241
PACSnumbers: 06.60.Vz,62.20.fq, 62.20.Qp, 81.20.Vj,81.40.Lm,81.40.Pq,82.80.-d
Effect
of
Friction
Time
on
Microstructure
and
Mechanical
Properties
of
Friction
Welded
AISI
304
Stainless
Steel
to
AISI
1060
Steel
HakanAtesandNihatKaya*
GaziUniversity, FacultyofTechnology, 06500Ankara,Turkey
*
GaziUniversity,
InstituteofScienceandTechnology, 06500Ankara,Turkey
Theinfluenceoffrictiontimeonthehardness,tensileproperties,and micro-structure of the welded samples is investigated. With increasing friction time,thehardnessandtensilestrengthvaluesincreasetoo.Thehardnessof theweldinginterfacesishigherthanthatofthebasealloys.Microstructures of theweldsareexamined bytheopticalandscanningelectronmicroscopy. Theresultsshowthatallweldingjointsareperfectlybondedandfreeofany cracks, pores,anddefects. Thetensilefracture of weldedjointoccurred in the AISI304 side. The chemical composition of the interfaces and defor-mationzonesaredetectedusingenergydispersivespectroscopy.Diffusionof theCrandNiisevidentontheAISI1060side.
Дослідженовпливтривалоститертянатвердість,властивостіприрозтягу та намікроструктуру зварнихзразків.Призбільшеннітривалоститертя твердість та границяміцности нарозрив такожзбільшуються.Міцність поверхні зварнихконтактіввиявляється більшою,аніж у основних сто-пів.Мікроструктурузварнихшвівбулодослідженометодамиоптичноїй електронноїмікроскопії.Показано,щозварніз’єднаннябулидосконало зваренимитанемалиніякихтріщин,порідефектів.Руйнуванняпри ро-зтягувідбувалосязбокукриціAISI304.Хемічнийскладзварних контак-тіві зондеформаціїдосліджено задопомогоюенергодисперсійної спект-роскопії.СпостерігаласядифузіяCrтаNiвбіккриціAISI1060. Исследуется влияние времени трения на твёрдость, свойствапри растя-женииинамикроструктурусварныхобразцов.Приувеличениивремени трения твёрдость и предел прочности на разрыв также увеличиваются. Прочность поверхностисварныхконтактов оказываетсявыше,чему ос-новных сплавов. Микроструктура сварных швов исследуется методами Фотокопированиеразрешенотолько всоответствиислицензией НапечатановУкраине.
оптической и электронной микроскопии. Показано, что все сварные со-единениябылиидеальносвареныинеимелиникакихтрещин,пори де-фектов. Разрушение при растяжении происходит со стороны стали AISI304. Химический состав сварных контактов и зон деформации ис-следуютсяспомощьюэнергодисперсионнойспектроскопии.Наблюдается диффузияCrиNiвсторонусталиAISI1060.
Key words: AISI304SS—AISI1060, friction welding, microstructure,
me-chanicalproperties.
(ReceivedDecember29,2011; in final version, February 12, 2013)
1.INTRODUCTION
AusteniticstainlesssteelsbasedonrelativelyhighamountofCrandNi
are excellentcandidatesto beusedas structuralmaterialsin chemical
and foodindustryas wellasin hospitalsurgical equipments.They
ex-hibithighductility,excellentdrawingandformingproperties.The
var-iousfusion andsolidstateweldingtechniquescanbeusedtojointhese
steels.Weldingofmachinerypartsisnecessaryinthemostofthe
engi-neeringapplications.Justafewdecadesago, materialswereclassified
asweldableandnon-weldable,butinnovationsintechnologyenabledthe
joining ofmajorityofmaterialsbyfusionandsolidstatewelding
tech-niques.Someofthefusiontechniquesareappliedforvariousmaterials
[1—3]. Thetypicalsolid-statejoining applicationsofdiffusion welding
for variouscontent ofthealuminium metalmatrixcomposites [4]and
FSW[5]havebeenstudiedindetail.Weldingtime,appliedload,
weld-ing temperatureandchemicalcompositionofthesteel aresomeofthe
keyparameterstocontrolsolid-statediffusionweldingprocess.
Rotaryfriction weldingisone ofthesolid statejoiningtechniques,
which is applied to join similar and dissimilar counterparts. In this
technique, machinery components are brought into contact with one
another. While one of them remains stationary, the other is rotated
withtheappliedpressure.Whenthetemperatureoftheinterfaces has
reached anoptimum value for the extensiveplastic deformation, the
rotation isstopped,while theforging pressure remains unchangedor
increased.Theapplicationofanaxialforcemaintainsintimatecontact
betweenthepartsandleadstoplasticdeformationinthematerialnear
theweldinterface.Ifsufficientfrictionalheathasbeenproduced
dur-ingsoftening,largerwearparticlesbegintoexpelfromtheinterfaces,
and axial shortening of the components begins as a result of the
ex-pelledupsetting.Ingeneral,heatisconductedawayfromthe
interfac-es,andplasticzonedevelops.Theplasticizedlayerisformedonthe
in-terfaces and the localstress system with the assistance ofthe rotary
shown thattheweld integrityisstrongly affectedbytherateofflash
expelledundertheappropriateconditions[6].Frictionweldinghasalot
of advantages overotherwelding processes. Among these advantages
arenomelting,highreproducibility,shortproductiontime,lowenergy
input, limited heat affected zone (HAZ), avoidance of porosity
for-mation and grain growth and the use of non-shielding gases during
weldingprocess.Inaddition,thetechniqueisnotonlyappliedtoround
specimensbutisalsousedforrectangularcomponents.Itiscalled
line-arfrictionweldingandhasrecentlybeenappliedtothesteel[7].
Although alargeamountof previous studies[4—11] insimilarand
dissimilar materials have been studied, mechanical properties of the
plain carbon steel to stainless steel joint by friction welding method
haveneverbeenreporteduptodate.Therefore,inthispaper,theAISI
1060 and AISI 404stainless steelare welded by frictionweldingand
themechanical,microstructuralpropertiesandoptimalwelding
condi-tionsoffrictionweldedsamplesareexamined.
2.MATERIALSANDMETHODS
Chemical composition of plain carbon steel and stainless steel
em-ployed is presented in Table.Cylindrical test specimens of 20 mm in
diameterand160mminthelengthwerepreparedforfrictionwelding.
Before thefriction welding, the surfaces facing each other were
ma-chined usingalathe.Before welding, thesurfaces of theworkpieces
werecleanedbyastainlesssteelbrushandacetonetoremovetheoxide
layerandstains.Joiningofthesetwodissimilaralloyswasperformed
onacontinuousdrivefriction-weldingmachineof300kNcapacityata
constant rotation speed of 2000 rpm, and constant friction pressure
(P1) of40MPa andconstant upsetting pressure(P2)of60 MPa.
Fric-tion and upsetting pressures can be observed on the screen, and the
stages ofweldingsequencearecontrolled by solenoidvalve.The
fric-tiontime(t)waschosenasthree,five,andsevenseconds.Tensiletest
specimenswerepreparedaccordingtoASTME8M-00b.Ultimate
Ten-sile Strength (UTS) and yieldstrength of thewelded specimens were
determined. Hardness values(VH1)were determinedusingShimadzu
TABLE.Chemicalcompositionoftestedmaterials.
Material C Cr Ni Si Mn P
AISI 1060 0.42 — — 0.21 0.45 0.02
Elements,wt.%
Material C Cr Ni Si Mn P
HMVdevice.Measurementsweretakenintheweldingcentrethrough
base metal witha distance of1 mm oneither sideof horizontal line.
Hardness valuesweremeasuredintheweldingcentre throughthe
di-ameterwithadistanceof2.5mmonverticalline.Fortensilestrength
and hardnesstest values,atleast, threespecimenswere prepared for
each parameter. The microstructure was investigated with the SEM
and optical microscopy.The AISI1060 steelspecimens werepolished
andetchedwithasolutionconsistingof4%Nital.AISI304SSetched
with asolutionof95%HCland5%OH—C6H2(NO2)3.Thegrainshave
various shapes ranging from10 μm to300 μm indimension. Typical
microstructuresofmaterialscanbeseeninFig.1.
3.RESULTSANDDISCUSSIONS 3.1.HardnessDistribution
Figure2,ashowstheeffectoffrictiontimeonthehardness
distribu-tion in the direction perpendicular to the weld interface of the
as-welded specimen. The maximum hardness values of joints were
ob-tained on the welding centre line. The hardness values generally
in-creased with friction time.Increasing hardness inthe welding
inter-facecanberelatedtothemicrostructureformedinthejointsurfaceas
aresultoftheheatinputandextensiveplasticdeformation.The
hard-nessofAISI304SSis129HVandthehardnessvalueoftheAISI1060
steelis254HV,butmeasuredhardnessofweldinginterfaceisaround
300HV.Wecansupposethatmicrostructuralevaluationsaroundthe
interface,diffusionofelementsonthesides,workhardening,
disloca-tion density, grain refinement, and formation of precipitations may
causethisincrease.AlthoughtheworkhardeningiseffectiveontheSS
side, phasetransformation, finegrainsand formationofprecipitates
and rapidcoolingfromtheweldingtemperatureresultedinthe
hard-ening in the AISI1060 steel. Similar hardness profile was also
ob-servedfordissimilar materialscouples[11].One cansupposethat the
hardnesspropertiesareinfluencedbyaninteractiveeffectoffriction
time,whichiscombinationsofheatinput,rangeofplasticdeformation
and degreeof strainhardening.The influenceofthe frictiontimeon
theverticalhardnessdistributionisalsoshowninFig.2,b.Bothsides
of materials are hardened duetothe austenitetransformation to the
martensite formation,whichis stronglyaffected by thecooling rate.
Allhardnessvaluesarehigherthanthoseinthebasemetalare.Itwas
observed that hardnessvalues increasedfrom thecentre through the
thicknessofthespecimen.Thephasetransformationandcoolingrates
weretheoptimumaround5mmawayfromthecentre.
3.2.TensileProperties
TheinfluenceoffrictiontimeonthetensiletestresultisshowninFig.3.
It can be seen that the yield strength slightly increased with the
in-Fig. 3.Theeffectof friction time on the tensile properties.
Fig.2.Theeffectoffrictiontimeonhorizontal(a)andvertical(b)hardness distribution.
crease offriction time.The UTSvalue forthe AISI1060 is 814MPa
and forthe AISI304SS it is 505 MPa [12]. The UTSstrength
corre-spondstoabout130%ofthatforAISI304SSpartand81%ofthatfor
AISI1060structuralsteelpart.AlthoughtheUTSvalueswerenot
in-fluenced by friction time,thehighest yieldstrength is obtained with
thehighestfrictiontime.
Fabricatedjointswerequiteductileinnatureand3sand5ssamples
failed in the AISI1060 side due tothe low elongation, which is only
17%comparedto70%fortheSS.However,whenthefrictiontime
in-creasedtothe7s,failureoccurredontheSSside(seeFig.4)duetothe
extended deformation and heat affected zone. Almost all specimens
failed on the base metals, showing that perfect joint interfaces were
obtainedunderthevariousfrictiontimes.Itisreportedthatthechoice
ofweldingparametersaffectsthemicrostructure[13—15].Ifthe
fric-tiontimeisheldlong,abroaddiffusionzonewithintermetallicphases
maybegenerated.Suchparametersasshortfrictiontime,lowfriction
and low upsetting pressures will result in a weakly bonded joint.
Hence, the quality ofthe welds stronglydepends onthe temperature
attainedbyeachsubstrateduringweldingprocess.Therefore,theflow
stress—temperaturerelationshipforeachmetalwillhaveanimportant
influence onthe friction welding parameters.However, it is also
re-ported that a good strength can be obtainedby a sufficientdiffusion
andmechanicallocking[11].Ifoptimumweldingparametersareused
infrictionwelding,aperfectbondingcanbeachieved.Ingeneral,high
friction andupsettingpressuresresultedinhightoughnessand
hard-ness in steel alloys [8]. It is also well known that for achievinghigh
strength at welding interface the brittle phases such as σ-phase,
δ-ferriteandoxideformationsshouldbepushedawayfromthejoint
sur-face. In thiscase, toachieve higherstrength, the frictionand
upset-ting pressures should beas high as possible forthe constant friction
time.Asthefrictionandwearhelptogetridofcontaminantslike
ox-ideontheinterface,anewfaceshouldbesurfaced.Atthesametime,
theupsettingandfrictionpressureandfrictiontimesetintobringthe
newfacewithinthescopeofattractionrange[14,15].
3.3.Microstructure
Figure5showsthetypicalmicrostructuralfeaturesofthewelded
sam-ples.Asseenfromthemacrographs,flashformationwasobservedinall
weldedsamplesbecauseofplasticdeformationduringwelding.Friction
Fig.5.Typicalmacro-andmicrostructureofthefrictionweldingparts;P1 = =40MPa,P2 =60MPa,t=3 s (a), 5 s (b), 7 s (c).
timeplaysanimportantroleinflashformation,whichgetsbiggerwith
increasingfrictiontimeonbothsidesofmaterialsduetothehigherheat
input and extensive plastic deformation. Thus, the burn-off (axial
shortening) increases with the increase of plastic deformation (see
Figs5,a,c).Allofthe weldedjointswereperfectlybondedandfreeof
anycrack,poreanddefect.
Twomainregionsareobservedontheweldedsamples:adynamically
recrystallizedzonewithextremelyfinegrainsandplasticallydeformed
heataffectedzone(seeFigs.6,a—c).Thewidthofbothzoneschangesdue
tothematerialsproperties,i.e.yieldstrength,coefficientofthermal
ex-pansion(CTE),andductility.Lowyieldstrength,highheatconductivity
andhighductilityatelevatedtemperaturescausesmoredeformationof
theAISI1060sidecomparedwithAISI304side.Thus,deformation that
is more plastic,extendedHAZandmoreaxialshorteningwereobserved
onAISI1060sideandthesepropertiesexpandedwithincreasedfriction
time. Figures 6, a—c also show the grain orientation inside the HAZ
zones. The intensityoforientation dependson the welding parameters
[11].Itisclearlyseenthatsufficientfrictiontimeonthejoiningsurface
resultedinanadequatelockingofthe surfaces,whereextensiveplastic
deformation and microstructural changes occurred, and these facts
caused better tensile response of the joint, where recrystallization
oc-curredontheAISI1060side(higherdeformation)andgrainrefinement
and strain hardening occurredon the AISI304 steelside (lower
defor-mation).Thewidthofrecrystallizedzoneismainlyaffectedbythe
fric-tion time.Anincreasein the frictiontime resultsin widersizeoffine
grainregion.Thechangeofmicrostructureandhigheraxialshortening
isalsoobservedintheAISI1060side.
Heat flow is observedpreferentially in the material with the higher
thermal conductivity. Due to the different thermal conductivities of
AISI304 (16.2 W⋅m−1⋅K−1) and AISI1060 (49.8 W⋅m−1⋅K−1), the specific
heatwassolarge,andhencemostofthefrictionalheatgeneratedinthe
AISI1060 side.Thedifferencein thermalconductivityexplainsthe
mi-crostructuralchanges,whichoccurpreferentiallyintheAISI1060side.
SimilarmicrostructuralobservationsfortheCu—Ticoupleswerealso
re-portedin[16].Mostofthechangestookplaceinthecopper,noevidenceof
microstructural variationwas seenin the vicinityofthe interface,and
neithergraingrowthnorgrainboundaryprecipitationwasobserved.This
behaviourcanbeattributedtothedifferenceofthermalconductivity[16].
Figures 5and6alsoshowthemicrostructural changesof AISI304
sideand AISI1060 sideofweld. Duringthefriction weldingprocess,
thetemperatureneartheweldinginterfacereachesjustaboveA1
tem-perature, which is almost equal to the recrystallization temperature
forthe steel.Therefore,extremely fineequiaxedgrainsformed from
ferriteandpearlite.Martensiticstructurecouldalsobeseeninthis
re-gion,when theweldinginterfacereachesA3 temperaturewiththe
oc-currenceofrapidcooling. Therefore,themicrostructureofAISI1060
steelturnedintoaustenite.Whenausteniticstructurewasexposedto
top cooling from evaluated temperatures, the microstructure turned
into martensite duetothe diffusion of carbonto thegrain boundary
and rapid coolingrate.These results meanthat the weld experienced
different thermal histories. The relatively quick cooling rate
influ-encedthemicrostructureoftheweld.
Figures 7 (1) and (2) show energy dispersive spectroscopy (EDS)
analysis points observedon theAISI304 side, while(4)and (5) show
EDS analysispointsobservedontheAISI1060 side.Itisseenthat
al-loyingelementconcentrationsignificantlychangedatthejointsurface
(Point3)andadjacenttothePoint3duetothemechanicallockingand
atomicdiffusion.InthejointsurfaceCr(1.53%)wasmeasuredfor3s
friction time.However,the increase ofthe friction timecaused high
densityofCrinthesurface,wheremoreheatgeneratedasaresultof
highfrictionpressure.ItisobviousthatCratomswithsmaller
diame-teraremoremobileanddiffusiveinsidethisregion.Significantlyhigh
amounts of Crand Niare detectedonthe AISI1060 sideinPoint(4)
and (5) under all conditions. Çelikyürek etal. [17] have reported the
Fig.7.JointsurfacesandlocationoftheEDSanalysissites.Theeffectof fric-tiontimeontheelementaldistribution.
High amount ofAl, Cr, and Niatoms were diffused into interface
andHAZregionduetothelongperiodoffrictiontime.
4.CONCLUSIONS
Based ontheresultsobtained inthisstudy,thefollowingconclusions
couldbedrawn.
1.Frictionweldingcanbeusedsuccessfullytojoinausteniticstainless
steel(AISI304)toAISI1060structuralsteel.Theprocessedjoints
ex-hibitedbettermechanicalandmetallurgicalcharacteristics.
2.Thehighesthardnessvalueswereobservedintheinterface,butthey
significantly varied with increasing distance and friction time. The
increaseinthehardnessatthejointzonemaybeattributedtothe
mi-crostructuralevaluation.
3. The tensile strength of friction-welded joints was not affected by
friction time.Thetensile strength ofobtained weld joints canexceed
AISI304SS strength by 30%. Fracture usuallyoccurs intheHAZ of
AISI1060steelside.
4.AnextensivedeformationoccurredmostlytotheAISI1060sidedue
toitsflowstressandhighthermalconductivity.
5. Typical microstructural features were observed in thewelding
re-gion. Recrystallizedfineandelongatedgrainswereobservednearthe
jointsurface.Thewidthofrecrystallizedregionismainlyaffectedby
thefrictiontime.
6.Thevarianceinweightofalloyingelementsisseenfromanalysisof
elements spectra. EDS measurements clearly show that steel joints
consistofdifferentcrystalstructure, carbidesandintermetallic
com-pounds.DiffusionoftheCrandNiisobservedintheAISI1060side.
ACKNOWLEDGEMENTS
Theauthors wishtothankProf.Dr.A.Kurt,Prof.Dr.N.Kahraman
and Prof.Dr.I.Uygurfortheirinvaluable contribution.Alsospecial
thankstoTurkishTrackFactoryfortheirsupportandguidanceduring
theexperimentalstage.
REFERENCES
1. I.UygurandB.Gulenc, Metallurgiya, 43,No.1:35(2004).
2. I.Uygur,IndustrialLubricationandTribology,58,No.6:303(2006). 3. I.UygurandI.Dogan, Metallurgiya, 44,No.2:119(2005).
4. I.Uygur,Mater.Sci.Forum,546—549:671(2007).
5. A.Kurt,I.Uygur,andH.Ates, Mater.Sci.Forum, 534—536:789(2007). 6. A.VairisandM.Frost,Mater.Sci.Eng.A,271:477(1999).
7. W.Y.Li,T.J.Ma,S.Q.Yang,Q.Z.Xu,Y.Zhang,J.L.Li,andH.L.Liao, Ma-ter.Lett.,62:293(2008).
8. P.Sathiya,S.Aravidan,andA.NoorulHaq, Int.J.Adv.Manufac.Technol., 31:1076(2007).
9. S.D.Meshram,T.Mohandas,andG.M.Reddy, J.Mater. Proces.Techn., 184:
330(2007).
10. K.Jayabharath,M.Ashfaq,P.Veugopal,andD.R.G.Achar, Mater.Sci.Eng. A,454:114(2007).
11. N.Ozdemir,F.Sarsilmaz,andA.Hascalik, Mater.Design, 28:301(2007). 12. www.matweb.com
13. A.Kurt,I.Uygur,andU.Paylasan, WeldingJournal, 90:102(2011). 14. H.Ates,M.Turker,andA.Kurt,Mater.Design,28:948(2007).
15. M.Sahin,H.E.Akata,andK.Ozel, Mater.Design, 29,No.1:265(2008). 16. S.Y.Kim,S.B.Jung,andC.C.Shur,J.Mater.Sci.,38:1281(2003). 17. I.Çelikyürek,O.Torun,andB.Baksan, Mater.Sci.Eng.A, 528:8530(2011).
