldingthejointswhichsuitedthesmallsizesoftubeandgavesmoothjoints.Theyperhapshaveabetterfatiguelifethanthesamejointsmadewithmetal-arcweldingandwhichinlateryearshavebeenreproducedwithTIGwelding.8.1.3.2Tubularsinbuildings,offshoreplatformsandotherstructuresThetube,orhollowsection,hasbeenusedbymanfromtimeimmemorialassuppliedbynatureintheformofbamboo.EventodayinindustrialisedSouthEastAsiancountriesbambooisusedforquitelargescaffoldingsaroundbuildings;thejointsaremadewithlashingsmadeofplasticsinplaceoftraditionalvinesorgrasses.Sincethemiddleofthetwentiethcenturysteeltubeshavebeenusedextensivelyforstructuralpurposesnotonlyascircularhollowsectionsbutincreasinglyassquareandrectangularhollowsectionswhichhavefoundfavourinbuildings,smallbridgesandotherarchitecturalapplicationswheretheirpropertiesandappearancegavethemadvantagesoverthetraditionalrolledsteeljoist,IandHsections.Extensiveresearchintothepropertiesofjointsinthesehollowsectionshasbeenfundedbythesteelmakersaspartoftheirmarketingstrategy.Thishasledtoadetailedunderstandingoftheperformanceofweldedjointsinhollowsections,andthedevelopmentofoptimumconfigurationsofthejointsforvariousloadcombinations.Moststeelforhollowsectionsusedinbuildingsiscarbon±manganesesteel,althougharatherunusualbuildinginCannonStreetinLondonhasanexposedtubularlatticemadeofferriticstainlesssteeltubewhosemembersarefilledwithwaterforfireresistance.Oildrillingandproductioninstallationshavebeenconstructedsincetheearlypartofthetwentiethcentury.Asexplorationandproductionmovedfromdrylandtoswamptolakeandthentotheopensea,thedrillingrigandthentheproductionequipmenthadtobesupportedabovethewateronthetypeofplatformwhichhasbecomesocommontodayandwhichwasinitiallydevelopedforuseinthePersianGulf,astheArabianGulfwasthen Structuraldesign87called,theGulfofMexicoandSouthEastAsia.TheseplatformsareconstructedmainlyofsteeltubeswithweldedjointsandthissubjectisexpandedoninChapter9.Awholebranchofstructuralengineeringpracticegrewuparoundthem,eventuallybeingembodiedinstandardsandcodesofpracticesuchasRP2ApublishedbytheAmericanPetroleumInstitute.Thenecessarydiametersandwallthicknessesofthetubesatthepointwheretheymeteachother,nodaljointsastheybecametobeknown,wererelatedtotheloadsthroughsimpleandempiricalformulaesuchas20punchingshearlatertoberefinedbythehotspotstressconcept.8.1.3.3DesigningtubularjointsFromanearlystageintheirtrainingstructuralengineersaretaughttoavoiddesigningintotheirstructureseccentricitiesandout-of-planeloadsbecausetheysetuplocalbending(secondary)stressesinadditiontotheprimarystresses.Primarystressesarethosestressescalculatedbytheconventionalglobalmethodsofstructuralanalysisbutcalculatingsecondarystressesrequiresmoredetailedmethodssuchasthoseusingfiniteelements.Theeffectofsecondarystressescanleadtolocalinstabilityorplasticcollapseunderloadslowerthanthedesignloadsor,inthecaseoffluctuatingloading,ashortenedfatiguelife.Thesesecondarystressesarecustomarilyavoidedbythesimpleexpedientofdesigningmemberstotransferloadsinlineorbyintroducingback-upmembersacrossplates.Examplescanbeseeninthedesignofbridgegirdersoverthesupportsandthegirdersoftopsidemodulesofthebigoffshoreplatformswherethereare`stiffeners'orback-upmembersintheplategirderswherethetransverseloadsarereacted(Fig.8.3).Historicallythisconceptwasnotadoptedonmosttubularnodaljoints.Inthese,ajointwasmadewheretwoormoretubularmembersmeetbystandingtheendsofthebracesonthesurfaceofthechord.Thisplacesthechordwallinbendingwhichwillbeseentocontradictthestructuralengineer'strainingandreallyoughttobeseenasdownrightbadpractice(Fig.6.5).Howwasitthenthatthedesignersofthetubularstructuresmadenodaljointsbetweentubesbyplacingtheendofonetubeagainsttheunsupported8.3Detailofheavygirderconstructionshowingback-upmembers. 88Weldeddesign±theoryandpracticewalloftheothersodevelopinglocalbendingstresses?Whydothesedesignersoftubularstructuresnotfollowthegoodpracticewellestablishedfordecadesifnotcenturies?Theanswermaylieintheoldhumanqualitiesofconservatismandlackofvision,or,inthevernacular,theycouldn'tseethewoodforthetrees.Soletuslookatthecurrentdesignpracticesdesignfortubularjointswhetherforabuilding,anoffshorestructureoraroadvehicle.Thefirststepistodecidewhatshapeandsizeoftubeistobeused.Thiscannotbedoneforeachmemberinisolation.Afeatureoftubularstructuredesignisthatthejointstendtocontroltherelativemembersizes.Ingeneralwestartwiththemainmemberswhetherwecallthemcolumns,legsorchords.Theirsizewilldependontheloadtheyareexpectedtocarryeitherstaticallyorasafluctuatingload.Localbucklingwilldecidetheproportionsofthecrosssectionwhichmayormaynothavetobestiffened.Overallbucklingwillinfluencethespacingofbracingmembers.Thesizeofthesebracesmaywelldependonthejointwhichhastobemadebetweenthemandthecolumnorchord.Forarchitecturaluses,theselectionofrelativemembersizesatthejointsmaybebasedonappearanceratherthantheirstructuralperformance,whichofcoursestillhastobeadequate.8.1.3.4HowtubularjointsworkWecanstartwithasimpleTjointbetweencirculartubesconsistingofachordontowhichisfixedabraceatrightangles.Itisasimplesymmetricaljointwhichwillhelptoexplainhowtubularjointsworkingeneral.Whenthebraceisloadedaxially,i.e.alongitslength,theforceisresistedbythechord.Fig.8.4showshowthistransferofloadoccurs.Whenthetwotubesareofequalsize,mostoftheloadtransfertakesplaceattheflankswherethejointstiffnessishighest.Whenthebraceisverymuchsmallerthanthechordittriestopunchthroughthechordanditsloadisresistedbytheshearingforcethroughthechordwallwhichdistortsundertheloadunderthelocalbendingeffect.Thedistortedshapeofthechordwalliscontrolledbyitsbeingattachedtothebrace,andtheloadedmemberisitselfactingasastiffener,sothereisaverycomplexpatternofstressessetupbothinthechordwallandintheendofthebrace.Whenabracethesamesizeasthechordisloadedlaterally,intheplaneofthejoint,thebendingloadatthechordisresistedbyshearattheflanksandalsobyshearinthechordwallelsewhere;iftheloadisoutoftheplaneofthejointtheloadisresistedmoreattheflanksthanelsewhere.Ifthebraceismuchsmallerthanthechord,thechordwallisputunderhigherlocalbendingandshearfromeitherthein-planeorout-of-planeload.Whatisparticularlysignificantisthatitisatthesepointsofhighstressthattheweldsareplaced. Structuraldesign89(a)(b)8.4(a,b)Braceandchordofequaldiameter,loadreactedmainlyonchordflanks;bracesmallerthanchord,loadreactedonchordface.8.1.3.5MoredetailedinformationThereisasmallnumberofauthoritativeworksreviewingtheknowledgeoftubularjointbehaviouranddesignindetailandthereaderwhowishestoreadmorewillfindthefollowingworksofgreatvalue.In1982ProfessorJaapWardenierofDelftUniversityintheNetherlands28publishedacomprehensiveworkonthedesignandperformanceofhollowsectionsinsteelpresentingtheoutcomeofresearchacrosstheworldonthestaticstrengthandfatiguecharacteristicsofjointsinbothcircularandrectangularhollowsectionsandinjointsbetweencircularandrectangularsectionsandbetweencircularbracesandopenrolledsectionchords.SometenyearslaterDrPeterMarshalloftheShellOilCompanypublisheda21commentaryonweldedtubularconnectiondesign.ThisworkwaswrittentoexplainthebasisoftubularjointdesignasexpressedintheAmericanWeldingSociety'sStructuralWeldingCodeD1.1tothoseengineerswhohadnotbeeninvolvedinthedevelopmentandapplicationoftheexperience 90Weldeddesign±theoryandpracticeindesigningtubularstructuresfortheoffshoreindustry.ForthoseengineerswhojustwanttoknowwhattodowhendesigningastructurewithhollowsectionsforconventionalbuildingsorsimilarpurposestheBritishSteel29publicationSHSWeldinggivesthenecessarydetailsandallowablestressesbasedonBS5950`Structuraluseofsteelworkinbuilding'.Takentogetherthesethreeworkscouldbesaidtoencompassmostofthebackgroundtothedesignofweldedtubularjointsanditwouldbesuperfluoustoreproducethedetailhere.8.2Designphilosophies8.2.1ElasticmethodofdesignTraditionalstructuralsteeldesignswerebasedontheideathatifthecalculatedstressinanypartofthestructuredidnotexceedanallowablestressthenitwouldsafelysupporttheloaditwasdesignedtocarry.Thisallowablestress,orworkingstressasitissometimescalled,inbothtensionandcompressionwassetasafractionoftheyieldstressortensilestrength.Forpressurevesselsandsomestructuresthisallowable,ordesign,stresswasoncesetataquarteroftheultimatetensilestrengthbutlaterthiswaschangedtotwo-thirdsoftheyieldstress.Thisapproachiscalledtheelasticmethodofdesignbecauseunderthedesignloadnowhereinthestructureisthenominalstressintendedtoexceedtheyieldstress;butclearlytherewillbestressconcentrationsatboltandrivetholesandotheropeningswherethestressmaybeuptoyieldstress.Whilstsuchconcentrationsareacceptedinframedstructuressuchasbuildingsandcranes,pressurevesseldesign(a)(b)8.5(a,b)Reinforcingoftheshellbyadoublerplatearoundanozzle;reinforcementoftheshellbythenozzleitself. Structuraldesign91practiceistointroducereinforcingatopeningstoavoidgeneratinglargeareasofyieldedmaterial;thisreinforcingmaybeintheformofaplateorasthebranchornozzleforwhichtheholeexists(Fig.8.5).Inbridges,cranesandothertypesofstructuresubjecttofluctuatingloads,boltedjointsaredesignedsothatthestressconcentrationsaretakenintoaccountintheworkingstressestoavoidprematurefatiguecracking.Formembersincompressionbucklingisavoidedbyareductionintheallowablestressdependingonthelengthandcross-section.Forsometransientloadingconditionssuchaswindgustloadsonbuildingsthemaximumstresscanexceedthenormalallowablestressbyperhaps25%onthebasisthatthestructurewillnothavetimetoreactdynamicallyinthebriefperiodforwhichtheloadexists.8.2.2PlastictheoryofdesignAlthoughbeingasimplemethodofdesigntheelasticmethodproducesratherinefficientsteelframestructuresintermsoftheweightofsteelusedtosupportaload.Thesizeofamemberisbasedonthemaximummomentanywhereinit;forasimplysupportedbeamwithadistributedorpointloadthisisatonlyonepointonthebeam.Theresultisthattheremainderofthebeamisincreasinglyover-designedtowardsitsends.Bymakingtheendjointsrigid,themaximummomentisreducedandmomentsareintroducedattheends;asmallerbeamsectioncanthenbeusedandmoreeffectivelysincemoreofitslengthisworkingneareritsdesignstrength(Fig.8.6(a)).Beyondthis,rigidjointsofferafurtheropportunityinsteelwhichwastobeexploitedbywhatwastobecomeknownastheplasticmethodofdesign.Thiswasdevelopedinthe1930sbyJFBaker(latertobecomeLordBakerofWindrush)andcolleaguesatBristolUniversityundertheaegisoftheSteelStructuresResearchCommittee.Thismethodwasbasedontheobservationthatarigidlyjointedstructurewouldnotcollapseuntilsufficientmembershadplasticallydeformedinsuchawayastoformamechanism(Fig.8.6(b)).Thisoccurredwhenatthepointsofmaximummomentthewholesectionwouldyieldandactasahinge,aplastichingeasitwascalled.Thecorollaryofthiswasthatindeformingplastically,thesteelwouldabsorbenergy.Intheevent,thefirstpracticalapplicationoftheplasticmethodofdesignwasnottobeasoriginallyenvisaged,inbuildingframes,butinatypeofdomesticairraidshelter,theMorrisonShelter,introducedinBritainintheSecondWorldWar.UntilthattimetheshelterscommonlyusedbyindividualfamiliesassomeprotectionagainstGermanairraidsallovertheBritishIsleswereAndersonShelters.Theseweredug-outsingardensreinforcedwithanarchofcorrugatedsteelsheetingwhichwascoveredwithearth.Intheinnercityareas,wheretherewereflatsandofficebuildingswith 92Weldeddesign±theoryandpracticePPBendingmomentMBendingmomentMM=0Pl/4Pl/12Pl/8Simplysupported(pin-ended)Fixedends(encastreÂ)(a)PPBendingmomentsMBendingMomentMM=0Mp-MpMpSimplysupported(pin-ended)Fixedended(encastreÂ)beamcollapseswhencentrebeamcollapseswhenmoment=Mpandplasticendmoments=Mphingeforms.andplastichingesformattheendsandcentre.(b)8.6(a,b)Elasticbendingmomentinabeamunderapointload;plasticbendingmomentinabeamunderapointload. Structuraldesign93noconvenientgardensandinschools,bothintownandcountry,wherethereweretoomanychildrentouseadug-out,brick-builtcommunalshelterswereconstructedonthestreetsandplaygrounds.InLondonthesesupplementedexistingundergroundspacessuchasundergroundrailwaystationswhichwereusedasshelters.Theeffectivenessofsucharrange-mentsinprotectinglifereliedonforewarningpeopleofraidsinthedaytimesothattheycouldtakecoverinthenearestshelterandontheirsleepinginthematnight.Evensothetragicfactsarethatduringthefirsttwoyearsofthewar,uptotheendof1941,the190000bombs,bothexplosiveandincendiary,droppedbyGermanaircraftonGreatBritainkilledsome44000civilians,including5500children;theyseriouslyinjured3050000people,4000ofthemchildren.LaterinthewarmannedGermanbomberswerereplacedbytheV1orFlyingBomb,alsoknownmoreinformallyastheDoodlebugandwhich,inthelightoffutureweapons,hassincebeencalledthefirstcruisemissile.TheseweresentoverthesoutheastofEnglandin1944and1945inthousandsmostlyaimedatLondonbutinpracticefallingoveralargeareaofsouthernandeasternEngland.Theirsmallsize,speed,numberandunpredictabilityofthesiteoftheireventualfalltoearthmadeanyformofusefuladvancewarningimpracticable.TotallyunpredictablewasthelaterV2ballisticmissileplungingtoearthatsupersonicspeed.Eachofthesetypesofweaponcarrieda1000kghighexplosivewarheadwhoseeffect,aswithbombsdroppedfromaircraft,wasnotonlytokillandmaimpeopleanddemolishthebuildingsnearwhereitfellbuttoradiateablastwavewhichwouldtypicallysuckawalloutofahousesoremovingthesupporttothefloorjoistsleavingthefloortocollapseasaslab.HavingtosleepinundergroundstationsorothershelterswasnotreallysatisfactoryforlongperiodssotheMorrisonShelterwasdeveloped,namedafterHerbertMorrisonthethenHomeSecretary.Itwascolloquiallycalledatableshelter.Itwasactuallyinstalledinsidehouses,givingpeopleashelterwhetherornottheyhadagardenandsoenabledthemtoremainintheirhomes,albeitstillatsomeconsiderableriskofdeathorinjuryfrombombs.Thisshelterprotectedtheoccupantsofahousefromflyingandfallingdebrisbutmorespecificallyfromthecollapseoftheupperfloors.ShowninFig.8.7,itcomprisedsteelportalframesintwoplanescoveredonthetopbyasteelsheetandonthesidesandbasewithwiremesh;ordinarilyitservednotonlyasatablebutasabedstead.Thecomponents,asteelsheetandanumberofpiecesofrolledsteelanglewithboltholes,steelmeshandboltscouldbeassembledquicklybyunskilledlabour,ifnecessarytherecipientfamilythemselves.Thefamilycouldtakerefugeintheshelterintheeventofanairraidwarningandcouldplaceamattressinitonwhichtosleepatnight.Hadtheframebeendesignedontheconventionalallowablestressbasisit 94Weldeddesign±theoryandpracticeSheetsteeltopRolledsteelanglesCollapsedmodeAssemblyboltsSteelmeshdebrisscreens(non-structural)mattresssupportmeshatthebottom.Approximatedimensions2m61.5m60.75m.8.7PrincipalfeaturesofaMorrisonShelter.wouldhavebeenfartooheavytohavebeensupportedbythetimberfloorofahouse.Sincetheshelterwasintendedtosurviveonlyasingleevent,themembersizesandthecornerjointsweredesignedsothattheframewouldprotecttheoccupantsbypartiallycollapsinginacontrolledplasticmannersoabsorbingtheenergyofthedescendingfloorratherthanbyofferingrigidresistance.Thisallowedthememberstobelighterthantheconventionaldesignpracticewouldallow.Ashelterwasdeliveredtoeachhouseholdasakitofsimplesteelpartswithpre-drilledholesforbolts;itwasassembledinaroomonthegroundfloororinthecellarofahouse.Afterthewar,theapplicationofweldingforsteelbuildingframesofferedamuchgreateropportunityfortheexploitationoftheplasticdesignmethod.Oneoftherequirementsofsuchastructurewasthatthejointsshouldbeabletodevelopthefullplasticmomentofthebeamsorcolumns,acharacteristicwhichweldingwasparticularlyabletoproduce.BakerhadbythenbecomeProfessorofEngineeringattheUniversityofCambridgeand31,32withhiscolleaguesdevelopedthisdesignmethodwhichwasfirstusedforthesteelframeofaschoolatHunstantoninNorfolk.ThesecondbuildinginwhichitwasusedwastheFatigueLaboratoryatTheWeldingInstitute.Theplastictheoryisnotapplicableinallcircumstances,forexamplewheredeflectionorfatiguelifeisaconstraint,andauseful33commentarywillbefoundintheSteelDesigners'Manualandinthebook34byDaviesandBrown. Structuraldesign958.3LimitstatedesignTheplasticdesignmethodisanexampleofwhatwenowcalllimitstatedesign.Thisapproachtodesignisbasedonthedefinitionofacondition,orstate,ofthestructurebeyondwhichitwillnotbeallowedtogo.Ifthisstateisforthenormalserviceinwhichthestructureisneithertodeflectmorethanacertainamountnortoshowanypermanentdeformationofthemembersthenitmaybecalledtheserviceabilitylimitstate.Ifthestateistobedefinedintermsofpartialorcompletecollapseofaframe,forexamplebyyieldingorbuckling,itcanbecalledtheultimatelimitstate.Themethodcanbeappliedwithothercriteriasuchasfatiguecrackingorstructuraloscillationorresonance.Itisverydifferentfromtheelasticdesignmethodinwhichthestressiscalculatednottoexceedsomearbitraryvaluewhichmaynothavearationalrelationtotheactualloadbearingcapabilityofthestructure.Thelimitstatedesignprocedurescanplacefactorsonthematerialpropertiestoallowfornaturalvariationsinthoseproperties.Factorscanbeputontheloadstoallowfortheprobabilityofeachtypeandsizeofloadoccurring.Ittherebycanbeamuchmorediscriminatingdesignprocessforsometypesofstructureandhasthepotentialforproducingmoreefficientandoptimiseddesigns. 9Offshorestructures9.1TheneedsofdeepwaterstructuresThedevelopmentofweldingdesignphilosophiesfordeepwateroffshorestructurestookplaceoveraveryshorttimeandsodeservesachaptertoitselfprovidingacompactscenarioofhowsuchphilosophiescanevolve.Thestructuralhollowsectionsusedinbuildings,cranesandsoonarerelativelysmallcomparedwiththosewhichcametobeusedintheconstructionofdeepwateroffshoreplatformsintheNorthSeainthe1970sbutthereisacommonalityofapproachwhichhasbenefitedbothscalesofstructureinprovidingarationaldesignmethodologythroughouttubularjoints.Thedesignofweldednodaltubularjoints,particularlyforoffshorestructures,whichwasalreadywellestablishedforshallowwaters,becamealmostaspecialistsub-disciplineofengineeringintheearly1970s.InEuropethissituationmaybesaidtohavestartedinthemid-1960swhentheNorthSeaexplorationoffthecoastsoftheNetherlandsandtheUKhadfoundgas;butasexplorationmovednorthopportunitiesforoilproductionwererevealed.Boththewaterdepthandtheextremesofweatherincreasetothenorth.ThemuchgreaterwaterdepththanwascustomaryinthesouthernNorthSea,theGulfofMexico,thePersianGulfandtheFarEastfromwhichmostoffshoreoilandgashadsofarbeenextractedwouldrequireplatformsofamuchlargersize.Theirsizewasnotonlyaresponsetotheirenvironmentdirectly;theirisolationandsothedistanceoverwhichtheproducthadtobepipedgavetheneedtoconductsomepreliminaryprocessingoftheproductwhichrequiredon-boardplant.Theisolationalsorequiredafairlylargeresidentcrewwithappropriatequartersandstoresforlongstays,andahelideckcapableofacceptingthelargesthelicoptersinuse.Afterajourneyofoneandahalfhours,theyrequiredrefuellingfromastoreofaviationfuelonboardtheplatform.Thisresultedinaplatformwhichwasmoreamulti-storeyhotelandofficeblockcumprocessplantthanthetraditionalspindlyshallowwaterplatformwithoneortwooperatingcabinsandafewbunksonlyhalfanhour'shelicopterflightfromland. Offshorestructures97ThedriveforthedevelopmentofnewplatformdesignswastheurgetoproduceoilassoonaspossibleafterthediscoveryofviablequantitiesintheNorthSea.Inadditiontotheirsizeotherconsiderationswerethemorecontinuousoccurrenceoflargerwavesizesandlowertemperaturesthaninpreviouslocations.Lloyd'sRegisterofShippinghadbeenreviewingthepositionwithoilcompaniessuchasShellandBPandjointlyconcludedthattherewasinsufficientvalidinformationonwhichtopassjudgementonthelongtermintegrityofsteelstructureswhichwerebeingdesignedforthenorthernNorthSeaoilfields.Theyfeltthatitwouldbeunwisetoextrapolatethedesignrulesusedforthejointsinthesmallerplatformsandapplythemtothelargerones.Twoareasinparticularinthestructuralfieldweredeemedtorequireattention,namelyfatigueandfracture.TheimportanceofsuchmattershadbeentragicallydemonstratedinadisasterwhichhadoccurredintheearlystagesofNorthSeaoilandgasdevelopment.InDecember1965,13mendiedwhentheselfelevatingbargeSeaGemcollapsedintheNorthSea.Thiscatastrophewastohavealargeandpermanentinfluenceontheapproachtomaterialspecificationandfabricationpracticesforoffshoreplatformsand,eventually,onshorestructuresallaroundtheworld.ThelossdemonstratedapressingneedtoensuretheadoptionofsteelspecificationsandfabricationpracticeswhichwouldenablesteelstructurestooperateinthenorthernNorthSeaconditionswithouttheriskofcollapsefromtheeffectsofabrittlefractureofamember.Itisworthlookingatthemainpointsoftheevent.OriginallybuiltintheUSAin1952,SeaGemwasbasicallyasteelpontoonabout100m630minplanand4mindepth.Ithadbeenusedinseveraldifferentpartsoftheworlduntilin1964thehullwassubstantiallymodifiedontwooccasionsatyardsinFrance.AtthetimeofitslossSeaGemhadbeenonstationintheNorthSeaforsomesixmonthsduringwhichtimeithadbeendrillinggaswells.Thecatastrophehappenedasthebargewasbeingloweredbyjackingitdownonitslegspriortoitsbeingmoved.Twotiebarsmadebygascuttingfromsteelplatefracturedleadingtoasequenceofeventswhichculminatedinthebargecapsizing;ofthe32menwhowereonboard13died.Amonthearliertwosimilartiebarshadfracturedwithoutconsequenceandthesehadbeenreplaced,apparentlywithoutanyattemptbeingmadetodiscoverthereasonforthefractures.Althoughthebargewasworkingsome43miles(74km)offshoreitsoperationsweresubjecttotheUKlegislationapplyingontheContinentalShelfandthedrillingwasbeingperformedunderaproductionlicenceissuedbytheMinistryofPower.TheUKContinentalShelfAct(1964)empoweredtheMinisterofPowertomakeregulationswhichweretoincludeprovisionsforthesafety,healthandwelfareofpersonsemployedonoperationsundertakenunderanylicenceundertheAct.ApublicenquiryintothelosswasthereforesetupbytheUK35Government.Aninvestigationofthetiebarsforthetribunalofenquiry 98Weldeddesign±theoryandpracticeshowedthattheyhadpoorfracturetoughnessandwereofasteelsusceptibletostrainageembrittlement.Thegascutedgeswereirregularandinplaceshadbeenrepairedbyweldingandthefractureshadoriginated`fromseverenotchessuchasfatiguecracks,welddefectsandthefilletradiibetweenthespadeendandtheshank'.Theconsequentmodeoffailureoftheremainderofthebargeshowedthatthematerials,thejointdesignandthefabricationpracticeswereunsatisfactoryfortheservicerequiredofthebarge.ThelossoftheSeaGempromptedtheUKGovernmenttointroducefurtherlegislationwhichcamefiveyearsaftertheeventintheformoftheMineralWorkings(OffshoreInstallations)Act1971.By1973onlythreeofitstenproposedsectionswereinforceandfurthersectionsincludedtheprovisionfortheGovernmenttoappointanindependentcertifyingauthoritytoissuecertificatesoffitnessforoffshoreinstallations.Ata36publicmeetingconvenedbytheSocietyforUnderwaterTechnologyinLondonin1973thisproposalwasstronglyarguedagainstbyconsultingengineersandothersinstatementssuchas:`TherearealreadyinsufficientdatafortheNorthSea,andtheoriesareasmanyinnumberasthereareexpertstoexpoundthem.Who,then,canbeexpectedtosaywithcertaintywhichoneiscorrect?Thecertifyingauthorities?Ifthedesignershoulddisagreewiththecertifyingbody,whowilldecidebetweenthem?'ItisinterestingtorecognisethatatthetimethismeetingwasheldsomelargeplatformshadalreadybeeninstalledintheNorthSea!Intheeventseveralcertifyingauthoritieswereestablishedandexercisedtheirrolesuccess-fully.9.2TheNorthSeaenvironmentTraditionheldthatroundtubesweregoodforoffshoreplatformsfortworeasons:firstly,asstructuralmemberstheyweregoodincompressionastheyhaveaveryeffectiveandbalanceddistributionofmaterialacrossthesectionwhichprovidesstabilityagainstelasticbuckling.Secondly,roundtubespresentthesamelateralresistancetothewavesintheseafromalldirectionsandthisresistanceisgenerallylowerthanwouldbegivenbyaprismaticshape;associatedwiththistheirbendingstrengthandstiffnessisthesameinalllateraldirections.Significanthydrodynamicwaveactionisconfinedtoadepthofwaterclosetothesurfaceofthesea,whichincludestheheightknownasthesplashzone,theareaofthestructureintermittentlywettedbythewavesurfaces.Belowandabovetheregionaffectedbywavesthebenefitsofthecircularsectionarenotasstrong.RecognisingthistheClevelandBridgeandEngineeringCompanyLimiteddesignedalargeoffshoreplatformin1973whichtheynamedthe`Colossus'(Fig.9.1);itsprimarystructureinthesplashzonecomprisedcirculartubesandtheremainderwasofflatplateconstruction.Castingsweretobeemployedfor Offshorestructures999.1TheCleveland`Colossus'deepwaterplatform(bycourtesyofDCCDixon). 100Weldeddesign±theoryandpracticecomplexjointstherebyminimisingstressconcentrationssoastoprovidealongfatiguelife.Argumentsforthatdesign,whichwasneverfabricated,werethatthelowerstructuralmembersweremadeinthetraditionalrectangularboxconfigurationcommonlyusedinbridgessothatsecondarystresseswerelowandalongfatiguelifewasenvisaged.Inadditionthestructureemployedwellprovenfabricationmethodstherebyreducingtheuncertaintiesarisingfromnovelty.Caststructuraljointswerenottobecomeapracticalityinoffshoreplatformsuntilmanyyearsafterandthe`Colossus'wasaheadofitstimeinthatrespect.Itisinterestingtoseethatinpractice,wherethelegsofatubularsteelplat-formwereparticularlylarge,e.g.10mindiameter,thestructuralconfigurationoftentendedtobemoreconventionalwithexternalmemberspassingintothelegwiththeirloadsbeingreactedbyinternaldiaphragms.Insomemediumsizetubularsitbecamethepracticetointroduceinternalstiffeningringswheretheloadsfrombracingmemberswerehigh.Whatseemstohavebeeninthemindsofengineersinthismatteroftubularjointswastheexperiencethatwithsmalltubestherewasnoproblemwithweldingtheendofonetubeontothesideofanother.Familiaritywiththisdetailperhapspersuadedthemthatthiswasacleanjointwhichoneshouldnotinterrupt,althoughinsomecasesexternalgussetsorringswereintroducedto`strengthen'thejoint.Thesearenotparticularlydesirableinthesplashzonesincetheyattractmorehydrodynamicloadthanthecleantubeprofileandwheretheyhavebeenusedithasbeenonlyonjointsabovethesplashzone.Therewereproposalsforjointconfigurationsinwhichthebracewascontinuousthroughthechordtherebyavoidingthegenerationofsecondarybendingstressesbutsuchdesignswerenotacceptedbytheoffshoreoperatorsandtheirfabricators.Wasthereatthebackoftheengineers'mindsaresistancetotheintroductionofinternaldiaphragmsfromafearofthesortofweldingproblemsofthe1960s?Thefabricatorshadonlyjuststartedtoovercometheproblemsoflamellartearinginplateandchevroncrackinginweldmetalwhichwerethebaneofheavyfabricatorsinthelate1960sandearly1970s.Howeverneithertheset-onnortheset-throughjointwasmoreorlesssusceptibletotheseproblemsthanothers.Regardlessofallthiswhatweseeisabandofdiametersandwallthicknessesoftubewhichattractedtheset-ontypeofjointandwhichhavethepoorfatigueperformanceasevidenceofthesecondarystressessetupinthisconfiguration.Therewascertainlyapracticaldifficultyinweldinginternalstructureintotubesoflessthan,say,1mindiameter.Thiswasundertakeninsomecircumstancesbutwithpre-heattemperaturesof100ëCandhigher.Weldershadtowearaheatproofsuitwhichmadeworkingveryuncomfortableandweldingveryslow.Butwasthisreallytheproblem?Perhapsweshallneverknow.Wemightbeforgivenforthinkingthatwhat Offshorestructures101thefollowingstoryreallytellsus,withhindsight,isthatsome£10millionwasspentfindingouthowtousepoorlydesignedjointsbutperhapsthatisbeingalittlecynical.Weldedjointsinthelargeoffshoreplatformstructurewereconsideredtorequirespecialresearchforthreereasons:.Therewaslittleexperienceintheuseofweldedjointsinthickcarbon±manganesesteeldowntotemperaturesof±10ëCwheretherewerehighlocalstressessuchasthosesetupbythenatureofthejointsintubularmembersthenbeingdesigned.Fracturetoughnessrequirementshadbeensetdownforprocessplantandforsubmarines,bothofwhichusedheavyplate,castingsandforging.Howeverthedesigndetailsensuredthatanylocalstressconcentrationsweremodest.ThematerialrequirementsforsuchproductsrestedonasubstantialbodyofworkonfracturewhichhadbeeninstigatedduringtheSecondWorldWarfollowingfracturesofanumberofweldedmerchantships.Tocopewithverylowtemperaturestheprocessplantindustryusedalloysteels,austeniticsteelsorothermetalssuchasaluminium..Thefatigueperformanceofweldedtubularjointsinsteelwasnotestablishedinanywherelikethesamedegreeofdetailandconfidenceasforweldedplateandsections.Bythe1960sanumberofcountrieswerepublishingstandardscontainingfatigueclassificationsforweldedjoints,mainlyforapplicationtosteelbridges.AsdescribedinChapter6,intheUKtheBritishStandardSpecificationforsteelbridges,BS153,hadcarriedweldedjointfatigueclassificationsinits1958edition.Extensiveresearchdirectedatthedesignofbothcivilandmilitarybridgeshadledtothepublicationofmorecomprehensivedataintheamendmentto15Parts3Band4ofBS153in1962andforweldedaluminiumalloysinBSCP118in1969..Theeffectofaseawaterenvironmentonthefatiguelifeofweldedjointsincarbon±manganesesteelswasnotwellestablished,andnorwastheeffectofitscorollary,corrosionpreventionsystems.Thenuclearpowerindustryhadinvestigatedthechemicaleffectofseawateronmetalsparticularlyinrespectoftheircoolingsystems,whichallusedseawater,butfatiguewasnotamechanismwhichhadbeenofconcern.9.3TheresearchAsaresultoftheseconcernstheUKGovernmentdecidedtosupportaninvestigationoftheengineering,materialandfabricationneedsoftheproposeddeepwaterplatforms.TheDepartmentofTradeandIndustrythoughitsShippingandMarineTechnologyRequirementsBoardsetupacommitteecalledtheMarineMaterialsPanel,whichitselfsetupasub- 102Weldeddesign±theoryandpracticecommitteeknownastheStructuralSteelsWorkingParty.Thisworkingpartywaschargedwiththetaskofdefiningwhatrelevantknowledgeexisted,whatfurtherknowledgewasrequiredandhowthisknowledgeshouldbeacquiredthroughresearchprogrammes.During1972and1973thisworkingpartyexaminedexistingpracticesandresearchinprogressintheuseofheavystructuralsteelinotherindustries,particularlytheprocessplantindustryandfromthisdevelopedthebasisofpreliminaryrecommen-dationsforthedesignandfabricationofoffshoresteelstructures.AtthesametimetheworkingpartycommissionedTheWeldingInstitutetoundertakeastudytoidentifyareasofresearchwhichshouldbepursuedtoreducetheprobabilityoffracturesoccurringinoffshoresteelstructuresintheNorthSea.Theconclusionsandrecommendationsofthatstudywereset3738outinareportwhichwaslaterpublishedinsummaryform.Therecommendationsinthatreportleadtothedraftingofaresearchprogramme,whichcametobeknownastheUKOffshoreSteelsResearchProgramme(UKOSRP).Thiswastobeconductedtoexaminetheeffectofseawateronthefatiguelifeofweldedjointsinsteelandthefatigueperformanceoflargeweldedtubularjoints.Theprogrammewouldalsoexaminethefracturebehaviourofthetypesofsteelenvisagedsoastobeabletodevelopfracturetoughnessrequirementsofsteelstoavoidbrittlefracture.TheprogrammewastobefundedbytheUKGovernmentandinterestslookedafterbytheUKAtomicEnergyAuthorityasprojectmanager.ThefatigueandfracturetestingwastobeperformedbyTheNationalEngineeringLaboratoryatEastKilbrideandTheWeldingInstitutenearCambridgewiththesupportoftheHarwellCorrosionServiceoftheUKAtomicEnergyAuthority.Thisprogrammewasformallyapprovedin1973andthedesignandmanufactureoftestrigsandspecimenswasputinhand.ProgresswastobemonitoredbyanewcommitteecalledtheOffshoreSteelsSteeringGroupcomprisingrepresentativesofclassifica-tionsocieties,designers,operators,theBritishSteelCorporationandgovernmentdepartments.Theinitialadviceofthisgroupwasthatasignificantlyenlargedprogrammewouldberequiredtomeettheagreedobjectives.Thiswasapprovedin1975bytheOffshoreEnergyTechnologyBoardwhichhadbythensubsumedtheinterestsoftheShippingandMarineTechnologyRequirementsBoardinmattersofoffshoreoilandgasexplorationandproduction.Theburgeoningdevelopmentofthedeepwateroilfieldsrequiredthattheprogrammebedesignedwiththeaimofprovidingimmediateinformationonmaterialandjointperformancewhichcouldbeappliedtodesignintheshorttermaswellasofproducingbasicknowledgeandunderstandingforthelongtermdevelopmentofthetechnology.TheenlargedprogrammesawtheintroductiontotheworkoffurthercontractorsintheformofLloyd'sRegisterofShippingResearchLaboratory,AtkinsResearchandDevelopment,theUniversityofNottinghamandthe Offshorestructures103SpringfieldsLaboratoryoftheUnitedKingdomAtomicEnergyAuthority.VeryquicklytheprogrammeexpandedintoaEuropeanprogrammewithfinancialcontributionfromtheEuropeanCoalandSteelCommunityandwithco-operationbetweenresearchlaboratoriesinanumberofcountriesparticularlyNorwayandtheNetherlandswhichhadverystrongdomesticandcommercialinterestsinNorthSeaoilproduction.InternationalnetworksofresearcherssuchasSub-CommissionXV-EoftheInternationalInstituteofWeldingandtheWorkingGroupsofCIDECT(ComiteÂInternationalpourleDeÂveloppementetl'EtudedelaConstructionTubulaire,CommitteefortheStudyandDevelopmentofTubularStructures)enabledmoreandmorepeopletocontributefurthertotheunderstandingoftheparticularcharacteristicsoftubularjoints.TheUKGovernmentthroughitsDepartmentofEnergyrequiredthatplatformdesignsbeexaminedbyindependentbodies.TheseweretobecalledCertifyingAuthoritieswho,iftheyweresatisfiedwiththedesignandconstruction,couldissueacertificateoffitnessfortheplatformtobeoperated.TheseCertifyingAuthoritieswereprivatebodiesinitiallyconstitutedofconsultingengineerswitharangeofdisciplines.ToprovideabasisfortheassessmentoftheintegrityofdesignsofnewsteelplatformstheDepartmentofEnergycommencedthepreparationandpublicationof39GuidanceontheDesignandConstructionofOffshoreInstallations.TheresultsoftheUKOSRPworkwereusedtoenhancethisGuidance.Inparallelwiththiswerebeingdevelopednationalstandardsandcodesofpractice.IntheUnitedKingdomtheBritishStandardDraftforDevelop-ment,DD55,wasbeingprepared,eventuallytobecomeBS6235butwhichsufferedtheindignityofbeingwithdrawnalmostassoonasitwaspublishedforreasonswhichwerenotmadepublic.Theresultsofthousandsofhoursofintensiveworkbythemembersofthecommitteeandtheirsupportingstaffswerenullifiedbythisaction.TheNorwegianPetroleumDirectoratepublisheditsownregulations,whichfortubularjointscloselyfollowedtheUKGuidance,sharingtheirresearchinputandtheresults.IntheUnitedStatesAPIRP2AsectionontubularjointswasexpandedbyincludingfatiguedesigndatabasedontheresultsoftheUKOSRPprogramme.Inthefracturetoughnessfieldthework,mostlyconductedatTheWeldingInstitute(nowTWI),intheearlyandmid-1970sasanextensiontotheirpreviousresearchprogrammes,wouldleadtoCTODcriteriaforsteelandweldmetalsasusedinoffshoreinstallations.Weldmetalsinparticularrequiredconsiderabledevelopmenttoofferadequatefracturetoughnesswithoutresortingtoexoticcompositionswhichmightinthemselvescreateacorrosionproblemowingtodifferentialelectrochemicalpotentials.Therequirementssetdownforthematerialandweldmetalproperties,particularlyinthefieldoffracturetoughness,wereverydemandingandcouldbeachievedoftenonlymarginallybyextremelycloseadherenceto 104Weldeddesign±theoryandpracticeapprovedweldingprocedureswhichdemandedverycloselimitsonheatinput.TherequirementsasinterpretedbysomeNorthSeaoperatorsandcertifyingauthoritiesendedupplacingasubstantialcostandtimeburdenonfabricators.CTODtestingwasnotthenaregularserviceofferedbytesthouses.Itrequiredsophisticatedloadingandmeasuringequipmentandpersonneltrainedandexperiencedinthetestingtechniquesandtheinterpretationoftheresults.ExasperationinfabricatorswasinducedbytheirhavingtoproducePQRs(WeldingProcedureQualificationRecords)whichforsomejointsrantosome50pages!Thiswasalsothetimewhenwelddefectacceptancelevelsofaseverityseenbeforeonlyinpowerandprocessplantwerebeingappliedtocomplexthreedimensionaljointsinthicksteelplatewhentheonlymeansofdefectdetectionwasultrasonicexamination,whichatthattimewasnotsuchareliabletechniqueasitbecameinthe1990s.Storiesaboundoflengthsofweldsinthicksteelplatemeasuredintensofmetresbeingexcavatedbecauseofareportedwelddefectwhichwasfoundnevertohaveexisted.SmallwonderthatonewellknownandhighlyexperiencedScottishweldingengineer,exasperatedaftersuchanexperience,wasdriventosayataseminarin1974:`WhatIwantisaweldmetalsotoughthatitdoesnaematterifyoucanseethroughit!'Intheeventtheimprovementsinweldmetalandparentmetaltoughnessoverlateryearshavealmostmadehiswishcometruefortheconsumablemanufacturersinarelativelyshorttimewereabletoofferweldmetalswhichcouldwellsatisfytherequirements.9.4PlatformdesignandconstructionThecontractualarrangementsforconductingoffshoreprojectswereverydifferentfromthetraditionalcivilengineeringproject.Inthistheclientretainsaconsultingengineer,`theEngineer',todesignandsupervisetheconstruction.Heisentirelyresponsibleforitsexecutionandsuccess.Intheoffshoreindustrythemajoroilcompanies(orthatsubsidiaryofanoilcompanyassignedto`operate'thefield)frequentlyexercisedveryclosecontrolovertheworkofthedesignerswhowereinrealityjustdesigncontractors.Thedesigncontractorsthenhadnoconventionalresponsibilityforthespecificationandsupervisionofconstructionwhichwasexercisedbytheoilcompany'sengineeringstaffmanyofwhomwereextremelyexperienced.Muchoftheprogressinmaterialsandweldinginthe1960sand1970swasowingtotheirassiduousattentiontotheintegrityoftheirstructuresdestinedforthenorthernNorthSea.Whilstitmaybeinvidioustodefinethecontributionsofindividuals,itwouldbeacceptedbymanythattheenergiesofthelateHarryCotton,BP'sChiefWeldingEngineeratthetime,movedforwardmanyofthedevelopments,drivenbyhisdetermina-tionnevertoseethelikeoftheSeaGemtragedyagain. Offshorestructures1059.5ServiceexperienceTherecordofstructuralperformanceofthelargerfixedplatformstructureshasbeenverygood.Anumberofrepairshavehadtobemadebecauseofboatcollisiondamageordamageduetodroppedobjects.Therehavebeenrelativelyfewoccurrencesofearlyfatiguecrackinginmajorfixedplatformmembers;thosewhichhaveoccurredhavetendedtobeinhorizontalbracingnearthewaterlinewhichsufferedwaveactioncausingoutofplanebending.Techniquesofstrengtheningsuchjointsinsituweredevelopedandappeartohavebeensuccessful.Thereappeartohavebeennobrittlefracturesinfixedplatforms.Sucharecordiscauseforsomesatisfactionbearinginmindthenoveltyofmuchoftheconstructionoftheearlyplatformsandthelackofserviceexperience.Itmightbesaidthattheexperienceshowsthatdesignswereconservative,andifthatissothenitmustbeseenasagoodthing.Theexperienceformedasoundbasefordevelopinglowercostmethodsofoilandgasextraction.Howmuchbetterthanascenarioofhavingearlydisastersandthenhavingtoimprovedesigns. 10Managementsystems10.1BasicrequirementsItisgenerallyacceptedthatanybusinessactivitymustbeconductedwithinanoveralldisciplinewhichensuresthatwhatisneededorwantedisdefinedandthatactionstakentosatisfythatneedorwantareputinhandandexecutedinaneffectivemannerandattheappropriatetime.Thisrequiresthreebasicresources:.peoplewiththenecessaryknowledgeandskill.facilitiestoenablethosepeopletoexercisethatknowledgeandskill.inputsormaterialswhichcanmeettherequirementsofthejob.Withtheseresourcesitisthennecessarytohave:.aplanofwhatactivitiesaretobepursuedwithwhichmaterialandwhen.meansforconveyinginstructionstothepartiesandindividualsinvolved.meansofcontrollingactivitiesand/orofdemonstratingthattheiroutputsconformtotherequirements.Thesefeatureswillberecognisedashavingbeenpartofengineeringforaslongasithasbeenpursuedandsothereisnothingnewhere.10.2ContractsandspecificationsAcommercialcontractisusuallyanagreementbetweentwopartieswherebyonepartysuppliestoordoessomethingfortheotherpartyforaconsideration.Suchaconsiderationisusuallyapaymentinmoneyorsomeothernegotiabledevice.Whathastobesuppliedordone,inotherwords,whatistobebought,purchasedorprocured,whicheverwordisused,needstobespeltoutclearlyinadescription.Forsimplepurchasesthisdescriptionmaysimplybethenameofaproprietaryarticleandcommonbusinesspracticeistoenterintoasimpleagreementonthebasisofapurchaseorder.Thiswillsimplyaskthatacertainquantityofaniteminthecataloguebe Managementsystems107suppliedforapricewithinsomanydaysorweeks.Therewillbestandardconditionsofthesupplywhichmaybethoseofthebuyerorthesupplieroranegotiatedcompromise.Usuallythedescriptionoftheproductisthatofthesupplierasitappearsinhiscatalogueorbrochure.Suchamethodmaybeusedformorecomplicatedengineeringproductsbutonlywherethatproductismadeofstandarditems.Anexampleisasimplereactionorseparatingvesselwithvariousnozzles,allmadetostandardspecifications.Thepurchaserdescribeswhatisrequiredinadrawingorsketchcallingupstandardparts.Formorecomplicatedandone-offitemssuchasabuilding,craneorpowerstation,thepurchaseorderisaninadequateinstrument.Therearemanystagesinthedesignandconstructionofsuchaprojectandthemoneyispaidoutinaccordancewiththeworkdone.Evidenceofachievementandverificationofthequalityoftheworkhastobeallowedforasdoproceduresforaddressingdisputes.Theproductisaone-offandneedstobedescribedinaspecification;thereisnopriorproducttouseasasampleormodel.Atthispointwemayseeadivergenceintheapproachestospecificationwriting.Atoneextremethespecificationsaysonlywhattheitemisrequiredtodo;thisiscalledaperformancespecification.Forexample,astockyardcranetoliftxtonnestoaheightofymetresandcarryitforadistanceofzmetres;thiswillbeaccompaniedbythedetailsofthesite.Thecustomerleavesitentirelytothesuppliertodesign,manufactureandinstallthecraneandmakeitwork.Thestockyardownermayknownothingofthedesignofcranesoreventhelawsandregulationssurroundingtheirconstruction,althoughhewillhavetolearnaboutthoserelatingtotheuseofcranes.Hecanaskonlyforwhatheneeds.Theotherextremeiswheretheclientdescribestheitemingreatdetail,eventotheextentofprovidingconceptualordetaildesignsfromwhichthesuppliermayneedtodevelopworkingdrawings.Asanexample,aspecialistchemicalmanufacturingcompanymaydesignitsownprocessesanditschemicalengineersknowexactlywhattheirplantneedstobe.Theywillwriteaspecificationingreatdetail.Whichapproachisuseddependsgreatlyonthenatureoftheclient.Inpracticesuchextremesarerarelyfollowedforanumberofreasons.Thecranewillrequirefoundationsfortherailswhichwillrequireknowledgeofthesoilconditionsanditwillrequireanelectricalpowersupply.Theclientwillthenhavetoretainspecialiststodealwiththosematters.Theywillhavetoagreerequirementswiththecranesupplier.Intheendtheclientmayfinditeasiertoemployaconsultingmechanical/electrical/civilengineerfamiliarwithcranestoactonhisbehalf.Theengineer'sexperiencewilltellhimthattherearecertainthingsaboutcraneswhichhavetobespecifiedindetailforthatyardbecausethestandarditems,althoughacceptable,arenotthebest.Attheotherextremethechemicalcompanymaygetwhatitasksforbutthepricemaybehighbecausethespecifiedworkisnotthewaythatthesuppliersusuallygoabouttheirwork.Inadditionthereisadangerforthe 108Weldeddesign±theoryandpracticeclientthathisdetailedspecificationisdeemedtobeaninstructiontothesupplierwhich,ifitturnsouttocauseaproblem,mayinlawplacetheresponsibilityforanyconsequentlossesordamageontheclient.Thetrendsoverthepastfewyearstopartnershipsratherthantraditionaladversarialagreementshelpstoavoidthesetwoextremesbutitisstillessentialthatanagreedspecificationfortheworksbederivedasearlyaspossible.Theuseofaneffectivemanagementsystemwillensurethattheintentisrealised.10.3FormalmanagementsystemsMostclientsandcustomershavebecomelessandlesstolerantoflatedeliveriesandofproductswhichdonotperformasrequiredorasexpected.Amajormovetoreducelateandpoorperformingproductsbyformalisedmanagementapproacheswasmadeinthe1950sparticularlyinthecontextofprocurementforthedefenceandenergyindustries.Thecausesandsourcesofdelaysandpoorproductperformancewereanalysedandactionssetoutasamanagementsystemwhichwouldpreventthesehappening.Theseformedthebasisofvariousdocuments.TheCanadianStandard,Z299,initiallyissuedin1975,describedasystempreparedfornuclearpowerstationconstruction.Thiswasparticularlysuitedtositeconstructionasopposedtofactorymanufacture.IntheUK,BS5179,basedondefencestandards,wasissuedasaguidetotheevaluationofqualityassurancesystemsbutwaswithdrawninfavourofBS5750,QualitySystems,whichwasfirstpublishedin1979,andapparentlydirectedmainlyatmanufactur-ing.Thissetoutthemostimportantfeatures,orelements,ofamanagementsystemwhich,whenoperatedwithinacontractualsituationbetweenapurchaserandasupplier,wouldleadtodesignandproductionofengineeringproductsconformingtoaspecification.AfteraperiodthisstandardwasusedtoformthebasisoftheISO9000seriesofdocuments,whichhaveseenvariousamendmentsandadditionsinattemptstoapplytheoriginalsystemelementstonon-engineeringandunspecifiedproductsandservicesandtoothercontractual,orevennon-contractual,situationsforwhichtheywereneverintended.Howeverwithinthescopeofthisbookwedonotneedtoaddressthosepoints.Itisgenerallyagreedthattheachievementofanddemonstrationofconsistencyofconformanceoftheproducttothespecificationisanessentialrequirementofanyindustrialprocess.Inmanymanufacturingprocessespartoralloftheoperationsmaybeperformedbymechanicaldevices.Thesedevicescanbedesignedtoincludetheirowncontrolsystemssothatwithadequatemaintenanceandwiththecorrectinputofmaterials,barringexternaldisturbances,theycanbeusedtomanufactureproductstoahighlevelofaccuracyandprecision.Thereareclaimsbysomethatsuchpractices Managementsystems109renderfinalinspectionunnecessarywhilstotherswouldnotbesobold.Attherawmaterialandsemi-finishedmaterialsstagestheconformanceoftheproductcanbeassessedreadilybytestsonsamplestakenfromproduction.Oncethemanufacturingoperationshavemovedalongtomakingthefinalproduct,suchdestructivetestingisnolongeralwaysaviableapproachtocontrolexceptforlowunitcostormassproductionitemswhichcanstillbesampled.Evenso,postmanufacturinginspectionhaslongbeenconsideredaninefficientapproachtoqualitycontrolforitdetectsnonconformingitemsafteranythingcanbedonetocorrectthem.Theresultisthatthepartsthemselvesarewastedaswillbeothersmadeataboutthesametime.Amoreeffectiveapproachistocontroltheprocesssothattheparameterswhichaffecttheconformanceoftheproductaremaintainedwithinthelimitswhichhavebeenshowntoproduceconformance.10.4Weldedfabrication41Formalmanagementsystemshavebeenappliedtoweldedfabricationactivitiesforlongerthanmost.Aweldingprocedure,ortobemoreprecise,aweldingprocedurespecification,ismorethanwhatisgenerallyunderstoodasaprocedure.Itisastatementofthewholeinputtothemanufactureofaweldedjoint.Itdefinesthematerial(s)tobejoined,theweldingprocess,anyweldingconsumables,edgepreparationsandweldingposition.Alsoincludedareweldingconditionswhichinarcweldingmeanquantitiessuchasweldingcurrentorwirefeedspeed,voltageandelectroderunoutlengthorweldingtravelspeed.Preheattemperature,sequenceofweldruns,anypostweldheatingorheattreatment,andnon-destructivetestingareothermajorpointsinaweldingprocedure.Muchweldingisstilldonemanuallywhichgivesapotentialvariabilityintheresultsasisthecasewithanyprocessoperatedbythehandofman.Inamanualweldingoperationthewelder,thematerialsandtheequipmentareallpartofthemanufacturingsystem.Toacquireconfidencethattheproductwillconformtothespecificationthewelder,theequipmentandtheproceduretobeusedmustbeconfirmedasbeingcapableofmakingtherequiredproduct.Thisconfidenceisacquiredbycheckingthattheequipmentwillbeworkingwithinitsoperatinglimitsandbygivingthewelderatrialjointtomakewhichrequiresthesameskillsastheactualjob.Thistrialjointmaythenbenon-destructivelyexaminedbyradiographyorultrasonicsandthencutintopieces,someformetallographicexaminationandhardnesstests,othersformechanicaltestsofstrengthandductility.Samplejointsmayhavebeenmadewiththeweldingproceduretoconfirmthatitiscapableofproducingtherequiredjointcharacteristics.Toprovidepriorevidencethattheweldersandtheweldingproceduresarecapableofprovidingtherequiredfabricationandsoavoidtheneedfor 110Weldeddesign±theoryandpracticetestsforeveryjob,recordsofwelderapprovaltestsandweldingprocedureapprovaltestsarekeptbythefabricator.Theserecordsaremaintainedascertificatessignedbythetestingandwitnessingauthorities.Dependingontheinterestsofthepartiesinvolvedinthejob,weldertestsandweldingproceduretestsmaybewitnessedbythecustomerorsomeindependentthirdpartyonhisbehalf.Therearesomeindustrywideschemesinwhichindependentsurveillanceisundertaken.Thisisintendedtoprovideconfidencethatamanufacturer'spersonnel,equipment,organisationandoperationaresuchthatcustomerscanacceptthatthemanufactureriscompetenttodothework.Thisreplacesseparatetestsforeachcustomerwhichtheywouldhavetopayforthemselvesaspartofthecontract.Suchschemeshavetheirlimitationsandthedisplayofacertificateofconformancetosomemanagementsystemstandarddoesnotrepresentanyguaranteethatafirmoranindividualwillperformasrequiredinanyparticularsituation.Asinanyotherbusinesstheproductionofweldedfabricationsrequireseducated,knowledgeable,trainedandcommittedpeopleworkingwithinanappropriatemanagementsystem.Insummary,basicconfidenceintheweldedfabricationofaviabledesignisachievedbyhaving:.competentweldersasdemonstratedbywelderapprovalcertificates.relevantweldingproceduresasdemonstratedbyweldingprocedurespecificationswhichhavebeentested.competentinspectionpersonnelasdemonstratedbyrelevantcertificates.Mechanisedorroboticweldingoperationsstillrequiretheinputofaknowledgeable,skilledandqualifiedoperatorandweldingprocedurespecificationsstillhavetobepreparedandverified. 11Weldquality11.1WelddefectsWhyhavedefects?Acynicmightobservethattherecanbenootherbranchofmanufacturingtechnologywheresomuchemphasisislaidupongettingthingswrongandthenattemptingtojustifyitasthereisinwelding.`Welddefect'isatermalmostinseparablefromweldinginthemindsofmanyweldingengineers.Itseemstobeanessentialpartoftheweldingculturethatwelddefectsshouldbeproduced!Whatiscommonlymeantbyawelddefectissomelackofhomogeneityoraphysicaldiscontinuityregardlessofwhetheritdiminishesthestrengthordamagesanyothercharacteristicoftheweld.Weldingisoneofthefewfinalmanufacturingprocessesinwhichthematerialbeingworkedonexistssimultaneouslyatvariousplacesintwophases,liquidandsolid.Thistogetherwiththelargetemperaturerangeandthehighratesofchangeoftemperaturegivesrisetothepotentialforgreatvariabilityinthemetallurgicalstructureofthejointanditsphysicalhomogeneity.Itisaswelltorecollectthatdiscontinuitiesinmaterialsarenotnecessarilyundesirable.Indeedthestrengthofmetalsdependsupontheircontainingdislocationsonanatomiclevel,i.e.disturbancesorwhatmightbecalledinothercircumstances`defects'intheregularlatticeoftheatoms,withoutwhichmetalswouldhaveverylowstrength.Ithastoberecognisedalsothatwelddiscontinuities,betheylackoffusionorpenetration,porosityorcracks,donotnecessarilyresultinadefectiveproductinthelegalsensethatitisnotfitforitsstatedpurpose.Attemptshavebeenmadetogetroundthesituationbycallingthesefeaturesflaws,discontinuitiesorimperfectionsbuttheworddefectremainscommonparlanceeventhoughitisreallyratherirrational.Wewouldnotcallforapolishedfinishonasteelbarifanas-rolledorroughturnedfinishwereadequateforthejobithadtodo;wedonotcallthelattersurfacesdefective,wecallthemfitfortheirpurpose.Thisconceptoffitnessforpurposebegantogainacceptanceinrespectofweldedproductsinthelate1960swhen42Harrison,BurdekinandYoungshowedthatthecommonlyusedweld 112Weldeddesign±theoryandpracticedefectacceptancelevelsinsomestandardspecificationsandcodesofpracticewereextremelyconservativeandhadfrequentlyledtounnecessaryrepairsandreworkduringfabrication.Suchrepairworkwasnotonlyanaddedcostbutalsodelayedfabricationandevenwholeconstructionprogrammes.Furthermoretheconditionsunderwhichrepairshadtobeconductedresultedintherepairedweldsometimesbeingofpoorerperformancethantheoriginalso-calleddefectiveweld.Themethodsoffracturemechanicsweredevelopedtomakeitpossibletoreviewtheeffectofawelddefectintermsoftheservicerequirementswithintheprocessknownasengineeringcriticalassessment(ECA).Withinthewholedisciplineofmetallurgy,weldingmetallurgyisaspecialisedsubjectwhichisdistinguishedbyhavingtoaddressthebehaviourofmetalswheretherewere:.highratesoftemperaturechange.hightemperaturegradients.changingsolubilityofgasesinthemetal.smallvolumesofmetalrapidlychangingfromsolidtoliquidstateandbackagain.transferandmixingofmetalsandnon-metalsinacomplexgaseousandelectricalenvironment.Thesefeaturessetuprapidlychangingfieldsofstrainandtheresultingstressesaddtothephysical,metallurgicalandmechanicalcharacteristicsof43welds.Thefeatureswhicharecalledwelddefectscanbeattributedbroadlytotwomainsources±workmanshipandmetallurgy.Workmanshipdefectsarethosewheretheskillsofthewelderhavenotmatchedthedemandsoftheweldconfiguration.Examplesofsuchdefectsarelackofpenetration,over-penetration,lackoffusion,undercutandpoorprofile.Metallurgical43,44defectsarisefromthecomplexchangesinmicrostructurewhichtakeplacewithtimeandtemperaturewhenaweldisbeingmade.Theycanarisefromthenaturalcompositionofthesteelorfromtheintroductionofextraneoussubstancestothemetalmatrixwhichisincapableofabsorbingthemwithouttheinductionofhighstrainandconsequentfracture.Examplesofsuchtypesoffeaturearehydrogeninducedcracking,hotcrackingandlamellartearing.Notwithstandingtheirdifferentoriginstheoccurrenceofeachofthetwogroupsofdefectscanbeavoidedbypropermanagementofthefabricationoperations(seeChapter10).Bearinginmindthesubjectofthisbookweneedtoseeiftheoriginsofwelddefectscanlieinthenatureofthedesign.Ifweacceptthatitisthedesigner'sresponsibilitytospecifyboththematerialsandtheconfigurationoftheproductthenwestartwiththematerials.Anessentialpartofthespecificationisthatthematerialshallbeweldablewhateverwetakethatto Weldquality113mean.Itisatthispointthatthisconceptofthe`designer'startstorunintotroubleforweknowthattheselectionofthematerialisintimatelyboundupwiththechoiceoftheweldingprocessandtheweldingprocedureandviceversa.Similarlytheweldpreparationswillbedecidednotonlybythetypeofwelddeterminedbyoperatingrequirementsofthejointbutalsobytheweldingprocess,thepositioninwhichweldingisdoneandthesequenceofassemblyofthefabrication.Itisforthisreasonthatdesigndrawingsmaybefollowedbyfabricationdrawingsfollowedpossiblybyshopdrawingswhichwillthemselvescalluptheweldingproceduresorsummariesofthemcalleddatasheets.Designinthesecircumstancesisaniterativeprocessconvergingasquicklyaspossibletoasolutionwhichmeetstheprojectrequirements.Fromnowonifwespeakofadesignerwearereallyreferringtooneofanumberofpartieswithdifferentroles.Aswesaw,welddefectsoccurfromtwomainsourcesbetweenwhichthereissomeinteraction.11.1.1SomecommonworkmanshipbaseddefectsshowninFig.11.111.1.1.1LackofsidewallfusionThearcfailstomelttheparentmetalbeforetheweldmetaltouchesit.Themoltenweldmetalrestsagainsttheparentmetalwithoutfusingintoit.11.1Somecommonworkmanshipbaseddefects(photographsbycourtesyofTWI).(a)Lackofsidewallfusion. 114Weldeddesign±theoryandpractice11.1(b)Lackofpenetration.11.1(c)Undercut.11.1(d)Poorprofile. Weldquality115Thiscanoccurbecause:(a)thearcisnothotenoughforthethicknessofthemetal(b)thearctravelstooquicklyalongthejoint(c)thearcisnotdirectedattheparentmetal(d)theweldmetalflowsaheadofthearc,preventingitfromimpingingontheparentmetal.Only(c)iswithinthescopeofthedesigner'sinfluenceifaninaccessiblejointoranunsuitableweldpreparationpreventsthewelderfromdirectingthearcattheedgepreparation.11.1.1.2LackofrootfusionThearcfailstomeltthemetalattheroot,causingasimilarconditiontothatinlackofsidewallfusion.Thecausesaresimilartothoseoflackofsidewallfusion.11.1.1.3LackofrootpenetrationThewelddoesnotreachthroughthefulldepthofthepreparation.Thiscanarisebecause:(a)therootgapistoosmallfortheweldingconditions(b)therootfaceistoolargefortheweldingconditions(c)theweldermaynotbesufficientlywellpractisedortrainedinthetechnique,particularlyinpositionalwelding,asinpipes.11.1.1.4UndercutTheparentmetaliswashedawayadjacenttotheweld.Thiscanarisebecause:(a)theweldingcurrentistoogreatfortheweldingposition(b)thewelder'stechniqueencourageswashingoutoftheparentmetal.11.1.1.5PoorprofileTheweldsurfaceiserraticallyshaped,peaky,underfilledoroverlapstheparentmetal.Mainlycausedbywrongweldingconditions,lackofwelderskill,practiceordiligence. 116Weldeddesign±theoryandpractice11.1.2Somecommon`metallurgical'defectsshowninFig.11.211.1.2.1Hydrogeninduced(cold)cracking11.2(a)Somecommonmetallurgicaldefects(photographsbycourtesyofTWI).Theoccurrenceofcoldcrackinginsteelisafunctionofbothmicrostructureandhydrogendissolvedinthemetal;insimpletermsitcanoccurifthemicrostructureisveryhard,usuallyinamartensiticheataffectedzone,andthedissolvedhydrogenlevelistoohighforthishardness.Itiscalledcoldcrackingbecausewhenitoccursiswhenthemetalhascooledtoambienttemperature.Thistypeofcrackingcanalsooccurintheweldmetalbutthisislesscommon.Itispreventedbytwomeasureswhicharepartofnormalgoodpracticeinthedesignofweldingprocedures:(a)ensuringthatacombinationofpre-heatandweldingheatinputisusedsothattherateofcoolingoftheheataffectedzoneisnotsohighastoquenchittoahighhardness(b)minimisingtheamountofhydrogentakenupbythesteelbyensuringcleanmetalsurfaces(nogrease,paintormoisture)andusinglowhydrogenweldingconsumables.Inhigheralloysteelsitissometimesimpossibletoreducethehardnesssufficientlyandpost(weld)heatingisapplied.Thisallowshydrogen,whichmightotherwisecausecracking,todiffuseoutofthesteeloveraperiodofhours.Inmarginallyhardenablesteelsthesameeffectisachievedjust Weldquality117throughpreventingthesteelcoolingdownquicklyafterweldingbycoveringtheworkwithheat-proofblankets.11.1.2.2HotcrackingHotcrackingcanoccurintheheataffectedzoneasliquationcrackingwhenonbeingheatedbytheweldingarcnon-metallicsubstancesinthesteel(usuallysulphides)meltwhilstthesteelissolidandformlayersofweaknesswhichfractureunderthethermalstressesofwelding.Inweldmetalthisform11.2(b)Hotcracking.ofcrackingisknownassolidificationcracking;astheweldmetalcoolsdownthesteelsolidifiesleavingthenon-metallicsstillliquid.Thishasthesameresultasliquationcracking.Thepatternofitsappearancecanbeinfluencedbythefreezingpatternoftheweldmetal,sometimesappearingalongthecentrelineoftheweldasthelastareatosolidifyafterthemetalcrystalshaveformedinasinglerunweld.Theseformsofcrackingarepreventedbyattentiontothesulphurcontentofthesteelandweldmetal.Thereisabroadrequirementinsinglerunbuttweldsthattoavoidthecircumstanceswhereweldmetalhotcrackingmightbeaproblemitiscustomarytorestrictthedepthtowidthratiooftheweld.11.1.2.3LamellartearingMicroscopicislandsofsulphidesandothercompoundsareproducedinsomesteelsandwhenthesteelisrolledintoplatestheseislandsbecomeplateletsorlamellaeThosenearthesurfacecanbemeltedbytheheatofanyweldingandifthecombinationofheatandthermalstressesissufficient 118Weldeddesign±theoryandpractice11.2(c)Lamellartearingtheselamellaewillbecomepointsofweaknessandallowthesteeltofractureinaratherwoodylookingwayknownaslamellartearing.Theincidenceoflamellartearingcanbeincreasedbyhydrogencrackingwhichcanactasaneasystartingpointforatear.LamellartearingismostlikelytooccurunderTjointsinsteelwhichhasbeenrolleddownsufficientlythintoproducelamellñbutwhichisstillthickenough,incombinationwiththeotherpartsjoined,torestraintheincipientthermalcontractionswhichcansetuphighstressesatrightanglestothesteelsurfaceundertheweld.Inpracticethistendstomeansteelplateinthicknessesbetween20and50mm.Theriskoflamellartearingcanbeavoidedbyusingplatewhichhasbeenprocessedinsuchawaythatitdoesnotcontainthelamellae.Thisisachievedeitherbymodifyingtheshapeoftheinclusionssothattheydonotcauseplanesofweaknessorbyremovingthematerialofwhichtheyconsist.Thelatterapproachwasfoundtoencouragetheoccurrenceofhydrogencracking;previouslythenon-metallicinclusionshadactedas`sinks'forhydrogendrawnintothesteelandwithoutthemthehydrogenenteredthesteelmatrixandcausedtrouble.Thelevelofresistanceofasteeltolamellartearingisconventionallyindicatedbytheobservedreductionincrosssectionalareaofatensiletestspecimentakeninadirectionatrightanglestothesurface,theZdirectionasitisknown.Steelmakersoffersteelsgivingvariousguaranteedlevelsofreductioninarea,e.g.15%,20%or25%.Inpractice,asteelfromareputablemakercangivevaluesofupto75%.Whichlevelischosenforaparticularcircumstancemaycomefromspecialistexperienceoranapplicationstandard. Weldquality11911.2Qualitycontrol11.2.1QualityinweldedjointsThemeansofthecontrolofqualityaremanyandvariedbuttheyarealldirectedatensuringthattheproductmeetsthespecification.Specificationscanbeverytightlydefinedortheycanbeveryloose;theymaydealwithmanycharacteristicsoftheproductorjustafew.Aphrasecommonlyfoundinmoretraditionalspecificationsisgoodworkmanship;thishasnomeasurablemeaningandsoisverysubjective.Itistakentomeansomethingmadeinawaywhichhasbecomecommonlyacceptedintheindustryasachievablebyatrainedandexperiencedworkmanandgenerallymeetsitspurpose.Manysuchworkmanshipcriteriaareveryoldandincorporatesoundtechniquesdevelopedasaresultofperhapscenturiesofexperience.Theuseofsometechniqueshoweverisbasedonamisunderstandingofeverydayobservations;anexampleistheoftheardexplanationthatoneheatsasteelplatebeforeweldingwithagastorch`todriveoutthemoisture'.Thisisarrantnonsenseandarisesbecausepeopleseeingthewatervapourofcombustionintheflamecondenseonthecoldplatetakeitthatthewatermusthavecomeoutoftheplate.Confirmationthattheproductmeetsaworkmanshiprequirementlayswiththeopinionofthepersonchargedwithexaminingtheproductforacceptance,theinspector.Thisapproachcanworkquitewellinanindustrywithastableworkforcemakingsimilarproductsrepeatedlyoveralongperiod.Judgementismadeonthebasisofpastsatisfactoryperformanceandacceptancebythecustomer.Inthecurrentworldofmorefluidworkforcesandwherethereislesstoleranceallowableontheproductperformance,itmaybenecessarytodefinetheacceptabilityoftheproductbycertainmeasurableparameters.Thismaybeassistedbythecomparisonwithsamplesorreplicasofanacceptableproduct.Bothofthesemethodsofjudgingtheacceptanceoftheproducthavethedrawbackthattheyexaminetheproductaftertimeandmoneyhasbeenspentonit.Iftheitemisunacceptableithastoberejectedaltogetheroritmayberepairedbothofwhichactionsrepresentawasteofresourcesandmoney.Statisticalanalysisofinspectionresultsinmassproductioncandetecttrendsawayfromthedesiredproductcharacteristicsandtheequipmentcanbeadjustedtocorrectthistrend.Muchweldingworkdoesnotlenditselftothatapproachwhichreliesonidentifyingdiscreteitemsofproduction.Inparticular,manualweldingofalongseamhastobecompletedbeforeitcanbeinspected.Any`defects'thenhavetobeexcavatedandre-welded.Suchworkifitistobefreefrominterruptionsneedstheattentionoftrainedandcompetentweldersandwelldesignedjointswhichdonotrequireunusualfeatsofskilltoweld.Whenweldingwithmechanisedequipmenttheweldingconditionscanbe 120Weldeddesign±theoryandpracticesetupandautomatically,orevenroboticallycontrolledduringwelding.Thetrulyadaptivesystemwillmakeallowancesinweldingconditionsforvariationsinfitupastheymayaffecttherootpenetrationoralternativelythepenetrationwillbemonitoredandtheweldingconditionsadjustedaccordingly.Thisisthebasisofprocesscontrol,usedinmanyindustries,asameansofensuringthattheoutputconformstothespecification.Inaperfectworld,postweldinspectionwouldthenbeunnecessary;fewofuswouldhavesuchconfidenceorrelinquishtheopportunityofpassinganeyeoverthecompletedwork.Itiseasytoconcentratehardonthemeasurementofdetailandforgettocheckthatalltheweldsareintherightplaceoreventhereatall!Inweldedfabricationwelddefectsarenottheonlysubjectforqualitycontrol.Dimensionsandmaterialsarealsoimportanttothequalityofthefinaljobasinallengineeringworkbutbothareofparticularconcerninweldedconstruction.Theconceptofadimensionaltoleranceiswellestablishedandsuchtolerancesrepresentabandofdimensionsbasedonaspecifiednominalfigure.Thesedimensionaltolerancesarenecessaryforanumberofreasons.Aprimereasonisthatitisimpossibletomakeanythingtoanexactdimensionandhavingmadeittobeabletomeasureitexactly.Abatchofnominallysimilaritemscannotallbemadeexactlythesameandthemagnitudeofthetolerancewhichhastobeallowedtoaccountfordifferencesbetweenonenominallyidenticalitemandanotherisameasureoftheprecisionofthemanufacturingprocess.Thesmallerarethetolerancesallowedthefinerhastobethecapabilityandcontrolofthemanufacturingprocesseswhichimpliescost,areasonformakingtolerancesaswideasispossible.Clearlythetolerancescannotbesowideastopreventmatingpartsfromfittingproperly,e.gline-upofboltorrivetholes,fit-upofpartstobewelded.Bucklingstrengthmayplacealimitontheflatnessofplatesandstraightnessofcolumns.Tolerancesareappliedtoshaftsandotherroundpartswhichhavetofitintoholes.Oneendofthescaleofsuchtolerancesmayhavetogivearunningfittoallowrotationorslidingandtheotherendaninterferenceorforcefittoconnectpartsfirmly.Asanexample,thedimensionaltolerancesonthedimensionsofthepartsofamachinetoolmaybesetforreasonssuchas:.togivethemachineitscorrectoveralldimensions.sothatindividualpartswillfittogetherandwillbeinterchangeable.toprovideasealbetweenpartscontainingfluids.togivetherequireddegreeoffit,whichmaybeaforcefitforfixeditemsorarunningfitforjournals/bearings.totransferloadsuniformly.toprovidebalanceinrotation. Weldquality121Insteelfabricationstolerancesareplacedondimensionstorecognisethatitmaybeimpossibletofabricatebeamsandcolumnswhichareexactlystraightorplateswhichareexactlyflatorcylindersanddomeswhichareexactlytotherequiredshape.Thedimensionaltolerancesaresettoavoidinstabilityortokeepsecondarystressestowithindefinedlimits.Inaircraft,shipandcarbodymanufacturetolerancesareplacedonmaterialthicknesstocontrolweightandonaircraftandshipstolerancesondimensionstoprovideinadditiontotheabovefeaturesthenecessaryaerodynamicandhydrodynamicperformance.Inweldedfabricationsourcesofdimensionalvariationarethermaldistortionandresidualstress.Theweldingarcisapointheatsourceataveryhightemperaturewhichmovesalongthejoint.Thesequenceofheatingandcoolingwhichtakesplaceleadstoexpansionandcontractionofmetalthrougharangeoftemperaturesandstrengths.Theresultisthatinsomecircumstancesthereremainlockedinstresses,calledresidualstresses,andinothersdistortionfromtheoriginalordesiredshapeofthepart(Fig11.3).11.3Distortioninaweldedjoint.Evenbeforeaweldingoperation,partsmayhavetobesetsoastoneutralisethedistortionwhichisexpected.Asimpleweldbeadonathinplatewilldemonstratehowthermaldistortionmanifestsitself.Amulti-runweldmadefrombothsidesintroducesamorecomplicatedsequenceofheatingandcoolingwhichwillhaveitsowneffects.Weldingisnottheonlycauseofdistortioninmembers.Rolledsectionssuchasuniversalbeamscontainresidualstressesowingtothedifferentthicknessesofthesectioncoolingatdifferentratesafteritisrolled.Thisisoflittleconsequencewhenthesectionisusedcomplete.Howeveritissometimesconvenienttoslitasectionalongitsmid-linetomaketwoTsections.Veryoftensuchaslittingoperationwillreleasethebalancedresidualstressesandthesectionthenadoptsacurvedshape.Distortioncanbeexperiencedinothertypesofconstructionthanwelding.Rivetedaluminiumalloystructuresasusedinairframescandistortunderthebuildupofthelocalstrainsintroducedbyeachrivetsetting.Thesequenceofrivetinghastobeplannedtominimisethistypeofdistortion.Insomefabricationsdistortionmaybeaccepted,ascanbeseeninmanyshiphulls.Distortioncanappearduringmachiningasstressedlayersareremovedandeveninserviceasresidualstressesare 122Weldeddesign±theoryandpracticeredistributedwithtimeorbytheeffectofserviceloading.Thisisundesirableinsomeproductsandparticularlyinmachinetoolsorfixtureswheredimensionalaccuracyandstabilityareoftheutmostimportance.Toavoidsuchin-servicedistortionitisusualtothermallystressrelievesteelfabricationsbeforemachiningorevenatintermediatestages.Thisstressoreliefisachievedbyheatingcarbon±manganesesteelstosome580±620C,andholdingatthattemperatureforatimedependingonthethicknessofthesteel.Thisrelaxesthestressestoadegreebutwillnoteliminatethemaltogether.Alessfrequentlyusedmethodistovibratethefabricationthrougharangeoffrequencieswhichwilllocallyyieldouthighresidualstresses.Completionoftheprocessissignalledbythereductiontoaconstantleveloftherequiredinputenergyfromthevibrator,akintoacessationofhysteresis.Thistreatmentdoesnotaffectthemicrostructureofthesteelandsodoesnotofferthesamebenefitsasheattreatmentwhereimprovedresistancetobrittlefractureisrequired.Itisbasictotheengineer'sroletorecognisethatitiseitherimpracticable,unnecessaryornotcosteffectivetodefinenotonlydimensionsbutmechanicalpropertiesandothercharacteristicsofmaterialsorstructurestoexactlevelsofaccuracyorprecision.Tolerancesmaybebasedontheabilitytoperformameasurement,ontheconsistencyofrawmaterialsupply,onthecapabilitiesofmanufacturingprocessesorontheperformancerequirementsofthestructureinrelationtothecostofmanufacture.Steelmaybemadefromrawmaterialsorscrap,bothfromanumberofsources,andtheskillofthesteelmakeristoendupwithsteelofacompositionwhichmeetsaspecification.Clearlyhandlingandmixingtonnesofwhitehotliquidmetalmakesexactcontrolofcompositiondifficultandsotolerancesareplacedinsteelspecificationsnotonlyintermsoftheirchemicalcompositionbuttheirmechanicalproperties.11.2.2Inspectionmethods11.2.2.1VisualinspectionItmightseemsoobviousastonotrequiredescriptionbutthisisakeymethodofinspectionwithoutwhichtheothermethodsareblind.Itrequiresaqualifiedandexperiencedweldinginspectornotonlytoobserveacompletedweldbuttobeabletodiagnosetheconditionswhichhaveledtoitscondition.Thevisualinspectionwillrevealsuchfeaturesassurfacebreakingporosity,undercut,coldlapping,lackoffusion,cracking,lackofpenetration,over-penetration,poorprofileand,perhapsevenmoreimportant,thecompleteabsenceofaweld,whichisnotunknown. Weldquality12311.2.2.2MagneticparticleanddyepenetrantTherearetwowaysofrevealingthepresenceofcertainfeaturesatthesurfaceofametalwhichotherwisewouldbetoofineforthenakedeyetodetect.Inthemagneticparticlemethodthemetal(itmustbeaferriticsteel)islocallymagnetised;discontinuitiessuchascracksatornearthesurfaceconcentratethemagneticfieldwhichisthenshownupbyironpowderorasuspensionofironpowderinaliquidsprinkledorsprayedontothemetalandwhichisattractedtotheconcentratedfield.Thedyepenetrantmethodisusedonnonmagneticmetalssuchasaluminiumandstainlesssteel.Astrongdyeissprayedontothemetalandsoaksintoanycracksorothersurfacebreakinggaps.Thedyeremainingonthesurfaceiswipedawayandthesurfaceisthensprayedwithafinechalkemulsion.Thiswilldrawupanydyenearthesurfacesothatthepositionofcracksandsoonwillshowupascolouredlinesorpatchesinthewhitechalk.11.2.2.3Radiography(X-rays)Radiationpassingthroughanobjectstrikesasensitivefilmgivinganimagewhosedensitydependsupontheamountofradiationreachingthefilm.Thiswillshowupthepresenceofporosityorcracksinaweldaswellasvariationsinweldsurfaceprofile.Hollowsintheweldsurfacewillshowupdarkerthantherest;cracksandporeswillshowupevendarker.Acrackwhichliesinaplaneparalleltoorclosetothatofthefilmwillnotshowupwell(Fig11.4).ThesourceoftheradiationmayabeanX-raymachineor,forsiteuse,aradioactiveisotope.Themethodrequiresthatbothsidesofthesubjectbeaccessible,forthefilmononesideandthesourceontheother.Thefilmisdevelopedlikeaphotographicfilmandhastobeviewedinaspecialisedlightbox.Thefilmcanbestoredforaslongasisnecessary.11.2.2.4UltrasonicsAbeamofhighfrequencysoundisprojectedfromthesurfaceintothemetal;theechofromtheoppositefaceoranyinterveninggapisreceivedbackandthetimebetweentransmissionandreceptionismeasuredelectricallyanddisplayedonanoscilloscopescreen(Fig11.5).Atrainedoperatorcantellwhatsizeoffeatureasignalor`indication'representsandwhereitis.Thereisnorecordoftheexaminationexceptthatrecordedinwritingbytheoperator. 124Weldeddesign±theoryandpractice11.4Radiographofabuttweld.Thelackofrootfusionisnotasclearlyrevealedasthelackofsidewallfusion.Signalfromdefect11.5Ultrasonicexaminationofthebuttweld.11.2.3ExtentofinspectionTraditionalmassproductionreliedforqualitycontrolbyeitherinspectingeveryitemduringand/orafteritsmanufactureor,tosavetimeandcost,inspectingasample.Insomeproductsthesamplemayhavebeentestedtodestructionorcutuptoconfirmitsconformancetothespecification.Statisticaltechniquesareusedtodefinetherateofsamplingorthesamplesizetoachievetherequiredlevelofconfidence.Ifthesamplesshowthatoveraperiodoftimeaproductcharacteristic,e.g.diameter,ismovingtowardsatolerancelimit,perhapsbecauseoftoolwear,themachinecanbere-settogiveadiameterattheotherendofthetoleranceband.Inmoresophisticatedcircumstancestheactualparametersoftheprocesswillbecontinuouslymonitoredandadjustedtokeepwithinthelimitswhichwillgiveproducts Weldquality125withintheirownlimits.Clearlysuchsophisticatedmethodscanbeappliedtodiscretemechanisedweldingoperations,suchasresistancewelding.Theirapplicationdoesofcourserequirethatthepartstobejoinedarethemselvescontrolledinthicknessandfit-up.Samplingisvalidonlywherethereisabasicallyrepetitivemanufacturingoperation;instatisticalparlanceeachitemmustcomefromthesamepopulation.Owingtovariationsinmaterialcomposition,fit-up,arclengthandsoon,manualormechanisedarcweldingmayproducedefectswhich,exceptunderconditionsofgrossmalpractice,appearscatteredonamoreorlessrandombasis.Thisisverydifferentfromthegradualdeviationofadimensionfromanominalsize.Thepracticeofsamplingisoftentobefoundinspecificationsforarcweldedfabricationunsupportedbyanystatisticalbasis.Forexamplethespecificationmaysaythat`10%ofweldswillberadiographed'.Thismaygiveresultsofsomesignificanceiftheweldingismechanisedsothatitcanbesaidtobeallofonepopulation;evensothereneedstobeclarificationastowhetherthisisintendedtomean10%ofwelds,i.e.oneweldinten,or10%ofwelding,i.e.onetenthofeachweldlengthoronetenthoftheweldlengthchoseninarandommanner.Manysuchspecificationsfailtosaywhatactionistobetakenifunacceptablefeatures,`welddefects',arefoundinthe10%.Certainlymostofthemsaythatsuchdefectsmustberepaired.Thisthenleavesthequestionastowhathappenstotheremaining90%oftheweldforifthereisadefectin10%thereissomeprobabilityoftherebeingdefectsintherest.Somespecificationsdealwiththatlikelihoodbyrequiringthatallremainingweldsbe100%examineduntilthecauseofthedefectshasbeenascertainedandresolved.Othersseemtoleavethatpossibilityunconsidered.Ofcourseevenifnodefecthasbeenfoundinthe10%,thereisperhapsstillachancethattherewillbedefectsintheremainder.Somespecificationstrytokeepmanualweldsdividedintopopulationsbycallingfor`x%ofeachwelder'sweldstobeexamined'.Howeverthediscoveryofadefectinoneweldinamanualprocessmaynotberelatedtotheremainderofthesamewelder'sweldsand,again,noactionisusuallydefined.Thebestthatcanbesaidaboutsuchapproachesisthattheydetectgrossmalpracticeandthattheymaycreateanenvironmentinwhichweldersknowthatthereisachanceofanydefectsbeingdiscovered.Theworstisthattheydonotofferanytruelevelofconfidencethattheworkconformstothespecification.Inneithercasecanthisbesaidtobeasatisfactorywayofgoingaboutqualitycontrolyettheseclausesarestilltobefoundinspecificationsforimportantstructuralworks.Thedrivetoavoid100%examinationderivesfromadesiretosavecosts.Howeverithastoberecognisedthatthisisnotalwaysthecostofinspection,whichmaybequitetrivial,butinthedesiretoavoidgreatercoststhroughhavingtorepairthedefectswhichmightbefoundby100%examination. 126Weldeddesign±theoryandpracticeProjectspecificationsoftenrevealawoefulignoranceofweldingandnondestructivetestingintheircompilers.Typically:`...buttweldsshallbeultrasonicallyexamined....filletweldsshallbeexaminedbymagneticparticle(MPI)...'Theinferenceisthatthesearealternativemeansofexamination.Thisispatentlynotso,sincebothtypesofweldrequirevisualexaminationandMPItodetectsurfacedefects.Ultrasonicscanbeusedtodetectsub-surfacedefectsbutthisisfeasibleonlyonbuttwelds.FilletweldsareanunsuitablesubjectforultrasonicsexceptbyspecialisedmeansandshouldbeexcludedfromultrasonicexaminationpurelyonfeasibilitygroundsandnotbecauseMPIisasubstitute.Inspectionlevelsmustberelatedtothenatureoftheproduct,themethodofweldingandtheoverallstructureofthemanagementsystem.Qualitycontrolismosteffectivewhenexercisedontheinputsratherthanontheoutputs.11.3WeldedrepairsTheconsequenceoffindingafatiguecrackorothertypeofdamagewillbeaneedtodecidewhethertoscraptheitem,repairitoruseitasitis.Thisdecisionwilldependonthecostandfeasibilityofrepairagainstthesupplyofanewitem.Aquestionmayberaisedastowhetherarepaireditemwilllastaslongastheoriginal.Tosomeextentthiswilldependontheaccessibilityofthedamagedareaandsothequalityoftherepairwhichcanbemade.Ifafatiguecrackedfullpenetrationweldisreplacedwithapartialpenetrationweldbecauseoflackofaccessthentherepairisunlikelytogivethelongevityoftheoriginal;itmightindeedjustbeawasteoftimeandefforttoattempttherepair.Asinallweldingtheweldersmustbesuitablyqualifiedandthewholerepairproceduremustbeplannedindetailand,wherenecessary,verifiedbytesting.Forsomelargeandcostlyplantithasprovedjustifiabletocreateareplicaormock-upofthejointareaonwhichtodevelopspecialpurposeequipmentandtechniquesandtocarryouttrialsbeforeembarkingontheactualrepair.Inrepairingaweldthereisnoreasontobelievethattherepairedweldneedbeinferiortotheoriginal.Weldedrepairsinoriginallyunweldedareaswill,ofcourse,havethecharacteristicofaweldedjointinthatmaterial.Therepairmustcommencewithremovingallthedamage,confirmedifnecessarybyappropriatenondestructiveexaminationmethods,andanyassociateddistortioncor-rected.Onsometypesofmaterialsuchasalloysteelsitmaybedesirabletoremoveallofanyexistingweldandheataffectedzonesastheirpropertiesmaybeaffectedbymultiplethermalcycles.Asforallweldingwork,thesurfacestobeweldedandadjacentareasmustbecleanedofanypaintorothersubstancewhichhasaccumulatedduringserviceorduring Weldquality127anyrecoveryoperation.Suitableedgepreparationsmustbedesignedandexecutedandanyarrangementsforpre-heatinstalled.Weldingcanthenproceedasintheprocedureandanyinterpasstemperatureandpostweldcoolingratescontrolled.Postweldinspectionwillthenbeconducted,preferablybeforeanypostweldheattreatment.Thisinspectionwillrequirespecialplanningifanelevatedpostweldtemperaturehastobemaintaineduntilpostweldheattreatmentisstarted.Forinsiturepairsthesizeoftheitemorrestrictedaccessmayrequirealocalheattreatment.Equipmentgivingcontrolovertemperaturegradient/timeaswellastemperature/timemaythenberequired.Finalinspectionwilltakeplaceafteranagreedtimeaftertheitemcoolstoambienttemperature.Thisisacommonprocedurewithsteelstoallowtimeforanydelayedcrackingtooccur.Thispracticearosefrompastexperience,withthicksteelsinparticular;instanceshadoccurredwhenafabricationhadbeeninspectedandpassedbutwaslaterfoundtocontaincracks.Hydrogeninducedheataffectedzonecrackinginsteels,alsocalledcoldcracking,isknowntooccursomehours,orevendaysinthicksections,aftercoolingoncessationofwelding.Somepartiesheldthatthiscouldhavebeenbecausetheinspectorshadmissedthecracksinthefirstplace,whichmayhavebeentrueorjustuncharitable!11.4EngineeringcriticalassessmentLookingbackatman'srecentuseofironandsteelwefindthatin1854WilliamArmstrongembarkedonthedesignofarifledgunasareplacementforthecumbersomefieldgunswhichatthattimefiredcastironballs.JamesRendel,famousforhisworkincivilengineering,encouragedArmstrongtoconsidersteelinplaceofwroughtiron,atransitionwhichhadcommencedincivilengineeringsomeyearspreviously.Armstrongagreedthatsteel,havingamuchgreaterstrengththanwroughtiron,shouldbethebettermetalforstandinguptothepressureinthebarrels.Howeverexperimentsconvincedhimthatinresistingexplosiveloadstensilestrengthwasnotthecorrectcriterion.Hethereforeadoptedthetechniqueusedinmanufacturingsportinggunswhosebarrelsweremade`bytwistinglongslipsofironintospiraltubesandthenweldingtogethertheedgesbywhichmeansthelongitudinallengthoftheslipsbecomesopposedtotheexplosiveforce'.ArmstronghadagreatrivalinthepersonofJosephWhitworthwhoproposedagunmadefromlargeforgings,inatotalcontrasttoArmstrong'sslipmethod.Armstrong'scommentwas:`TomakelargegunsontheprincipleofsolidforgedtubeseitherofsteelorironIconsiderentirelyoutofthequestion,becausewecanneverpenetratetheinteriorofthemasssoastodiscovertheexistenceofflaws.'AlfredKruppinGermany,ofthesecondgenerationofthatdynasty,wasofcourseanaturalrivalofArmstrongand 128Weldeddesign±theoryandpractice11.6Structuralfailureinarailwaybridge.twoyearsyoungerthanhim.In1863ArmstrongwrotetoStuartRendel,oneofJamesRendel'sthreesonsandArmstrong'smanagerinLondon,reportingthatoneofKrupp'sgunshadburst,`...flyingintoathousandpieces.Allthefragmentsweresoundsothatthefailurewaspurelyduetotheintrinsicunfitnessofthematerial.'(Stuart'sbrotherGeorgewasmanageroftheordnanceworksofSirWGArmstrong&Companyandhisotherbrother,Hamilton,wasresponsiblefortheengineeringdesignofTowerBridgeinLondon.)AswesawinChapter2,castironwasknowntobesusceptibletofractureandtherewereanumberofinstancesofcatastrophicfailureofrailwaybridgesintheVictorianeraofwhichanexampleisshowninFig.11.6.ToquotefromTheIllustratedLondonNewsof9May1891:`ThedisasteronFridayMay1,attheNorwoodJunctionStationoftheLondonandBrightonRailway,fromthecollapseoftheironbridgeoverPortlandRoad,whenanexpresstrainwaspassingoverit,mighthavehaddreadfulresults....Therewasanundiscovered``latentflaw''inoneofthegirdersofthebridge,whichoughttohavebeenreconstructedlongsince,asitgavewaybeneathapilotenginefourteenyearsago.'Suchfailureswerearesultofthepoortensilepropertiesofcastironinconjunctionwithdefectswhichweremoreorlessacceptedfeaturesofcastingatthetime.Chapter2describeshowfromtheexperienceofthesefailuresaroseanumberofbridgedesigns Weldquality129employingcastironinthecompressionmembersandwroughtironinthetensionmembers.ThelattermaterialwasnotwithoutitsproblemswhichwerealludedtobyIKBrunelinalettertothecommissionappointedtoenquireintotheapplicationofirontorailwaystructures.InhisletterBrunelwrote:`Whowillventuretosaythatifthedirectionofimprovementisleftfree,thatmeansmaynotbefoundofensuringsoundcastingsofalmostanyform,andiftwentyorthirtytonesweight,andofaperfectlyhomogeneousmixtureofthebestmetal?'Brunel'svisionexiststodayinexamplesofcaststeelnodesusedinsomeoffshorestructures.Theseaccountsdemonstratethattheleadingengineersofthetimewereawarethatmetalsneededmorethantensilestrengthtosupportloadsandthatmaterialflawscouldaffecttheintegrityofastructure.Eventodaymostconventionalstructuralengineeringdesignproceduresassumethatthematerialandthejointscontainnorandomimperfectionswhichwouldpreventstructuresmadeofthemfailingtoperformtheirtaskandthatthemechanicalpropertiesareentirelyuniformthroughoutthematerial.Earlierinthisbookwesawthatweldingprocesses,particularlyintheirmanualforms,aresubjecttovariationsinbehaviourwhichcanresultinunplannedvariationsandevendiscontinuitiesinaweldedjoint.Earlierchaptersshowthattocopewiththissituationtherearewelddefectacceptancestandards,workmanshipstandards,whichhavegrownoutofcommonpractice.Theseoftenrepresentwhatisachievablebygoodpracticeordefectswhichcanbeeasilydiscoveredbutareinnowayrelatedtotheeffectofanyweldflawsontheintegrityofthestructure.Theirvalidityrestsonthepastsatisfactoryuseoftheminconjunctionwithcontrolledmaterialpropertieswhichagainhavenotheoreticalrelationtothetoleranceoftheallowablewelddefects.Inrecentyears,inspectiontechniquesandoperatortraininghaveimprovedsothatitispossibletodefinetheshape,orientationandsizeofaninternalweldormaterialflawfarmoreaccuratelythaninthepast.Everynowandthenanengineerisfacedwiththeproblemofwhattodowithafabricationdefectwhichislargerthanthespecificationwouldallowbutwhoseremovalwouldbedifficultorexpensive.Anothermatteriswhenacrackisfoundtohavedevelopedinserviceanditisnecessarytodecideifthecrackwillreducetheintegrityofthestructureandwhetheritislikelytogrow.Inbothofthesesituationstheengineerhastodecideifthestructureisfitforitspurposeinthepresenceoftheflaworcrack.Theengineercanthenturntoaprocedureknownasengineeringcriticalassessment(ECA).Whenappliedtothetolerationofwelddefectsthisinvolvesmakinganassessmentoftheeffectoftheflawontheintegrityofthestructure.Thisassessmentismadebyanalysingthewayinwhichthepresenceoftheflawmodifiesthelocalstressfieldandaffectsthepotentialforthepropagationofcracksbybrittlefracture,fatigue,stresscorrosioncrackingandsoon.Theapproachmakesuseoftheconceptoffracturemechanicswhichwasoriginallypostulatedby 130Weldeddesign±theoryandpractice45GITayloranddevelopedbyAAGriffithforexplainingthebehaviourofcracksinbrittlematerials.Ithassincebeenextendedtobeapplicabletocrackbehaviourundernon-linearstress/strainconditionssuchasexistinelastic/plasticmaterialsincludingsteels.Thetheoryandthetechniquesarequitesophisticatedandthesatisfactoryusageofthemethodsrequiresafundamentalunderstandingofthebasisoftheconceptsandtheirinherentunderlyingassumptions.Forthisreasontheiruseisbestlefttothefracturemechanicsspecialists.Howeverintheearly1980sitwasrecognisedthatthiswassuchapowerfulandpotentiallybeneficialtoolthatguidanceontheuseoffracturemechanicsinassessingweldedjointsinrespectoffatiguecrackingandbrittlefractureshouldbemadeavailablepublicly.SuchguidancewaspublishedintheUKasBritishStandard,PD6493.AcomparabledocumententitledTheFitnessforPurposeofWeldedStructureswaspublishedbytheInternationalInstituteofWeldingin1990asadraftfordevelopmentbutneverpublished.AnamendedversionofPD6493was46issuedin1991andadevelopmentofthiswaspublishedin1999asBS7910.CEN,theEuropeanstandardsbody,throughitsTechnicalCommittee121,isplanningtoissuethesamedocumentasoneofitsTechnicalReports.Althoughtheyshouldnotbeusedastextbooks,suchdocumentsasBS7910representacondensationofknowledgeofandexperienceintheapplicationofdefectassessmentmethods.Theyaredesignedtobeusedbypeoplewithsomebackgroundknowledgeoffracturemechanicsandaregenerallyconservativeintheirresults.Nonethelessitisessentialthattheuserensuresthattheinformationusedinderivingadecisionontheacceptabilityofacertaindefectisreliable.Thetwomostcommonformsoffractureagainstwhichwelddefectsareassessedarefatigueandbrittlefracture.BS7910setsouttheassessmentprocedureinanumberofsteps.Foranassessmentforbrittlefracture,aknowledgeofthefracturetoughnessofthematerialsurroundingthedefectisrequired.Thiscanbeasacriticalstressintensity,Kc,oracriticalCTODtypeofmeasurement,dc.SeverallevelsofassessmentareofferedinBS7910inincreasingdegreesofconfidenceaccompaniedbyincreasingcomputationandanincreasingneedforaccuratematerialsandstressdata.Forassessmentofawelddefectinrespectoffatigue,twoapproachesaregiven.OneequatestheeffectofthedefectwiththefatigueperformancecategoriesinBS7608.Theotherrequiresthecalculationofthehistoryofthecrackfrontgrowthbyaniterativeprocedure.Thiscanbequiteacomplexandtimeconsuming47exercise.Aswehavecometoexpectthereissoftwareavailablewhichcanperformthesecalculations. 12Standards12.1WhatwemeanbystandardsTheword`standard'asitiscommonlyusedinengineeringisacontractionofstandardspecification.Thisisaspecificationforamaterialormanufacturedproductwhichmaybewrittenbycompaniesforinternaluse,andbynationalandinternationalbodiesforpublicuse.Theword`standard'alsoreferstostandardproceduressuchasexaminationsandtestsofmaterialsandpersonnel.Thereareothertypesof`standard'inadifferentcontext,forexamplethestandardmetreisthebasicmeasureoflengthwhichwasoriginallyrepresentedbythelengthofaplatinumbarkeptinParis.Suchbasicstandardshavebeenreplacedbymoreesotericmeasuressuchasthedistancetravelledbylightinavacuuminacertaintime.12.2StandardspecificationsThesehaveanumberofpurposes.Atasimpleleveltheiruseminimisesthecostofproductionandmaintenanceofengineeringproductsthroughthereductioninvarietyandtheresultinginterchangeabilityofsimilarparts.Anhistoricalexampleoftheeffectoflackofstandardisationwasinscrewthreads.Untilthe1960ssomecountriesusedseveralthreadsizesquotedinincheswhichincludedsuchformsasWhitworth,BritishStandardFine(BSF),BritishAssociation(BA),UnifiedCoarse(UNC)andUnifiedFine(UNF).SomemanufacturersevenhadtheirownthreadssuchaswereusedintheBSA(BirminghamSmallArms)bicycle.TheownerofoneofthesebicycleshadtomakesurethatanyreplacementnutsorboltsweretotheBSAthread.Eventuallymetricsizeswereadoptedbymostcountriesandmattersbecamemucheasiertomanagebothinfactoriesandinthecustomers'maintenancedepartments.Thereductioninvarietyofferedlowercoststhroughincreasedproductionrunsofparts,reducedstockholdingsoffinishedpartsandfewertypesoftoolsusedinmanufactureandmaintenance,forexampletaps,diesandspanners.Standardpartsalsosavedesigntimewhethertheybefastenings,couplingsorcableterminations, 132Weldeddesign±theoryandpracticebecauseallthatneedstobedoneistocallupthestandardnumberasthepartnumberontheassemblydrawing.Productsmadetoastandardspecificationcanbeusedwithothersindifferentcountries,regionsorcontinents.Wecanbuythesametorchbatteries,photographicfilmsandcarfuelovermostoftheworldthankstostandardsbutstrangelyenoughwestillcan'talwaysusetheplugonourelectricalequipmentallovertheworld.Inthebusinessofinformationtechnologywestillhavetheridiculoussituationwherewehavemutuallyincompatiblesoftware.ThetextofthisbookcannotbesimplytransferredtosomeotherPCwhichdoesnothavethesamewordprocessingsoftwareunlessthereisaconversionprograminstalled.WecannotevencopytexttoafloppydiskandautomaticallyexpecttobeabletorunitonadifferentmakeofPCevenwiththesamesoftware.The`floppy'diskitselfisidentifiedbyitsdiametermeasuredininchesandnotmillimetreswhichistheinternationalunitbecauseintheUSA,wherethefloppydiskwasfirstmarketed,theinternationalsystemhasnotbeenimplemented,abouttheonlycountryintheworldnottodoso.Perhapsinafewyearsstandardisationwillreachtheinformationtechnologyindustry;thepotentialsavingsintimeandcostwillevennowbeevident.Asrecentlyas1999aspacecraftfailedtolandintactontheplanetMars,evidentlybecauseinstructionshadbeengiveninmilesandnotkilometres,atrulycostlyexampleoftheconsequencesofafailuretousestandards.Asecondpurposeofastandardistodescribeaproductwhichhasaspecificlevelofperformance.Thisisofparticularimportancewhen,forexample,astandardspecificationincludesprovisionsforsafety.Standardscanensureconsistentperformanceofengineeringproductswhichrequiretobedesignedtoaparticularphilosophyonabasisofreliableperformancedata.Suchdesignmethodsanddataoftencomefromavarietyofsourcesandoveralongperiodoftime.Thefunctionofastandardspecificationisoftenthentoprovideadigestofthisdatawhichwillhavebeenassessedforvaliditytoensurethatitcanbeusedreliablywithinthecontextofthestandard.Examplesareseeninproductstandardssuchasthoseforbridges,chemicalprocessplant,offshoreplatformsandcranes.Suchstandardsaremoredifficulttocompilethanadescriptionofapartsuchasaboltandcannotspecifytheeventualproductasaphysicalitem.Theyhavetobemorecorrectlythoughtofascodesofpractice,leavingtheengineerfreetodesignandmanufacturetheproductashethinksfitwhilstconformingtotheintentofthestandard.Astandardmustbewrittennotonlysoastodefinethecharacteristicsoftheproductbuttodefinehowthatproductwillbedemonstratedtoconformtothestandard.Thisisfeasiblewhere,forexample,materialcompositionandstrengthorthedimensionsofascrewthreadarespecified.Howeverwhenwegettosomethingaslargeasabuildingorabridgehowdowe Standards133demonstrateconformancewiththestandard?Onewayofcourseistoperformacompletelyseparatesetofcalculationsfromtheoriginaldesigncalculations.Anotheristoputloadsonitandmeasurethestressesortomeasurestressesinservice.Itmustbesaidthatstandardsshouldbeusedonlyasasupportforgoodengineeringandnotasitsbasis.Standardsarederivedjointlybythepartiesinterestedinmakingandusingtheproductaswellasbyothersandthetimetakentoprepareandpublishastandardmeansthatitcannotbebasedonthemostuptodatetechnology.Theresultcanthenrepresentonlyrarelyanythingexceptthelowestspecificationacceptabletothoseparties.Inanyparticularapplication,thestandardalonemaynotrepresentalltherequirementsofthecustomerorthemanufacturerandthecreationofasoundengineeringproductrequiresthatitbespecified,designedandmanufacturedbycompetentengineers.Thereisaviewthattheavailabilityof,andadherenceto,detailedstandardsisnotnecessarilybeneficialbecause,aswehaveclaimedabove,engineerscanusethemasdesignaidsinsteadofseekingnewsolutionstorequirementstherebydiscouragingthedevelop-mentoradoptionofmoreadvancedapproaches.Furthermoretheexistenceofdetailedstandardsmakesitpossibleforpeopleoflittlefamiliaritywiththesubjecttoattempttodesignandmanufactureproductsaboutwhichtheyknowlittle.Itmustbeemphasisedthattheapplicationofastandardrequiresthattheuserunderstandthecircumstancesforwhichitwasprepared;itcanbeverydangeroustousestandardsinignoranceoftheirderivationandscope.Inthefieldofweldedfabricationtherearemanystandardsdescribingthematerials,weldingconsumables,weldingplant,themanagementofweldingoperations,inspectiontechniquesandproceduresandthefabricatedproductitself.Naturallymanyofthesestandardswillbecalledupbymanufacturersandcustomersintheirproductspecifications,unfortunatelynotalwayswithadequateknowledgeoftheirscopeandcontent.Thereareinexistencemanyin-housecompanyspecificationswhichhavebeenusedforyearsduringwhichtimetheymayhavebeenamendedbypeoplewithoutspecialistweldingknowledgetosuitnewjobsandforwhichtheoriginallyquotedstandardsareinappropriateorwhichevenconflictwiththebasisofthedesign.Thisisacircumstancewheretheweldingengineerwillbeneededtoadviseontheinterpretationoreventherewritingofthespecification.Itisanunfortunatefactoflifethatthefirstactofmanywritersofprojectspecificationsistoreachforthelistofstandards.Thisshouldofcoursebethelastthingthattheydo.Thefirstandmostimportantmattertobedecidedisthebasisofthedesign.Thatistosaywhattheproductisintendedtodo,inwhatwaywillitdothatandwhatmeansofrealisationwillsatisfythat.Inpractice,anddependingontheparticularindustry,thespecificationwillbemoreorlessdetailed,andwillcallupsuchstandardsasare 134Weldeddesign±theoryandpracticetechnicallyappropriateorasarerequiredbylegislation,thecustomerorotherauthorities.Projectspecificationswillgothroughseveralstagesaswillthedesign;forexampleincivil,structuralandotherheavyengineeringthesestagesmayincludeafeasibilitystudy,conceptualdesign,designspecifica-tion,detaildesign,fabricationspecification,fabricationorshopdrawings,designreportandfinallytheas-builtrecords.Therearestandardswhichareapplicabletoallthesubjectsofthechaptersinthisbookfromproductstandardssuchasbridges,buildings,cranesandpressurevessels,weldingmaterialssuchasweldingequipmentandelectrodesandtechniquessuchasnondestructiveexamination.OnaninternationalscalestandardsarepublishedbyISOandIEC;onaregionalscaletherearestandardssuchasthosepublishedinEuropebyCENandCENELEC.Onanationalscaletherearestandardspublishedbynationalstandardsbodiesaswellasbyprofessionalinstitutions,commercialbodiesandindividualmanufacturers.Ingeneralaccesstointernationalandregionalstandardsinanycountryisthroughthatcountry'snationalstandardsbody. References1NewmanRP,`Trainingforweldingdesign',ImprovingWeldedProductDesignConference,Abington,TheWeldingInstitute,1971.2CottrellA,AnIntroductiontoMetallurgy,2ndedn,London,EdwardArnold,1975.3Traininginaluminiumapplicationtechnologies,TALATCD-ROMVersion2.0,Brussels,EuropeanAluminiumAssociation,1999.4Glossaryforwelding,brazingandthermalcutting,BS499:Part1:1991,London,BritishStandardsInstitution.5Welded,brazedandsolderedjoints±symbolicrepresentationondrawings,ISO2553,Geneva,InternationalOrganisationforStandardisation,1992.(AlsoasEN22553publishedundertheirownprefixesbythenationalstandardsbodiesofCEN.)6Welding,brazingandsolderingprocesses±vocabulary,ISO857,Geneva,InternationalOrganisationforStandardisation,1990.7Standardweldingtermsandsymbols,A3.0,Miami,AmericanWeldingSociety.8TsaiKC,ChenC-Y,`Ductilesteelbeam-columnmomentconnections',Proc,IIWAsianPacificWeldingConference,Auckland,NewZealand,1996.9`Designrulesforarc-weldedconnectionsinsteelsubmittedtostaticloads',DocXV-358-74,IIW,1974(unpublished).10ClarkPJ,`Basisofdesignforfillet-weldedjointsunderstaticloading',ImprovingWeldedProductDesignConference,Abington,TheWeldingInstitute,1971.11Steelstructures,Part1:Materialsanddesign,ISO10721-1,Geneva,Interna-tionalOrganisationforStandardisation,1997.12WohlerA,`Teststodeterminetheforcesactingonrailwaycarriageaxlesandthecapacityofresistanceoftheaxles',Engineering,1871,11.13ShuteN,NoHighway,London,Heinemann,1949.14GurneyTR,FatigueofWeldedStructures,2ndedn,Cambridge,CambridgeUniversityPress,1979.15GurneyTR,TheBasisoftheRevisedFatigueClauseforBS153,London,TheInstitutionofCivilEngineers,1963(Discussion,1964).16SignesEG,BakerRG,HarrisonJD,BurdekinFM,`Factorsaffectingthefatiguestrengthofweldedhighstrengthsteels',BrWeldJ,1967143.17MaddoxSJ,FatigueStrengthofWeldedStructures,2ndedn,Abington,AbingtonPublishing,1991.18GurneyTRandMaddoxSJ,`Are-analysisoffatiguedataforweldedjointsinsteel',WeldResInt,1973,3(4). 136Weldeddesign±theoryandpractice19PilkeyWB,Peterson'sStressConcentrationFactors,2ndedn,Chichester,Wiley,1997.20MarshallPW,DesignofWeldedTubularConnections,Amsterdam,Elsevier,1992.21StructuralUseofAluminium,BS8118,London,BritishStandardsInstitution,1991.22HoneycombeRWK,Steels,MicrostructureandProperties,London,Arnold,1981.23BoydGM,BrittleFractureinSteelStructures,London,Butterworths,1970.24TipperCF,TheBrittleFractureStory,Cambridge,CambridgeUniversityPress,1962.25RoltLTC,VictorianEngineering,London,PelicanBooks,1974.26DwightJB,`Effectofweldingoncompressionelements',ImprovingWeldedProductDesignConference,Abington,TheWeldingInstitute,1971.27YoungWC,Roark'sFormulasforStressandStrain,Basingstoke,McGraw-Hill,1989.28WardenierJ,HollowSectionJoints,Delft,DelftUniversityPress,1982.29SHSWelding,TD394,15E.97,London,BritishSteel,1997.30FrontLine,London,HisMajesty'sStationeryOffice,1942.31BakerJF,TheSteelSkeleton,Vol1,ElasticBehaviourandDesign,Cambridge,CambridgeUniversityPress,1954.32BakerJF,HorneMRandHeymanJ,TheSteelSkeleton,Vol2,PlasticBehaviourandDesign,Cambridge,CambridgeUniversityPress,1956.33SteelDesigners'Manual,London,BlackwellScience,1994.34DaviesJMandBrownBA,PlasticDesigntoBS5950,Oxford,BlackwellScience,1996.35ReportoftheInquiryintotheCausesoftheAccidenttotheDrillingRig`SeaGem',Cmnd3409,London,HMSO,1967.36`GiantOffshoreStructures±Whoseresponsibility?'OffshoreServices,Vol6,No5,Kingston-upon-Thames,SpearheadPublications,July1973.37HicksJG,Astudyofmaterialandstructuralproblemsinoffshoreinstallations,WeldingInstituteResearchReportE/55/74,January1974.38HicksJG,`Astudyofmaterialandstructuralproblemsinoffshoreinstallations',WeldingandMetalFabrication,1974,6(9).39OffshoreInstallations:Guidanceondesign,constructionandcertification,London,HMSO,1990(Revoked1998).40ReidA,ProjectManagement:GettingitRight,Cambridge,WoodheadPublish-ing,1999.41BurgessNT,ed,QualityAssuranceofWeldedConstruction,2ndedn,London,ElsevierAppliedScience,1989.42HarrisonJD,BurdekinFMandYoungJG,`Aproposedacceptancestandardforwelddefectsbasedonsuitabilityforservice',Proc2ndconf,SignificanceofDefectsinWelds,London,TheWeldingInstitute,1968.43LancasterJ,MetallurgyofWelding,6thedn,Abington,AbingtonPublishing,1999.44LancasterJ,HandbookofStructuralWelding,Abington,AbingtonPublishing,1992.45GriffithAA,PhilTrans,A-221,163±8,London,RoyalSociety,1920.46GuideonMethodsforAssessingtheAcceptabilityofFlawsinFusionWelded References137Structures,BS7910:1999,London,BritishStandardsInstitution,1999.47Crackwise3,AutomationofBS7910:1999,Fatigueandfractureassessmentprocedures(softwareondisk),Abington,TWI,1999. BibliographySomepublicationswhichmaybeausefulbackgroundorprovidefurtherreferences:BS7608,Codeofpracticeforfatiguedesignandassessmentofsteelstructures,London,BritishStandardsInstitution,1993.Recommendedpracticeforplanning,designingandconstructingfixedoffshoreplatforms,APIRP2A,Washington,AmericanPetroleumInstitute.Specificationfortheuseofstructuralsteelinbuilding,BS449:Part2,London,BritishStandardsInstitution,1995.Structuraluseofsteelworkinbuilding,BS5950,London,BritishStandardsInstitution.Steelgirderbridges,BS153,London,BritishStandardsInstitution,1958.Steelgirderbridges,Amendmentno4toBS153:Part3Band4,London,BritishStandardsInstitution,1962.Steel,concreteandcompositebridges,BS5400:Part10:1980,Codeofpracticeforfatigue,London,BritishStandardsInstitution. Indexaccess,16,29,34,38,44,46,52,83,114,123,107,112,120126,127consumable,15,16,20,23,25,29,31,51,104,ageing,21109,118,123aircraft,5,7,8,9,20,60,61,63,86,93,121continuouscasting,18allowablestress,55,90,91,93converter,12alloyingelements,14,15,18,19,21copper,7,21,47aluminiumalloy,8,9,20,21,22,31,32,35,core,23,27,2954,61,63,78,86,101,122cornerjoint,36,50angles,4,39corrosion,8,9,10,14,15,16,17,20,21,24,arcwelding,8,10,15,21,23,24,25,26,27,68,76,101,103,12928,29,37,38,42,52,55,109,125coveredelectrode,27,41argon,29,31crackopeningdisplacement,80atomicstructure,13crackpropagation,21,76austenite,13,14,15CTOD,80,103,104,130austenitic,8,16,17,78,101cumulativedamage,61,62,73,74autogenouswelding,31cutting,19,32,44,45,46,49,97backingbar,44,47deeppenetration,27,28,42backingstrip,44,47,49defects,48,50,54,79,80,98,111,112,113,basiccoating,27,29,42120,121,125,126,128,129,130bendingmoment,92designmethods,132Bessemer,12diptransfer,30bevel,32,44,45,49,50distortion,23,34,35,43,44,46,49,52,83,boxsection,83,85121,122,126bridge,2,9,11,23,59,62,64,76,79,83,84,drawings,41,11385,86,87,90,100,101,128,132,134ductility,15,20,49,55,109brittlefracture,3,75,76,77,78,81,97,102,duplexstainlesssteels,17105,122,129,130dyepenetrant,38,123buckling,35,83,88,91,95,98,121burning,32ECA,129buttjoint,37,42eddycurrent,38buttweld,33,37,38,42,44,45,47,49,54,56,edgepreparation,28,32,44,45,47,49,50,52,67,6858,109,114,127elasticdesign,95carbon,12,13,14,15,16,17,19,23electronbeam,23,38carbondioxide,23,29,31electroslagwelding,24,38carbonequivalent,15environment,2,16,68,98,101castiron,11,127,128,129extrude,21,33cellulosiccoating,27,42Charpytest,20,77,78,79fatiguecracking,25,59±74,85,90,95,105,chromium,14,15,16130coldcracking,117fatiguelife,61,63,65,66,68,73,86,87,94,construction,9,12,18,52,79,82,103,104,100,101,102 140Indexfatiguelimit,63leglength,58ferrite,13,14limitstate,95ferritic,16,17,76,86,123fillermetal,20,23MAG,29fillet,37,38,39,51,55±8magnesium,21fit-up,50,58,121,125magneticparticle,123flashbuttwelding,24,38manganese,14,15flaw,112,127,128,129manualmetalarcwelding,25flux,20,23,26,27,28,29,30martensite,14fluxcored,30microstructure,12,14,25,35,80,112,117forgewelding,23MIG,29fracture,55,60,63,64,67,75±81,83,97,98,Miner,73102,118,128molybdenum,14fracturemechanics,73,75±81,112,130MorrisonShelter,91,93fracturetoughness,16,17,20,27,29,32,33,44,75±81,101,102,103nickel,14,15,16,17,27,29,78frictionwelding,21,23,38non-destructiveexamination,48,56,126fullpenetration,38,39,42,43,126NorthSea,18,96,97,98,103,104fusionboundary,28,55notchtoughness,20,54,77±9gascutting,19,32,44,46,48,97offshore,18,62,64,67,78,79,82,86,88,90,gasshieldedwelding,23,27,29,4696±105gouging,46oxy-fuelgas,23grain,13,14,15,19,49,78oxygen,29,31,32hardenability,14,15,29parentmetal,20,27,44,54,76,104,113,117hardness,8,15,109,117partialpenetration,38,39,58,68,126heataffectedzone,15,16,27,32,42,55,80,pearlite,15117,118,126,127pigiron,12heattreatment,11,12,14±16,21,33,35,78,plasma,31,32109,127plastichinge,54,91helium,23,29,31plasticmoment,91hollowsection,85plastictheory,91hotcracking,112,118porosity,29,39,112,123hydrogen,15,16,27,50position,27hydrogencracking,15,19,29,42,112,117,post-heating,15118,119,127postweldheattreatment,35,127preheating,15inclusions,18,20,32,39,43,49,50,78,119procedurequalification,104inertgas,29,31proceduretest,32ingot,17,18pulsing,30initiation,35,60,63inspection,38,39,44,83,109,110,120,122±qualification,376,127,129,133quality,7,17±19,79,107,108,111±30iron,11,12,13±20,22,24,26,27,33,82,84,quenching,11,14,15123,127,128,129radiation,25,123J-preparation,46radiography,18,38,109,123jointtype,65residualstresses,33±5,50,67,75,79,83,121,122lackoffusion,39,44,46,55,57,75,111,112,resistancewelding,24,125123rivets,7lackofpenetration,38,43,54,112,123rolledsections,33,34,121lackofrootfusion,57,116rootface,32,43,44,117lackofsidewallfusion,113rootgap,44,51,117lamellartearing,49,59,199,112,118,119rootpenetration,116lamination,50rutile,27,30,42lapjoint,37laser,23,38safelife,61leansteels,19SeaGem,97,98,105 Index141seamwelding,24T-joint,36,49,88shakedown,20tolerance,38,47,50±2,57,83,120,121,124,silicon,20,29129SNcurve,59±74torch,31,32,46specification,77,78,97,101,104,106±10,transformation,14,15112,117,119±26,129,131±4transitiontemperature,77,78spigotjoint,47,48tubular,9,32,65,67,76,79,84±90,96,100,spotwelding,8,24101,102spraytransfer,30stainlesssteel,8,16,17,27,29,32,33,86,123U-preparation,46steelmaking,12,18,19ultimatetensilestrength,90stirfrictionwelding,21ultrasonics,18,38,109,123,124stressconcentration,63,65,75,76,79,90,100UKOSRP,67,101±4stresscorrosion,8,15,16,17,24,129stresscycle,59±74welddecay,16stressrange,59±74weldface,42stressrelief,122weldmetal,16,20,27,28,29,37,42,44,45,structuralsteel,7,14,18,19,24,32,42,54,55,49,54,55,75,80,100,103,104,113,117,62,77,79,90,102118submergedarcwelding,23,27±9weldroot,42,47sulphur,14,118weldtoe,50,63,64,65,67symbols,41,58weldability,10,17,18,27weldingelectrode,26,41tackweld,47,49weldingengineer,8,10,15,19,28,133tempered,14weldingrod,41,49,52tensilestrength,90,127,129Wellswideplatetest,80thermitwelding,24wroughtiron,11,22,127,129thermomechanicaltreatment,18thicknesseffect,67yieldstress,55,67,90throat,55±8throughthicknessductility,49zinc,20,21TIGwelding,31,86Zqualityplate,50
此文档下载收益归作者所有
