Fatigue Behaviour of FSW and MIG Weldments

8/9/2019 Fatigue Behaviour of FSW and MIG Weldments

1/9

Fatigue behaviour of FSW and MIG weldmentsfor two aluminium alloys

P.M.G.P. Moreira *, M.A.V. de Figueiredo, P.M.S.T. de CastroFEUP, Faculty of Engineering, University of Porto, R. Dr. Roberto Frias, 4200-465 Porto, Portugal

Abstract

The increasing use of aluminium alloys in transportation, such as railways, shipbuilding and aeronautics, calls for moreefficient and reliable welding processes that would require more in depth understanding of fatigue failure. The objective of this work focuses on the contrasting difference of fatigue behaviour of joints made from the traditional process of metalinert gas (MIG) welding, and the emerging process of friction stir welding (FSW). Effort is made to relate the macroscopicmechanical behaviour to the microstructural feature of the weldments.

2007 Elsevier Ltd. All rights reserved.

Keywords: Aluminium alloy; Fatigue life; Friction stir welding; MIG; Scanning electron microscopy

1. Introduction

Conventional fusion welding of aluminium alloysproduces a weld prone to defects such as porosity,consequence of entrapped hydrogen gas not beingable to escape from the weld pool during solidica-tion. An example of a fusion process is the metalinert gas (MIG) welding [1]. In the MIG weldingprocess, the arc and the weld are protected fromatmospheric contamination by a gas shield, and an

electric potential is established between the elec-trode and the work piece causing a current ow,which generates thermal energy in the partially ion-ized inert gas. Friction stir welding (FSW) is a solid-state joining process developed and patented by The

Welding Institute (TWI), intended to be used inhigh strength alloys that were difficult to join withconventional techniques. In FSW, the interactionof a non-consumable and rotating tool with theworkpieces being welded creates a welded jointtrough frictional heating and plastic deformationat temperatures below the melting temperature of the alloys being joined.

In this work, a study of the fatigue behaviour of friction stir (FS) butt welds of two 3 mm thickness

age hardenable aluminium, 6082-T6 and 6061-T6alloys, was carried out. For comparison, MIG buttwelds of the same alloys were also performed andtested. Tensile tests and microhardness measure-ments of weld joints and base materials were per-formed in order to determine the inuence of eachwelding process in the mechanical properties. Thefatigue behaviour (SN curves) of specimens of bothwelding process was analyzed. Microstructure wasexamined and correlated with the macroscopic

0167-8442/$ - see front matter 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.tafmec.2007.06.001

* Corresponding author.E-mail address: [email protected] (P.M.G.P. Moreira).

Theoretical and Applied Fracture Mechanics xxx (2007) xxxxxxwww.elsevier.com/locate/tafmec

ARTICLE IN PRESS

Please cite this article in press as: P.M.G.P. Moreira et al., Fatigue behaviour of FSW and MIG weldments ..., Theor.Appl. Fract. Mech. (2007), doi:10.1016/j.tafmec.2007.06.001
mailto:[email protected]:[email protected]


2/9

mechanical behaviour. Scanning electron micros-copy (SEM) was carried out and the fractographicfeatures of both types of welds compared.

2. Welding processes

The MIG welding parameters used were: 128 A,17.1 V, 700 mm/min and Argon at a 20 l/min ow.A ller wire AWS ER5356 with a diameter of 1 mm [2] was used. The friction stir welds were per-formed using the same parameters for both alloyswere: welding speed of 800 mm/min; pitch angle of 2 ; rotating speed of 1500 rpm. The FSW processof the Al6082-T6 was performed using a tool witha 6 mm diameter threaded pin and the shoulderhad 15 mm diameter. For the Al6061-T6 a tool witha 4 mm diameter threaded pin and a shoulder of 10 mm diameter were used.

3. Tensile tests

Tensile tests were performed to determine themechanical properties of the welded and unweldedmaterial (yield stress r y, rupture stress r t and Youngmodulus E ). The stress/strain records of all tensiletests are plotted in Fig. 1 and the principal materialproperties are presented in Table 1 . It was foundthat the FSW specimens have lower yield stress val-

ues than MIG specimens. Nevertheless the rupturestress of FSW specimens presents higher values.Also the elongation of the friction stir welded spec-imens presented higher values. The Al6061-T6 hasan ultimate tensile stress about 6% higher than theAl6082-T6. All base material (BM) specimens failedin the same manner, 45 shear plane.

For the case of MIG welded specimens ruptureoccurred outside the welding seam in the heat

affected zone (HAZ). Similar observations are pre-sented by Ericsson and Sandstrom [3] where a 45fractured surface was also found. The Al6082-T6MIG welded specimens present a yield stress and arupture stress of 65% of the base material. The yieldstress obtained in this work is 20% higher than theresults presented by [3]. The Al6061-T6 MIGwelded specimens has a yield stress of 51% of thebase material and a rupture stress of 65% of the basematerial. Comparing MIG weldments of both alloysit was found that the yield stress of the 6082-T6 ishigher but the rupture stress is higher for the6061-T6 MIG welded specimens.

In the case of the 6082-T6 friction stir weldedspecimens fracture occurred near the weld edge,where a decrease of hardness occurs [4]. The frac-ture surface presents a 45 angle, as presented in

[5]. In the case of the Al6061-T6 the fracture startedat the weld root indicating that a lack of penetrationoccurred during welding (root aw). Dickerson andPrzydatek [6] suggested that root aws up to0.35 mm deep do not cause degradation in mechan-ical performance when compared to aw-free welds.

The Al6082-T6 friction stir welded specimenspresent a yield stress of 51% and a rupture stressof 70% of the base material. Obtained in [4] is alsoa relation of rupture stress of 76% between the basematerial and friction stir welded specimens. Valuesof the same magnitude are also reported in [3,7,8].The Al6061-T6 friction stir welded specimens havea yield stress of 52% and a rupture stress of 71%of the base material. Performed in [9] is the frictionstir weld of 4 mm thick Al6061-T651 plates. Theyield stress and rupture stress were found to belower compared with those obtained in this study.

4. Microhardness proles

Microhardness tests were performed to charac-terize the Vickers hardness prole in the vicinity of the weld area. Measurements were performed atFig. 1. Tensile tests of MIG and FS welded specimens.

Table 1Material properties for FS welded specimens, data acquired intensile tests

r y (MPa) r t (MPa)

Parent 6082-T6 276.2 322.9Parent 6061-T6 306.3 342.0MIG 6082-T6 176.8 210.0MIG 6061-T6 156.3 221.2FSW 6082-T6 140.5 226.1FSW 6061-T6 158.7 241.5

2 P.M.G.P. Moreira et al. / Theoretical and Applied Fracture Mechanics xxx (2007) xxxxxx

ARTICLE IN PRESS



3/9

the specimens middle thickness using a 100 gf load.Fig. 2 illustrates the hardness proles of the MIGwelded Al6061-T6 and Al6082-T6 specimens. Themajor softened areas are the weld centre line, andthe two transitions zones just at the limit of theHAZ. In these areas, the hardness reaches a mini-mum value near 50HV and the base material hasvalues over 90HV. Minimum average values arefound in the Al6082-T6 specimen.

Fig. 3 illustrates a low magnication overview of the friction stir welded Al6082-T6 along with corre-

sponding hardness proles. A hardness decrease isidentied in the thermo mechanically affected zone(TMAZ). The average hardness of the nugget zonewas found to be signicantly lower than the hardnessof the base alloy. There is a zone outside the nuggetzone which has the lower hardness value. The weld-

ing process softened the material reducing the hard-ness to 33% of the parent material, as shown in [4].The hardness minimum values are obtained in thewelding retreating side, e.g. [5]. As it also suggestedin [7] the variation of the microhardness values in

the welded area and parent material is due to the dif-ference between the microstructure of the base alloyand weld zone.

5. SN fatigue data

Fatigue tests were carried out in a servo-hydrau-lic MTS testing machine. Specimens dimensionswere chosen according to the ASTM standardE466. The weld was perpendicular to the load direc-tion in the SN tests and to the material rollingdirection. The maximum stress levels used were cho-sen as a function of the yield stress for each type of joint. Values of 12040% of the yield stress werechosen. A stress ratio of R = 0.1 was used and thefrequency was in the interval of 726 Hz dependingupon the stress level. Life was dened as the numberof cycles to failure and a total of 76 specimens weretested. The number of cycles considered as a thresh-old for innite life was 107 cycles. The fatigue livesresults for the MIG and friction stir welded speci-mens are plotted in Fig. 4.

In tests of friction stir welded specimens, data

show narrow scatter and were tted using a powerequation. For the friction stir welded 6082-T6, itwas found that for 65% and 60% of the yield stressthe fatigue life is considered innite. In friction stirwelded specimens fatigue cracks have propagatedat the centre of the weld, as veried in [5] whenstudying friction stir welded 5 mm and 10 mm thick6082-T6 specimens. Also, a study [3] on the fatiguelife assessment of the same alloy obtained a fatiguelife of 5 105 cycles to failure with a stress range of about 90 MPa, at a stress ratio, R = 0.5. In the pres-ent study, the fatigue life of 5 105 cycles isobtained at a stress range of 105 MPa. The frictionstir welded 6061-T6 specimens presented lower livesthan the friction stir 6082-T6 specimens when testedat stresses lower than 130 MPa. For the friction stirwelded 6061-T6 specimens cracks have propagatednear the limit of the tool shoulder. Fatigue scatter issomewhat higher in the MIG welded specimens anddata were tted using a power equation. Fatiguelives of MIG welded specimens are lower than thoseof friction stir welded specimen. The 6061-T6 MIGwelded specimens presented higher fatigue lives thanthe MIG 6082-T6 specimens.

Fig. 2. Micro hardness proles of the MIG welded Al6061-T6and Al6082-T6 specimens.

Fig. 3. Micro hardness prole of the FS welded Al6082-T6.

P.M.G.P. Moreira et al. / Theoretical and Applied Fracture Mechanics xxx (2007) xxxxxx 3

ARTICLE IN PRESS



4/9


5/9

ences high strain and is prone to recrystallization.Immediately at its side is the TMAZ which endsat the tool shoulder delimited by the dashed lines.After the TMAZ appears a zone affected only bythe heat generated during the welding process, e.g.[4,11]. FSW gives rise to microstructure changes.As observed in [5] the base material contains twosizes of grains owing to partial recrystallization.The recrystallized grains are approximately 20 l min size while the non-recrystallized grains can be lar-ger than 100 l m. Microstructure 3, structure at theweld nugget, shows dynamic recrystallized grainsmuch smaller and equiaxed when compared to theelongated base metal microstructure. The grain evo-

lution between the base material and the weldedaffected material is evident comparing microstruc-tures 3 and 4. The grains in the nugget zone areequiaxed [5]. In the HAZ the grain size is similarto the base metal. Second-phase particles in the

workpiece are essentially stirred into the weld zone.Similar observations are presented in [4]. Thedynamic continuous recrystallization microstruc-ture which characterizes the FSW process of thisalloy is well documented in [12].

7. SEM analysis

Fractured fatigue test specimens were analyzedby SEM. Measurements were performed consider-ing the crack initiation site as the origin and mea-surements were obtained through the crack length.The result in the coordinate is the average of fourmeasurements of the number of striations in a per-pendicular line to their orientation.

7.1. Base material striations spacing

A fatigue test specimen of Al6082-T6 andanother of Al6061-T6 subjected to fatigue tests witha maximum stress of 70% of the yield stress(193.3 MPa for the Al6082-T6 and 214.4 MPa forthe Al6061-T6) and R = 0.1 were analyzed. The

6082-T6 base material specimen analyzed had a fati-gue life of 485,858 cycles and the 6061-T6 a fatiguelife of 783,586 cycles. Fig. 7 shows fatigue striationsat different crack lengths. Figs. 8 and 9 present thestriation spacing vs. crack length. The crack growthcan be tted using an exponential approximation,with reasonable correlation.

7.2. MIG welded specimens, striations spacing

Two MIG butt welded specimens were analyzedafter fatigue testing ( R = 0.1). The specimen of Al6082-T6 was fatigue tested at a maximum stressof 60% of the yield stress (106.1 MPa), and the spec-imen of Al6061-T6 was tested at a maximum stressof 70% of the yield stress (109.4 MPa). The 6082-T6specimen presented a fatigue life of 46,645 cyclesand the 6061-T6 specimen had a fatigue life of 28,476 cycles. In the MIG welded specimens fatiguecracks appeared at the end of the weld seam, nearthe V notch effect due to the extra material depos-ited by the melted feed wire.

In the 6082-T6 MIG welded specimen the stria-tions identication process was very hard to carry

Fig. 6. FS welded Al6082-T6: (a) macrostructure of the FSwelded Al6082-T6; (b) microstructure 3, weld; and (c) micro-structure 4, BM.


ARTICLE IN PRESS



6/9

out. The fatigue crack surface presented a hetero-genic structure with different structures dispersedrandomly. For example, in fractograph 9 presented

Fig. 7. Fatigue striations for different crack lengths: (a) Al6082-T6 fractograph 19 crack length of 3.974 mm and (b) Al6061-T6fractograph 18, crack length of 3.322 mm.

Fig. 8. Fatigue striation spacing vs. crack length for specimen of Al6082-T6.

Fig. 9. Fatigue striation spacing vs. crack length for specimen of

Al6061-T6.

Fig. 10. SEM analysis of the Al6082-T6 MIG welded specimen:(a) fractograph 9 crack length of 2.582 mm and (b) fractograph 3,crack length of 0.316 mm.


ARTICLE IN PRESS



7/9

in Fig. 10 besides the larger striations at the top of the fractograph, a ner type of striations is identi-ed. The coarser striations are of a second orderthat contain several ner striations. Fatigue cracksinitiated and have propagated not at the centre of

the weld but at the end of the deposited material,an area subjected to high temperatures during weld-ing. This high temperature lead to the appearance of disperse micropores (diameter with less than 4 l m),as identied in fractograph 3.

Fig. 11 shows fatigue striations for the Al6061-T6 MIG welded specimen. The fatigue crack surfacewas a heterogenic structure. In fractograph 6,besides fatigue striations, some gaps in the materialare identied. These gaps were probably formed inthe heating and cooling of welding leading to theappearance of internal cracks. These gaps aremicro-cracks that can be signs of small hot cracks.Hot cracks can be formed in AlMgSi fusion weldswhen the low melting MgSi eutectic in the grain

boundaries remains liquid during weld solidicationand subsequent material shrinkage [3]. Heat treat-able aluminium alloys are sensitive to hot shortcracking, which results form HAZ liquidation dur-ing welding [13]. This specimen presents no less

micropores than the Al6082-T6 MIG specimen. Infractograph 12 striations with a V shape orientationwere found. The image shows the random distribu-tions that can be found in striations if the fatiguecrack is observed with high magnications. In thiscase, the two orientations found are probably dueto an internal gap formed during the coolingprocess.

Figs. 12 and 13 present the striation spacing vs.crack length for the Al6082-T6 and Al6061-T6MIG welded specimens.

Fig. 11. SEM analysis of the Al6061-T6 MIG welded specimen:(a) fractograph 6, crack length of 0.611 mm and (b) fractograph12, crack length of 4.886 mm.

Fig. 12. Fatigue striation spacing vs. crack length Al6082-T6MIG welded specimen.

Fig. 13. Fatigue striation spacing vs. crack length Al6061-T6MIG welded specimen.


ARTICLE IN PRESS



8/9

7.3. FSW welded specimens, striations spacing

The specimens were fatigue tested at a maximumstress of 70% of their yield stress (98.4 MPa for the6082-T6 and 111.1 MPa for the 6061-T6). The 6082-

T6 specimen presented a fatigue life of 258,827 cycles and the 6061-T6 specimen had a fati-gue life of 173,499 cycles. The friction stir welded6082-T6 specimen has an irregular structure in itsentire fracture surface preventing the identicationof measurable striations. For the Al6061-T6 frictionstir welded the fatigue crack propagated near theshoulder edge. Despite the heterogeneity found inthe fractured surface striations measurement pro-cess was feasible. In fractograph 5, presented inFig. 14a, several striations sets with different orien-tations are identied.

8. Conclusions

Sound welds have been obtained using MIG andFSW for 3 mm plate thickness of two aluminiumalloys Al6082-T6 and Al6061-T6. Tensile testing of

the weld joints and base material produced ade-quate tensile strength values.Yield and rupture stress of friction stir welded

and MIG welded specimens are lower than for basematerial. All welding processes lead to a decrease of the material mechanical properties, more pro-nounced in the MIG specimens. Detailed hardnessexamination revealed lower hardness values in theMIG welded specimens.

The friction stir welded 6061-T6 specimens pre-sented lower lives than the friction stir 6082-T6specimens when tested at stresses lower than130 MPa. Fatigue scatter is somewhat higher inthe MIG welded specimens. Its fatigue lives arelower than friction stir welded specimen. The MIGwelded 6061-T6 specimens presented higher fatiguelives than the MIG 6082-T6 specimens.

The general microstructure was in good agree-ment with previous published researches. In FSW,the nugget has recrystallized grains smaller thanthe base material. A SEM analysis of fatigue stria-tion measurements was performed for fatigue spec-imens of both welding processes and fatigue crack

growth rate can be estimated from using thesemeasurements.

Acknowledgements

The work was partially supported by PhD schol-arship FCT SFRH/BD/19281/2004 and FP6 projectDaToN (Contract no. AST3-CT-2004-516053 of theEuropean Union). The collaboration of R. Silva, D.Silva and F. Oliveira is gratefully acknowledged.

References[1] M.A. Wahab, M.J. Painter, M.H. Davies, The prediction of

the temperature distribution and weld pool geometry in thegas metal arc welding process, Journal of Materials Process-ing Technology 77 (1998) 233239.

[2] ESAB Welding Handbook Filler Materials for Manual andAutomatic Welding, fth ed., ESAB AB, Goteborg, Sweden,2003.

[3] M. Ericsson, R. Sandstrom, Inuence of welding speed onthe fatigue of friction stir welds, and comparison with MIGand TIG, International Journal of Fatigue 25 (2003) 1379 1387.

[4] A. Scialpi, L.A.C. de Filippis, P. Cavaliere, Inuence of shoulder geometry on microstructure and mechanical prop-

Fig. 14. Fatigue striations for Al6061-T6 FS welded specimen:(a) fractograph 5, crack length of 0.877 mm and (b) fatiguestriation spacing vs. crack length.


ARTICLE IN PRESS



9/9

erties of friction stir welded 6082 aluminium alloy, Materialsand Design 28 (2007) 11241129.

[5] L.E. Svensson, L. Karlsson, H. Larsson, B. Karlsson, M.Fazzini, J. Karlsson, Microstructure and mechanical prop-erties of friction stir welded aluminium alloys with specialreference to AA 5083 and AA 6082, Science and Technologyof Welding & Joining 5 (2000) 285296.

[6] T.L. Dickerson, J. Przydatek, Fatigue of friction stir welds inaluminium alloys that contain root aws, InternationalJournal of Fatigue 25 (2003) 13991409.

[7] D. Harris, A.F. Norman, Properties of friction stir welded joints: a review of the literature, in: Progress Reportpresented at the 6th PSG Meeting, June, 2003, pp. 1718.

[8] E.D. Nicholas, S.W. Kallee, Friction stir welding a decadeon, in: IIW Asian Pacic International Congress, Sydney, 29October2 November, 2000.

[9] S. Hong, S. Kim, C.G. Lee, S.-J. Kim, Fatigue crackpropagation behavior of friction stir welded AlMgSi alloy,Scripta Materialia 55 (2006) 10071010.

[10] Metals Handbook, Metallography and Microstructures,ASM American Society for Metals, vol. 9, Ohio, USA,1985.

[11] W. Woo, H. Choo, D.W. Brown, S.C. Vogel, P.K. Liaw, Z.Feng, Texture analysis of a friction stir processed 6061-T6aluminum alloy using neutron diffraction, Acta Materialia 54(2006) 38713882.

[12] G. Liu, L.E. Murr, C.-S. Niou, J.C. McClure, F.R. Vega,Microstructural aspects of the friction-stir welding of 6061-T6 aluminum, Scripta Materialia 37 (1997) 355361.

[13] ASM Speciality Handbook, Aluminium and AluminiumAlloys, ASM International, Ohio, USA, 1993.


ARTICLE IN PRESS


Fatigue Behaviour of FSW and MIG Weldments

Documents

Transcript of Fatigue Behaviour of FSW and MIG Weldments