Fracture toughness of ISO 3183 X80M (API 5L X80) steel friction stir welds

Fracture toughness of ISO 3183X80M (API 5L X80)steel friction stir welds T.F.A.Santos a ,T.F.C.Hermenegildo a ,b ,C.R.M.Afonso a ,R.R.Marinho c ,M.T.P.Paes c ,A.J.Ramirez a ,*a Laboratório Nacional de Luz Síncrotron (LNLS),C.P.6192,Campinas,SP 13083-970,BrazilbDepartamento de Engenharia de Materiais,Universidade Estadual de Campinas (Unicamp),Campinas,SP 13083-860,Brazilc PETROBRAS/CENPES,Rio de Janeiro,RJ 21949-900,Brazil a r t i c l e i n f o Article history:Available online 5August 2010Keywords:Friction stir welding CTOD fracture toughness ISO 3183X80M (API 5L X80)steela b s t r a c tFriction stir welding (FSW)is a solid-state joining process with numerous advantages suchas good dimensional stability and repeatability,which is widely used Al alloys and with agreat potential for critical joining applications involving high melting temperature alloys.Twelve millimeter thick plates of ISO 3183X80M (API 5L X80)steel was friction stir weldedusing two passes on both sides of the plate using ceramic tools.Different heat inputs wereobtained using a fix travel (welding)speed in combination with several spindle speeds.Thefracture toughness of the two-pass joints was evaluated at 25°C using the critical crack tipopening displacement (CTOD m ),revealing that joints produced with lower spindle speedspresented higher toughness at the heat-affected zone (HAZ)and stir zone (SZ),which arecomparable with the base metal (BM)toughness.On the other hand,joints produced usinghigher spindle speeds presented low fracture toughness at the SZ and elevated CTOD mtoughness at the HAZ.The joints produced with low spindle speeds showed CTOD m -valuesabove the offshore standard (DNV-OS-F101)requirements.Ó2010Elsevier Ltd.All rights reserved.1.IntroductionFriction stir welding (FSW)is a solid-state joining process,which eliminates the melting and solidification associated problems,such as liquation and solidification cracking.In addition,in the case of steels,it is expected that hydrogen induced cracking is severely reduced.Besides,among the most relevant advantages of FSW over traditional fusion welding processes are the inherent automation,lower distortion and residual stresses,and good mechanical properties and performance of the joints [1–4].The FSW process has been applied to join several materials,including aluminum,magnesium,lead,zinc,and copper alloys.Recently,FSW of high-temperature alloys including steels,stainless steels,titanium alloys,and Ni-base alloys has been increasingly drawing attention from the aerospace,shipbuilding,and petrochemical industries [1–5].However,the development and field deployment of such technology to join these high melting temperature materials continues to be a challenge.The development of affordable and relievable tools that withstand the imposed thermal and mechanical condi-tions remains as one of the barriers that needs to be overcome.Polycrystalline Cubic Boron Nitride (PCBN)and composites using this material have shown promising results as FSW tool materials to join a variety of metals [1–9].Moreover,PCBN is relatively inert when processing/welding iron and nickel alloys at high temperatures.However,the reported low fracture toughness of pipe-steels FSW joints has been a major issue that could delay or even prevent the deployment of such tech-nology [10].Keeping that in mind and confident that a metallurgically-based solution could be found to this issue,the LNLS 0013-7944/$-see front matter Ó2010Elsevier Ltd.All rights reserved.doi:10.1016/j.engfracmech.2010.07.022*Corresponding author.E-mail address:ramirez@lnls.br (A.J.Ramirez).2938T.F.A.Santos et al./Engineering Fracture Mechanics77(2010)2937–2945research group in collaboration with Petrobras engaged in a R&D program tofind a combination of welded material and bainitic transformation processing conditions that will provide sound and high toughness FSW joints.Such solution has been found,opening the way to the deployment of such technology to high strength steels welding,which may have a tremendous technological and economic impact on the energy,oil and gas,nuclear,petrochemical and shipbuilding industries[1].The study was based on Brazil-produced high strength steels,which have additions of titanium,niobium and vanadium for grain refinement and carbide precipitation hardening[11,12].The commercial ISO3183X80M steels,which are equiv-alent to the American Petroleum Institute(API)5L X-80steels have typical microstructures consisting of ferrite and bainite. In order to increase strength without significant losses in toughness,a high volume fraction of bainite is desirable.However, the thermal cycles imposed during the welding process may lead to localized brittle zones in the heat affect zone(HAZ). Additions of Mn preclude the bainite transformation and promote the formation of the martensite–austenite(MA)constit-uent,which may reduce HAZ toughness[13,14].The MA constituent morphology plays an important role in the fracture toughness,with elongated and massive MA resulting in lower and higher fracture toughness,respectively.Such morphology may be controlled by allowing additions,with silicon and aluminum promoting massive MA[15].Due to the normally expected high fracture toughness of modern high strength structural steels,the application of linear elastic fracture mechanical concepts with K IC does not lead to proper results.Therefore,a safer approach is to use elastic–plastic fracture mechanics,which can be accessed using CTOD or J-integral tests[16].The crack-opening displacements (COD)concepts bridges the elastic and plastic fracture conditions.According to Hertzberg[17]the COD is related to the ex-tent of plastic strain in the plastic zone,which is analogous to the toughness measurement from the area under the stress–strain curve in an uniaxial tensile test.Hence,the concept of critical crack-opening displacement near the crack tip has been introduced to provide the fracture criteria and an alternative measurement of fracture toughness.According to Bhadeshia and DebRoy[18],just few works describe the fracture toughness of friction stir welds of steels,and these studies used only elementary mechanical properties.Kaiser et al.[16]have shown that arc-welding of high strength steel results on CTOD-values reduction.High strength structural steels with355and460MPa yield strengths(S355and S460)have shown heat-affected zone minimum CTOD-values near0.25mm,which is the minimum requirement for steels offshore application.Such tests were performed on 20mm thick samples atÀ10°C.On the other hand,S690steel(690MPa yield strength)13mm thick joints tested at the same temperature presented CTOD-values between0.10and0.20mm.Fairchild et al.[19]performed two-sided friction stir welds produced with12.0mm plates in the temperature range be-tweenÀ100andÀ10°C.The authors found CTOD-values for the base metal between0.10and0.20mm within the temper-ature range betweenÀ110andÀ40°C.For the temperature range betweenÀ20andÀ10°C,the CTOD-values far the base material were between0.40and0.50mm.The friction stir welds produced using300rpm spindle speed and51mm/min welding speed,using PCBN tools,were between0.10and0.20mm,the CTOD results indicated that the weld toughness were significantly lower than those observed for the base metal atÀ10°C(0.50mm measured by British Standard7448-1[20]).2.Experimental procedureThe study was performed on a commercial grade ISO3183X80M(API-5L-X80)pipe-line steel with chemical composition indicated in the Table1.Table2shows the mechanical properties of the steel studied in this work.Both the chemical com-position and the mechanical properties were provided by the steel producer,Usiminas[21].Plates of400Â200Â12mm were used for the FSW samples preparation.These plates were machined from19mm thick plates and care was taken to keep the plate banding at the through-thickness center.The welds were performed normal to the rolling direction in order to compare them to pipe-line arc-welding circular joints.Ju et al.[22]have shown that CTOD fracture toughness of pipe-line steel changes significantly when the fracture propagates normal and parallel to the rolling direction.For API-5L-X65steel,the CTOD value changed from1.33to0.58mm when tested,at room temperature,normal and parallel(longitudinal joints)to the rolling direction.Two pass friction stir welds were performed on both sides of these12.0mm plates using the parameters shown in Table3. The joints were performed using a dedicated TTI FSW(Fig.2b),which allows position and force controlled welding.Downward forces between28and33kN(force controlled process)were necessary to produce sound welds using an un-tilted PCBN tool with threaded conical shape and6mm long pin and convex threaded shoulder.Welding speeds were limited to keep the forces in the welding direction(x direction)at or below5kN in order to extend the tool life.The radiographic examination did not reveal any internal defect in the performed joints.The fracture toughness of the joints was evaluated using the three-point bending crack tip opening displacement (CTOD)test,which was carried out according to ASTM E1290[23].The12mm thick CTOD specimens were standardTable1Chemical composition(wt.%)of the ISO3183X80M pipe-line steel.C Mn Si P S Al Nb V Ti Ni Cr Mo N a B a0.05 1.760.170.0160.0020.0350.0660.0250.0160.020.150.20593a ppm.T.F.A.Santos et al./Engineering Fracture Mechanics77(2010)2937–29452939Table2Fig.1.Three-point bend specimen geometry for fracture toughness testing in the stir zone.Fig.2.(a)Downward force and tool shoulder temperature evolution during the welding operation and(b)TTI friction stir welding system.2940T.F.A.Santos et al./Engineering Fracture Mechanics77(2010)2937–2945single-edge-notched-bending(SE(B))with through-thickness notch.For the BM,HAZ/BM and SZ tests,the pre-cracks were oriented in the through-thickness(L–T)direction.The schematic drawing of machined specimens is shown in Fig.1.All tests were performed at room temperature(25°C).3.Results and discussion3.1.Performed weldsThe welding parameters provided the tool advance(mm)per revolution changing from0.20to0.33(see Table3),which means higher temperature and lower temperature,and consequently,higher and lower heat inputs,respectively.Sound fric-tion stir welds were produced for the three different used parameters combinations.In the friction stir welding(FSW)system there is a hole,in which is possible to insert the thermocouple into the tool head and it is attached to the FSW machine.This thermocouple positioned by the tool shoulder does not provide the exact tool temperature at its interface with the processing material within the stir zone.However,due to the high thermal conductivity of PCBN,it provides a comparative value for different welding parameters andfinal microstructures[7].The peak temper-atures achieved at the PCBN tool during friction stir welding near the tool shoulder was of870,905and approximately 1000°C for the0.33mm/revolution,0.29mm/revolution and0.20mm/revolution welding conditions,respectively,as pre-sented in Fig.2a.There were some temperature measurements problems for the0.20mm/revolution joints,however, slightly higher temperatures are expected in the tool.Fig.2a shows the downward force evolution during the welding process for the different used parameters.The quase-stationary downward force during the transversing(actual welding)ranged between28and33kN(Fig.2b).A weld width of20mm at its face was kept approximately constant for all the used parameters.Preliminary tests were performed in position3.(a)Macrograph of the welded joint showing thefirst and second pass,the retreating(RS)and advancing sides(AS).Vickers microhardness traversefriction stir welded X80M steel were performed in(a)sample welded with0.33mm/revolution,(b)0.29mm/revolution and(c)0.20mm/revolution.The BM shows CTOD m -values of 0.43mm for the 12mm thick samples The BM stable fracture propagation region of the 12mm thick CTOD samples surface (see Fig.5a)indicates the presence of cleavage and dimples micromechanism,explaining the good CTOD m -value,typical for ductile materials.The CTOD m -values measured at the SZ and HAZ of samples welded with 0.33and 0.29mm/revolution parameters,which correspond to 300and 350rpm spindle speeds,respectively,indicate excellent fracture toughness.The CTOD m was of 0.75and 0.73mm for SZ and HAZ,respectively,for the 0.33mm/revolution joint.Similarly,the SZ and HAZ presented CTOD of 0.62and 0.79mm,respectively,for the 0.29mm/revolution joint.The fracture surface of these samples shows a ductile frac-ture surface formed by dimples micromechanism in both regions,the SZ and HAZ,which are shown in Fig.5.Different from the 0.33and 0.29mm/revolution joints,the SZ of 0.20mm/revolution joint (500rpm spindle speed)shows a much lower CTOD m -value.The SZ and HAZ show CTOD m -values of 0.19mm and 0.81mm,respectively.Here also the CTOD samples fracture surface evaluation indicates the fracture micromechanism,where the low fracture toughness SZ presented a cleavage fracture morphology formed mainly by small cleavage facets (see Fig.5f).On the other hand the high fracture toughness HAZ presented surface formed by dimples,indicating the micromechanism of ductile fracture behavior revealed by the CTOD test.Therefore,it becomes clear that the stir zone fracture toughness reduction with the factor mm/revolution reduction from 0.33to 0.20,which implies an increase of heat input,which clearly brings microstructural changes that are associated to different thermo-mechanical processing histories.3.3.Relationship between mechanical properties and microstructuresThe BM and the 0.20mm/revolution joint SZ were submitted to X-ray diffraction (XRD)analysis.The most intense peaks were related to ferrite phase ones,as presented in Fig.6.The cementite phase could not be identified because its more in-tense peaks coincide with ferrite’s peaks,which are much more intense due to the higher volumetric fraction.The BM and SZ show approximately the same phases,however the SZ ferrite peaks are slightly wider,which can be due to strain and smaller grain size in this region.In addition,the diffractograms shown in Fig.6reveal low intensity austenite peaks,which are due to such phase presence within the MA constituent.Fig.7a presents an overall view of the microstructure obtained in the optical microscopy (OM)of the steel as-received,consisting of ferrite (F),degenerated Pearlite (DP)[24]and martensite–austenite constituent (MA).In the Fig.7b scanning electron microscopy (SEM)using secondary electrons (SE)shows the DP can be observed in detail.This steel is manufactured in order to form some MA,which is desirable to increase the fracture toughness [13].The HAZ for all the studied parameters present elevated fracture toughness and very similar microstructure,which is shown in Fig.8a.This microstructure presents moderate amount of MA constituent with massive morphology in the ferritic matrix,called granular bainite [10],granular ferrite or granular bainitic ferrite [26].According to Bangaru et al.[10]this microstructure is beneficial to fracture toughness.Bott et al.[13]have reported that massive MA constituent may result in elevated toughness in these high strength steels,which has been corroborated by the elevated CTOD-valuesmeasured Schematic representation of the FSW process and actual welded joint cross section.The CTOD sample notch positions are represented lines for the stir zone and heat affect zone/base metal,respectively.RS –retracting side;AS –advancing side (right side)and (b)CTOD base material and for the stir zone (SZ)and heat affect zone (HAZ)of 12mm thick friction stir welds performed with 300,350speeds (left side)in the room temperature.in this work.However,Bonnevie et al.[27]claim that massive MA constituent morphology is detrimental for the material toughness.On the other hand,the stir zone microstructure has been more severely influenced by the welding process and parame-ters.The joints produced using 0.33and 0.29mm/revolution parameters,which have similar toughness,show the same microstructure,composed by acicular ferrite and MA (granular bainite)and some carbides,(Fig.8c and d).AccordingtoScanning electron microscopy (SEM)fracture surface of stable crack propagation region of CTOD samples fractographs.(a)BM;(b)SZ and welded with 0.33mm/revolution parameter;(d)SZ and (e)HAZ of samples welded with 0.29mm/revolution parameter;(f)SZ and welded with 0.20mm/revolution parameter.measurements performed on the BM and SZ for the joint produced usingBase material microstructure,modified Le Pera Etching[25]:(a)optical microscopy and(b)secondary electron image–scanningIkawa et al.[28],the MA constituent morphology varies according to the cooling rate.He shows that MA has an elongated morphology for low cooling rates and its morphology changes to massive as the cooling rate increases.The microstructures observed in this work,have shown that the MA morphology have changed from slightly acicular for the0.2mm/revolution joint to massive for the0.33and0.29mm/revolution joints.Lower values of mm/revolution have resulted on higher heat inputs.However changes in the deformation heat generation and friction coefficient with temperature may result in unex-pected heat generation variations.Therefore,the MA morphology evolution observed in this work does not agree with the results reported by Ikawa et al.[28].This may be due to the deformation effect unexpected heat generation rate changes or to the important chemical composition differences between the studied alloys.The large Mn content difference should be having a very important effect on the MA formation and therefore,on its morphology.The SZ microstructure of the0.20mm/revolution joint presents the MA constituent associated to the upper bainite-like constituent,which has been designated as degenerated upper bainite[10].Fig.8c shows the elongated MA constituent and the large amount of carbide(cementite)precipitation in this region of the0.20mm/revolution joint,which explains the low fracture toughness of this microstructure.Fig.8d presents a detail of this microstructure where is observed that the carbides (cementite precipitates)are located along the acicular ferrite boundaries,which would be the typical morphology of upper bainite.The elongated MA constituent has been highlighted in thisfigure.The SZ microstructure of the joints produced using0.33and0.29mm/revolution parameters,shown in Fig.9,reveal the lower carbide content of these microstructures,which has resulted in better fracture toughness,as measured through the CTOD test.Microstructural data are summarized in the Table4.The ferritic grain size(GS)was measured by electron backscatter diffraction technique(EBSD).These data shows that the ferritic grain size does not change from BM to SZ.Therefore,the SZ strengthening mechanism is due to the phase transformations process.However,other welded joints show different con-clusions,such as the2205[29]and2507[7]duplex stainless steels.These joints present the grain refinement as the strengthening mechanism and phase transformations was not observed.0.2mm/revolution joint(a)HAZ(optical microscopy–OM),(b)HAZ(scanning electron microscopy–SEM).Moderate amount ofpointed by the arrows,is observed in this microstructure.The darker phase corresponds to iron carbide(cementite)and the lighter matrix and(c)SZ(OM),(d)SZ(SEM).Large amount of carbides(cementite)and elongated MA constituent,pointed by the arrows,are microstructure.Modified Le Pera Etching[25].0.29mm/revolution joint SZ.Lower amount of carbides(cementite)and MA is observed in this microstructure:(a)optical microscopy–modified Le Pera Etching[25].Table4Typical microstructure and ferritic grain size of the welded joint.Joint(mm/rev)Welded joint Ferritic grain size(l m)BM HAZ SZ BM SZ0.20F+DP+MA GB(FB+MA)+C DUB(F+MA)+C530.29GB(FB+MA)+C40.334F–ferrite,DP–degenerated Pearlite,MA–martensite–austenite constituent,DUB–degenerate upper bainite,FB–bainitic ferrite,GB–granular bainite, C–iron carbide.T.F.A.Santos et al./Engineering Fracture Mechanics77(2010)2937–29452945 4.ConclusionsFriction stir welding process parameters that produce fully consolidated and defect free12mm thick joints in ISO3183 X80M(API-5L-X80)have been developed.The12mm thick base material(BM)fracture toughness measured at25°C by the CTOD m test was of0.43mm.The stir zone fracture toughness of the produced joints have varied from(i)unsuitable for pipeline application(CTOD 0.19mm)to(ii)exceeding the base material one(CTOD0.73–0.79mm),qualifying the joint toughness for commercial pipe-line application.The process parameters variation resulted in appreciable microstructural changes within the SZ.The0.20mm/revolution joint resulted in degenerated upper bainite associated with the elongated MA and high carbide(cementite)content which was detrimental to fracture toughness.On the other hand,the0.29and0.33mm/revolution joints presented massive MA associated with acicular ferrite(granular bainite)and lower carbide content which was beneficial to fracture toughness.For the studied welding conditions there was not appreciable microstructural,neither toughness differences in the ana-lyzed region which includes the HAZ and BM at the welding conditions.The measured HAZ/BM fracture toughness is suitable for commercial pipeline application.The microstructure of the HAZ was formed by moderate amount of massive MA constituent associated with acicular fer-rite(granular bainite)and moderated amount of carbides(cementite).AcknowledgmentsThe authors would like to acknowledge Petrobras,FINEP,and LNLS for thefinancial support and,TenarisConfab(S.Kojima) for the materials donation.The authors would like to acknowledge the National Council for Scientific and Technological Development(CNPq)for the scholarships of T.F.A.Santos and T.F.C.Hermenegildo.The authors are very grateful to Prof. Itamar Ferreira of Unicamp,who kindly contributed to the discussions of this work.We are also grateful to JoséLuis Lisboa for helpful assistance during experiments at Unicamp.References[1]Mishra RS,Mahoney MW.Friction stir welding and processing.Ohio:ASM International;2007.p.360.[2]Mishra RS,Ma ZY.Friction stir welding and processing.Mat Sci Engng R2005;50:1–78.[3]Defalco J.Weld J2006;85:42–4.[4]Thomas WM,Nicholas ED.Mater Des1997;18:269–73.[5]Fujii H,Cui L,Tsuji N,Maeda M,Nakata K,Nogi K.Mater Sci Engng A2006;429:50–7.[6]Okamoto K,Hirano S,Inagaki M,Park SHC,Sato YS,Kokawa H,et al.In:Proceedings of the4th international symposium on friction stir welding,Utah;2003.p.1–11.[7]Steel RJ,Sorensen CD,Pettersson C,Sato YS,Nelson TW,Sterling CJ,et al.Stainless Steel World2004;16:1–16.[8]Nelson TW,Su J-Q,Steel RJ.In:Proceedings of the fourteenth international offshore and polar engineering conference;2004.p.50–5.[9]Sorensen CD,Packer SM.In:Proceedings of the3rd international symposium on friction stir welding,Kobe;2001.p.1–17.[10]Bangaru NV,Fairchid DP,Macia ML,Koo JY,Ozekcin A,et al.Proceedings of the pipeline tech.Ostend2004:789–804.[11]Wang JP,Yang Z-G,Bai BZ,Fang HS.Mater Sci Engng A2004;369:112–8.[12]Pontremoli M,Bufalini P,Aprile A,Jannone C.Metals Technol1984;11:504–14.[13]Bott IS,de Souza LFG,Teixiera JCG,Rios PR.Metal Mater Trans A2005;36:443–54.[14]Ozekcin A,Jin HW,Koo JY,Bangaru NV,Ayer R,Vaughn G,et al.Int J Offshore Polar Engng2004;14:284–8.[15]Bayraktar E,Kaplan D.J Mater Proc Technol2004;153–154:87–92.[16]Kaiser HJ,Kern A,Niessen T,Schriever U.In:Proceedings of the eleventh international offshore and polar engineering conference,Stavanger;2001.p.227–33.[17]Hertzberg RW.Deformation and fracture mechanics of engineering materials.Wiley;1995.p.680.[18]Bhadeshia HKDH,DebRoy T.Sci Technol Weld Joining2009;14:193–6.[19]Fairchild D,Kumar A,Ford S,Nissley N,Ayer R,Jin H,et al.In:Proceedings of the8th international conference;2009.p.371–80.[20]BS7448-1,Fracture mechanics toughness tests.London:British Standard Institution;1991.[21]Usiminas.Inspection certificate no.1247149.[22]Ju J,Lee J,Jang J.Mater Lett2007;61:5178–80.[23]ASTM1290-10:Standard test method for crack-tip opening displacement(CTOD)fracture toughness measurement.West Conshohocken:AmericanSociety for Testing and Materials;2010.[24]Shanmugam S,Ramisetti NK,Misra RDK,Hartmann J,Jansto SG.Mater Sci Engng A2008;478:26–37.[25]Ale RM,Rebello JMA,Charlier J.Mater Charact1996;37:89–93.[26]Krauss G,Thompson SW.ISIJ Int1995;35:937–45.[27]Bonnevie E,Ferrière G,Ikhlef A,Kaplan D,Orain JM.Mater Sci Engng A2004;385:352–8.[28]Ikawa H,Oshige H,Tanoue T.Trans Jpn Weld Soc1980;11:87–96.[29]Santos TFA,Ramirez AJ.Brazilian synchrotron light laboratory(LNLS)internal report.。