Materials and Design 87 (2015) 405–413

Contents lists available at ScienceDirect

Materials and Design

j ourna l homepage: www.e lsev ie r .com/ locate / jmad

Microstructure and residual stress distributions in friction stir welding ofdissimilar aluminium alloys

Hamed Jamshidi AvalDepartment of Mechanical Engineering, Babol University of Technology, Shariati Avenue, PO Box 484, Babol, Iran

E-mail address: h.jamshidi@nit.ac.ir.

http://dx.doi.org/10.1016/j.matdes.2015.08.0500264-1275/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 19 January 2015Received in revised form 4 August 2015Accepted 11 August 2015Available online 14 August 2015

Keywords:Friction stir weldingDissimilar aluminium alloysResidual stressMicrostructures

The aim of this investigationwas to study the effect of welding heat input and postweld natural aging on residualstress, microstructure, and precipitation distribution in different zones of dissimilar friction stir welding of 8 mmthick plates of AA6082-T6 and AA7075-T6. It was found that atomic diffusion occurs at the interface of themate-rials in the stir zone of the joints. Transmission electronmicroscopic investigations showed that reprecipitation offine Guinier–Preston zone, β′, and η′ precipitates resulted in increased micro-hardness in the SZ after naturalaging. An increase in welding heat input resulted in decreased maximum tensile residual stress and increasedsize of the tensile residual stress region. Natural aging within the SZ and thermo-mechanical affected zoneresulted in 15–20 MPa reduction of the residual stress in these zones.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Friction stir welding (FSW) is a solid-statewelding technique, devel-oped by TheWelding Institute (TWI), UK, which has demonstrated veryhigh suitability for the joining of dissimilar aluminium alloys. As there isnomelting during this process, all the defects related to the presence ofbrittle interdendritic and eutectic phases are eliminated; thus, frictionstir welds may be produced with superior properties when comparedwith fusion welds [1]. The strength and microstructure of friction stirwelded joints are dependent on the employed welding parameterssuch as rotational and welding speed, geometric parameters of thetool and thework piece, and initial material temperature. In this regard,some studies on dissimilar FSW of aluminium alloys including AA6xxxand AA7xxx series were reported in the literatures [2–5]. For instance,Steuwer et al. [6] have examined the effect of process parameters onresidual stress in dissimilar FSW of AA5083–AA6082. They showedthat the rotational speed of the welding tool affects the residual stressin the weld more significantly when compared with the transversespeed. Moreover, they reported that the residual stress on the AA5083side is larger when compared with the AA6082 side. Jonckheere et al.[7] studied the torque, temperature, and hardening precipitation evolu-tion of 4.7 mm thick similar and dissimilar friction stir welds between6061-T6 and 2014-T6 aluminium alloys. They found that the power re-quired to perform dissimilar welds lies in between the torques requiredto perform similar welds of both alloys, and so do the temperatures.Furthermore, they reported that an increase in temperature causes anincrease in precipitate's radius and a decrease in precipitate's volume

fraction leading to lower yield strength. Ouyang and Kovacevic [8]studied the material flow and microstructural evolution in the dissimi-lar friction stirwelded AA6061–AA2024 joint. They classified the nuggetzone into three different regions: the mechanically mixed region(MMR) characterized by relatively dispersed particles of differentalloy constituents, the stirring-induced plastic flow region (SPFR)consisting of alternative vortex-like lamellae of the two Al-alloys, andthe unmixed region (UMR) consisting of fine equiaxed grains of the6061 aluminium alloy. They reported that the degree of materialmixing, the thickness of the deformed aluminium alloy lamellae, andthematerialflowpatterns dependon the related positions in the nuggetzone and the processing parameters.

Prime et al. [9] studied the residual stress profile of 25.4 mm thickdissimilar friction stir welded AA7050–AA2024. They reported thatthe stress in the test specimen peaked only at about 32 MPa and hadthe conventional “M” profile with tensile stress peaking in the heat-affected zone (HAZ) outside theweld. Da Silva et al. [10] have evaluatedthe effect of joining parameters on the mechanical properties,microstructural features, and material flow of 3 mm thick dissimilarfriction stir welded AA2024–AA7075. They found that the HAZ at theretreating side showed the minimum hardness value. Moreover, theyhave reported that high rotation speed yielded intense onion ring-likemixing pattern in the stir zone (SZ), but a reduced surface integrityand limited material mixing at low rotation speeds has been observed.Cavaliere et al. [11] studied themechanical andmicrostructural proper-ties of dissimilar 2024 and 7075 aluminium sheets joined by FSW. Theyreported that the presence of the FSW line reduces the fatigue behavior,but the comparison to the parent material is acceptable.

Cole et al. [12] have considered the influences of tool offset andtool–work piece interface temperature during dissimilar FSW of

Table 1The alloy chemical composition (wt.%).

Alloy Zn Mg Cu Mn Si Fe Ti Cr Al

AA6082-T6 0.20 1.12 0.14 0.91 1.20 0.34 0.08 0.04 BalancingAA7075-T6 5.21 2.30 1.32 0.10 0.21 0.37 0.12 0.18 Balancing

Table 2Mechanical properties of the base material from tensile tests.

Alloy Yield stress (MPa) Tensile strength (MPa)

AA6082-T6 288 337AA7075-T6 498 560

Table 3Welding parameters used in the experiments.

Designation Rotational speed (rpm) Welding speed (cm/min)

Sample A 1000 9Sample B 1000 12Sample C 1200 9Sample D 1200 12

406 H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

AA6061–AA7075. They reported that the weld tool offsets into theAA7075 side, and increasing amount of AA7075 stirred into the nuggetincreases the measured tensile strength. Guo et al. [13] studied the ef-fects of heat input on thematerial flow, microstructure, andmechanicalproperties of the dissimilar friction stir welded AA6061–AA7075. Theyfound that in conditions that AA6061 is located in the advancing side,better mixing of the material in the SZ was achieved. Furthermore,they reported that the tensile strength of the joints increases with thedecrease of heat input and corresponds very well to the minima in themicro-hardness profile. Bala Srinivasan et al. [14] investigated themicrostructural aspects and stress corrosion cracking (SCC) behaviorof dissimilar friction stir welded AA7075–AA6056 joint. They foundthat the mechanical mixing of the materials at elevated temperaturesin the SZ has led to the evolution of a fine-grained recrystallized struc-ture, but the dwell time at these high temperatures was insufficientfor diffusion of elements during friction stir processing. Moreover,they reported that though the weld nugget is resistant to SCC, thethermo-mechanical affected zone (TMAZ)/HAZ region of AA7075 isprone to SCC in 3.5% sodium chloride (NaCl) solution at lower strainrate than 10−7/s. İpekoğlu and Çam [15] have considered the effects ofinitial temper conditions and postweld heat treatment (PWHT) on themicrostructure and mechanical properties of the dissimilar friction stirwelded AA7075–AA6061 joint. They showed that PWHT could be usedin order to improve the joint properties for both O and T6 joints.

AA6082 and AA7075 Al alloys are increasingly used in automotive,rail transportation and aerospace industries. Joining of these two alloysis required in some of these applications, while AA7075 and AA6082

Fig. 1. The tool geometry

alloys cannot be welded by conventional welding processes, as fusion.FSW as a solid state welding process is capable of producing high-quality welds when optimized parameters are used. In this work, theeffects of rotational and linear speeds of tool on residual stress profilein dissimilar FSW of AA6082-T6 and AA7075-T6 are examined, andthe developed microstructures and precipitation distribution in theweld zone as well as themechanical properties of the joints are evaluat-ed. Accordingly, X-ray diffraction (XRD) residual stress analysis and ten-sile testing together with optical metallography and transmissionelectron microscopy (TEM) are performed to assess the effects ofprocess parameters on welded joints.

2. Experimental procedures

In this study, aluminium alloy plates AA7075-T6 and AA6082-T6 of8 mm thickness were investigated. The composition in weight percentand mechanical properties of both alloys are presented in Tables 1 and2. Welds were produced using a FSW machine in load control with atilt angle and axial load of 2° and 12 kN, respectively. Theweld configu-ration had the AA6082 alloy placed on the advancing side and theAA7075 alloy on the retreating side. Single-pass friction stir butt weldsweremade using a tool made of H13 steel with a shoulder 23mm in di-ameter and triangular frustum pin. The dimensions of the tool used inthis study are shown in Fig. 1. The tool rotational speeds of 800, 1000,1200, and 1400 rpm and linear speeds of 9, 12, and 15 cm/minwere ap-plied in the experiments, but only the welds made by using the rotatingspeeds of 1000 and 1200 rpm and the linear speeds of 9 and 12 cm/minwere satisfactory. Therefore, the last rotational and linear speeds wereinvestigated in this study. The welded samples were named accordingto Table 3. The thermal history of various regions in the advancingand retreating sides of the joints was measured using six K-typethermocouples with a 0.25 mm diameter wire as shown in Fig. 2.

used in this study.

Fig. 2. Schematic view of the samples employed in welding experiments and the thermo-couple positions.

Table 4Measured peak temperature and grain sizes of stir zone in the advancing and retreatingsides of various samples.

Sample Grain size of SZ (μm) Peak temperature (°C)

Advancing side Retreating side Advancing side Retreating side

A 7 5.9 471 460B 4.9 3.4 446 431C 8.4 7.2 493 481D 6.1 4.7 463 449

407H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

To determine the microstructure of the joints, the welded sampleswere transversely sectioned and polished. The optical metallographywas performed on polished samples using a reagent composed of 3 mlnitric acid (HNO3), 6 ml hydrofluoric acid (HF), 6 ml hydrochloric acid(HCl), and 150 ml H2O. In addition, thin sheets for TEM were preparedby means of mechanical polishing (thinning the sample piece to about0.2 mm) and double jet electropolishing. Themechanical polished sam-pleswere electropolished using a solution of 25%HNO3 inmethanol at avoltage of 12 V and temperature of−20±2 °C.Microstructural charac-terization was performed using an optical microscope Olympus PME3and Tecnai G2 transmission electron microscope operating at 200 kV.Besides, distribution of elements in the SZ was studied by Electron

Fig. 3. Macrostructure of welded samples; a) sample A (1000 rpm, 9 cm/min), b) samp(1200 rpm, 15 cm/min).

Probe Micro-Analyzer (EPMA). The linear intercept method based onASTM: E112-13 was used to find the average grain size.

The Vickers micro-hardness of the joints was measured along themid-thickness of the plates with a 100 gf loading for 15 s. Tensile testingwas conducted on the specimens with a 50 mm gauge length and12.5 mm width which were prepared according to ASTM: E8M-13standard at a constant cross-head speed of 1 mm/min. The deformationbehavior of tensile samples was studied by a digital image correlation(DIC) method. The longitudinal and transversal residual stress distribu-tions at 30 μmdepth from the top surface of the samples and perpendic-ular towelding directionweremeasured byXRD technique using Cr–Kα

radiation with the X-ray tube operating at 20 kV and 4 mA targetcurrents.

3. Results and discussion

The macrostructure of cross-section of the abovementioned speci-mens is shown in Fig. 3. It can be seen that no onion ring feature wasfound in the SZ. From Fig. 3(c) and (d) it is apparent that higher rotationspeeds and lower traverse speeds result in the material mixing of com-plex patterns. However, this phenomenon is not more sensible at thelower rotational speed of 1000 rpm. As shown in Table 4, the better

le B (1000 rpm, 12 cm/min), c) sample C (1200 rpm, 9 cm/min), and d) sample D

Fig. 4. Zinc content in samples A and C, measured in the points indicated in Fig. 3.

Fig. 5.Microstructure of different zones of sample B; a) BM of AA6082, b) BM of AA7075, c) HAand h) SZ of AA7075.

408 H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

mixing of the materials in the weld nugget may be attributed to thehigher heat generation and peak temperature of samples A and C thanthe others. Although, mixing of the materials qualitatively can bedetected due to the different etching behaviors of the base materials,to evaluate the distribution of elements in the cross-section of welds,EPMA analysis was performed. The quantitative analysis of zinc in sam-ples A andC on the points specified in Fig. 3(c) and (d) is shown in Fig. 4.Comparing the results, it is possible to confirm that zones 1 and 9 corre-spond to the base materials AA6082 and AA7075, respectively. The zinccontent of zones 2 and 3 suggests that these zones are composed basi-cally of AA6082, while zones 7 and 8 composed basically of AA7075.The zinc content of regions 4, 5, and 6 suggests that these zones areformed by a mixture of both materials. The zinc content between twobase materials can be related to atomic diffusion at the interface ofthese joints. The materials in the weld nugget experience frictionaland plastic deformation heating and intense plastic deformation duringFSW. Therefore, the diffusion rate of zinc atoms due to severe plastic de-formation of thematerial and high peak temperature in theweld nugget

Z of AA6082, d) HAZ of AA7075, e) TMAZ of AA6082, f) TMAZ of AA7075, g) SZ of AA6082,

Fig. 6. a) Micro-hardness vs. aging time of sample B, b) comparison between the micro-hardness of samples B and C after 365 days of natural aging.

409H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

can result in the increase of the diffusion rates higher than that in thestatic condition. Similar results for the atomic diffusion in the interfaceof samples during FSWof Al andMg alloys have been previously report-ed [16].

Fig. 5 illustrates the detailed microstructure in various zones ofsample B. Fig. 5(a) and (b) presents the base metal microstructure ofAA6082 and AA7075, respectively. The average grain sizes of AA6082and AA7075 are approximately 85 μm and 65 μm, respectively. The mi-crostructures of the HAZ in both sides are similar to the base materialsas shown in Fig. 5(c) and (d). Fig. 5(e) and (f) represents the TMAZ onthe advancing and retreating sides, respectively. The SZ and HAZ werebounded by the TMAZ,where bent and elongated grainswere observed.As can be seen in the images, the bent and elongated TMAZ grains in theadvancing side due to large relative deformation in this zone [17] aremore visible than those in the retreating side. The microstructure ofthe SZ experiences high temperature and severe plastic deformation,thus the microstructure of this zone is characterized by recrystallizedequiaxed grains, while the grain size in the AA7075 side is finer thanthat in the AA6082 side (Fig. 5(g), (h) and Table 4). The SZ grain sizedifference was also reported by Guo et al. [13] and Bala Srinivasanet al. [14] in dissimilar FSW of 6000 and 7000 series aluminium alloys.They attributed this difference to the initial grain size of the base

materials [14] and the higher content of second phase particles inAA7075 alloy [13]. Besides, the SZ grain size in the FSWalsowas affectedby plastic deformation rate and by welding temperature [18,19].Therefore, the coarser SZ grain size of AA6082 may be due to finerbase material grain size and higher alloying elements (and higher sec-ond phase particles) of AA7075 than AA6082 and higher temperaturein the advancing side (Table 4). As seen in Table 4, increasing ratio ofrotational to linear speed enhances the SZ grain size and peak tempera-ture in both advancing and retreating sides; for example, the grain sizesof the SZ in the AA6082 side of sample B with welding parameters1000 rpm and 12 cm/min and sample C with welding parameters1200 rpm and 9 cm/min are 4.9 μmand 8.4 μm, respectively. The plasticdeformation rate andwelding temperature are two parameters that actinversely on dynamic recrystallized grain size. In other words, the fineSZ grain size is produced by higher plastic deformation rate and lowerwelding temperature. Although increasing rotational to linear speedinduced more plastic deformation rate, the peak temperature effect onthe SZ grain size is dominant on deformation rate at the range ofrotational and linear speeds investigated in this paper.

Fig. 6(a) shows the micro-hardness profile of sample B measuredalong mid-thickness of the joints at six different natural aging periodsafter FSW. The base materials AA6082-T6 and AA7075-T6 show the

Fig. 7. TEM bright field images of different zones in sample B; a) base material of AA7075-T6, b) base material of AA6082-T6, c) TMAZ of AA7075 side, d) TMAZ of AA6082 side, e) SZ ofAA7075 side, and f) SZ of AA6082 side.

410 H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

average Vickersmicro-hardness values of about 95HV0.1 and 165HV0.1,respectively. The hardness profile changes in the SZ and TMAZ after nat-ural aging weremore significant than the other regions. The increase inmicro-hardness value after 365 days of natural aging within the SZ andTMAZ of the AA6082 side and the SZ of the AA7075 side compared tomicro-hardness value of these zones in as welded condition is notice-able. However, no obvious changes were observed in the TMAZ of theAA7075 side. The change in hardness in the SZ and TMAZ of weldedsamples is due to reprecipitation after natural aging. The micro-hardness of samples B and C after 365 days of natural aging is comparedin Fig. 6(b), and related TEM images of various regions from the basemetal to the dynamically recrystallized zone of sample B are summa-rized in Fig. 7. As can be seen in sample C, with the highest heat input,softening takes places at higher distances and maximum micro-hardness in both sides of sample C, which is lower than those in sampleB. It is worth noting that the SZ hardness decrease in the AA7075 side at2mm from theweld center line of sample C is related to AA6082 flow tothe AA7075 side due to stirring effect of pin as shown in Fig. 3(c). As it

was found previously described in the literature, in the Al–Mg–Si6XXX series aluminium alloys, the decomposition of supersaturatedsolid solution (SS) proceeds in the following steps: the formation ofneedle-like zones along the ⟨100⟩Al crystal direction; the ordering oftheir structure and the formation β″ needles; the transition from β″needles to β′ rods; and the formation of equilibrium β phase [20]. Inthe Al–Zn–Mg–(Cu) system (7XXX alloys) with Mg:Zn lower than 0.5,the precipitation sequence is generally reported as SS → Guinier–Preston (GP) zones (MgZn)→ η′ → η(MgZn2) [21]. Most of its strengthis associatedwith precipitation of GP zones and β″ and η′ semi-coherentphases. In the base material of AA7075-T6 (Fig. 7a), the TEM micro-graphs revealed a fine distribution of precipitates 10–90 nm in size. Inprevious studies by Rhodes et al. [22] and Su et al. [23], the precipitateswith a size less than 20 nmare GP zones and the coarser precipitate is η′.Besides, the AA6082-T6 basematerial presents needle-type precipitates(Fig. 7b), and the size of the precipitates was found to be in a range of10–50 nm length. The morphology of these precipitates indicates thatthey are β″ type [24,25]. In the TMAZ of the AA7075 side, the small

Fig. 8. Schematic representation of the precipitation sequence within the SZ and TMAZ.

Fig. 9. Residual stress profile of samples B and C; a) longitudinal residual stress, b) transverse residual stress, and c) longitudinal residual stress vs. aging time of sample B.

411H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

Fig. 10. a) The strain maps, and b) the local mechanical properties of different weld sub-zones for sample B.

412 H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

precipitates had dissolved and only round particles with an average di-ameter in the order of 70–250 nmremain (Fig. 7(c)). Fig. 7(d) illustratesthat needle-type precipitates from base material AA6082 becamecoarser and wider spaced with a round precipitate in the order of60 nmdiameter, and rodmorphology precipitates 80–250 nm in lengthand 9–15 nm in width in the TMAZ. With respect to size, morphology,and studies by Sato et al. [26] andOhmori et al. [27], the rod precipitatesare defined asβ′ precipitates and round precipitates belong to the Si andβ phases. The type andmorphology changes of the TMAZ precipitationsin response to the heat input and the deformation rate from FSW arereasonswhy this zone shows an abrupt hardness decrease as comparedto the base materials. As shown in Fig. 7(e), the SZ of the AA7075 side

Table 5Measured temperature of different zones of samples B and C.

Sample Temperature (°C)

Advancing side (AA6082 side) Retreating side (AA7075 side)

SZ TMAZ HAZ SZ TMAZ HAZ

B 446 411 307 431 389 295C 493 463 369 481 448 356

shows three different precipitates which were identified as GP zones,round type η′, and rod shape η precipitates. Besides, the SZ in theAA6082 side presents fine round type β and Si and rod β′ precipitates(Fig. 7(f)). The increase in micro-hardness in the SZ after natural agingcan be related to reprecipitation of fine precipitates β′ and η′ and GPzones. The schematic representation of the precipitation sequencewith-in the SZ and TMAZ is shown in Fig. 8. Curves 1 and 2 indicate thewelding thermal cycles in the TMAZ and SZ, respectively. In the TMAZ,growth of th base metal precipitates (GP and η′ in AA7075 and β″ inAA6082) occurred, whereas reprecipitation of β′, GP and η′ occurredin the SZ during cooling.

Fig. 9 shows the residual stress profile of samples welded with thelowest and highest heat inputs, that are, samples B and C, respectively.As shown in Fig. 9, for both samples longitudinal residual stress in thecenter of the weld in both advancing and retreating sides has tensilenature and away from the weld center first increases up to maximumtensile stress near the edge of the tool then in farther distance becomescompressed. The overall behavior of transversal residual stress is similarto longitudinal residual stress, but in the weld center line, transversalstress has compression nature and maximum residual stress isapproximately two times less than the longitudinal residual stress.The residual stress profiles of the longitudinal and transverse directions

413H. Jamshidi Aval / Materials and Design 87 (2015) 405–413

are asymmetric, and the maximum tensile residual stress occurs in theAA7075 side. Another important result (Fig. 9) is that with increasingwelding heat input, the maximum tensile residual stress decreases,and the size of the tensile residual stress region on either side of theweld line increases; for example, themaximum tensile residual stressesof the AA6061 side of samples B and C are 104MPa and 97MPa, respec-tively. Besides, the sizes of the tensile residual stress region in samples Band C are 48 mm and 67mm, respectively. The results of the longitudi-nal residual stress measured in sample B after two different naturalaging periods are shown in Fig. 9(c). It can be seen that there is a notice-able decrease in the residual stress from 7 to 365 days of natural agingwithin the SZ and TMAZ. Duringnatural aging, themicrostructuralmod-ification due to the vacancies and solute atom flowmay cause the creepbehavior and stress relaxation [28]. Therefore, the lower residual stressof the SZ and TMAZ after 365 days of natural agingmay be attributed tostress relaxation in these zones due to the natural aging phenomena.The strain maps correspond to strain distribution at maximum load,and the related local mechanical properties of different weldingsubzones for sample B achieved by DIC method are shown in Fig. 10.According to the strain maps, the related macrostructure of tensilesample is shown in Fig. 10(a). The tensile sample fracture in the TMAZof theAA6082 sidewhereminimumhardness and tensile residual stressin theweld zone are observed is illustrated in Figs. 6 and 9. It can be seenthat the residual stress distribution at the region between theHAZ of theAA6082 and AA7075 sides is a function of the local mechanical proper-ties and the fluctuating inhomogeneous temperature distributionsduring and after welding. In other words, the SZ, TMAZ, and HAZ ofthe AA6082 side with lower residual stress profile show higher peaktemperature (Table 5) and lower micro-hardness and yield stress thanthe related regions in the AA7075 side (Figs. 6 and 10(b)). Besides,AA6082 at moderate temperatures has a rather lower yield stress thanAA7075; therefore, the residual stress build up in the AA6082 side willbe lower than the AA7075 side [6,29].

4. Conclusion

The influence of rotational and linear speeds of triangular frustumpin on residual stress and microstructure of dissimilar friction stirwelded AA6082-T6 and AA7075-T6 is examined by optical microscopy,TEM, micro-hardness, and XRD residual stress tests. The results can besummarized as follows:

1- Zinc distribution across the weld shows that with increasing weldingheat input per unit length, themixing of thematerial in theweld nug-get of the joints is performedmore efficiently, anddue to severe plasticdeformation of the material and high-peak temperature in the weldnugget, the atomic diffusion occurs at the interface of the materials.

2- In allwelding heat input, the grain size in the AA7075 side isfiner thanthat in the AA6082 side. The peak temperature effect on the SZ grainsize is dominant on deformation rate.

3- Coarsening and dissolution of the strengthening precipitates presentin the AA6082 and AA7075 base metals result in softening in theweld zone. Besides, reprecipitationoffineGP,β′, and η′precipitates re-sults in an increase of micro-hardness in the SZ after natural aging.

4- The residual stress distribution at different zones is a function of thelocal mechanical properties and the temperature distributions of dif-ferent welding subzones. An increase in welding heat input resultedin decreased maximum tensile residual stress and increased size ofthe tensile residual stress region on both sides of the weld line.

5- The longitudinal residual stress in the SZ has tensile nature, and awayfrom theweld centerfirst increases up tomaximum tensile stress nearthe edge of the tool, then in farther distance becomes compressed,while transverse residual stress in the weld center line has compres-sion nature and maximum residual stress is approximately twotimes less than the longitudinal residual stress.

References

[1] W.M. Thomas, E.D. Nicholas, J.C. Needham, M.G. Murch, P. Templesmith, C.J. Dawes,Friction Stir Welding, 1991.

[2] P. Cavaliere, A. De Santis, F. Panella, A. Squillace, Effect of welding parameters onmechanical and microstructural properties of dissimilar AA6082–AA2024 jointsproduced by friction stir welding, Mater. Des. 30 (2009) 609–616.

[3] R. Palanivel, P. Koshy Mathews, N. Murugan, I. Dinaharan, Effect of tool rotationalspeed and pin profile on microstructure and tensile strength of dissimilar frictionstir welded AA5083-H111 and AA6351-T6 aluminum alloys, Mater. Des. 40(2012) 7–16.

[4] Youbao Song, Xinqi Yang, Lei Cui, Xiaopeng Hou, Zhikang Shen, Yan Xu, Defect fea-tures and mechanical properties of friction stir lap welded dissimilar AA2024–AA7075 aluminum alloy sheets, Mater. Des. 55 (2014) 9–18.

[5] P. Cavaliere, F. Panell, Effect of tool position on the fatigue properties of dissimilar2024–7075 sheets joined by friction stir welding, J. Mater. Process. Technol. 206(2008) 249–255.

[6] A. Steuwer, M.J. Peel, P.J. Withers, Dissimilar friction stir welds in AA5083–AA6082:the effect of process parameters on residual stress, Mater. Sci. Eng. A 441 (2006)187–196.

[7] Caroline Jonckheere, Bruno deMeester, Anne Denquin, Aude Simar, Torque, temper-ature and hardening precipitation evolution in dissimilar friction stir welds between6061-T6 and 2014-T6 aluminum alloys, J. Mater. Process. Technol. 213 (2013)826–837.

[8] J.H. Ouyang, R. Kovacevic, Material flow and microstructure in the friction stir buttwelds of the same and dissimilar aluminum alloys, J. Mater. Eng. Perform. 11(2002) 51–63.

[9] M.B. Prime, T.G. Herold, J.A. Baumann, R.J. Lederich, D.M. Bowden, R.J. Sebring, Resid-ual stress measurements in a thick, dissimilar aluminum alloy friction stir weld, ActaMater. 54 (2006) 4013–4021.

[10] A.A.M. da Silva, E. Arruti, G. Janeiro, E. Aldanondo, P. Alvarez, A. Echeverria, Materialflow andmechanical behaviour of dissimilar AA2024-T3 and AA7075-T6 aluminiumalloys friction stir welds, Mater. Des. 32 (2011) 2021–2027.

[11] P. Cavaliere, R. Nobile, F.W. Panella, A. Squillace, Mechanical andmicrostructural be-haviour of 2024–7075 aluminium alloy sheets joined by friction stir welding, Int. J.Mach. Tools Manuf. 46 (2006) 588–594.

[12] E.G. Cole, A. Fehrenbacher, N.A. Duffie, M.R. Zinn, F.E. Pfefferkorn, N.J. Ferrier, Weldtemperature effects during friction stir welding of dissimilar aluminum alloys6061-T6 and 7075-T6, Int. J. Adv. Manuf. Technol. 71 (2014) 643–652.

[13] J.F. Guo, H.C. Chen, C.N. Sun, G. Bi, Z. Sun, J. Wei, Friction stir welding of dissimilarmaterials between AA6061 and AA7075 Al alloys effects of process parameters,Mater. Des. 56 (2014) 185–192.

[14] P. Bala Srinivasan, W. Dietzel, R. Zettler, J.F. dos Santos, V. Sivan, Stress corrosioncracking susceptibility of friction stir welded AA7075–AA6056 dissimilar joint,Mater. Sci. Eng. A 392 (2005) 292–300.

[15] Güven İpekoğlu, Gürel Çam, Effects of initial temper condition and postweld heattreatment on the properties of dissimilar friction-stir-welded joints betweenAA7075 andAA6061 aluminumalloys,Metall.Mater. Trans. A 45 (2014) 3074–3087.

[16] Y.S. Sato, S.H.C. Park, M. Michiuchi, H. Kokawa, Constitutional liquation during dis-similar friction stir welding of Al and Mg alloys, Scr. Mater. 50 (2004) 1233–1236.

[17] H.J. Liu, J.J. Shen, Y.X. Huang, L.Y. Kuang, C. Liu, C. Li, Effect of tool rotation rate onmicrostructure and mechanical properties of friction stir welded copper, Sci.Technol. Weld. Join. 14 (2009) 577–583.

[18] L. Fratini, G. Buffa, CDRXmodelling in friction stir welding of aluminium alloys, Int. J.Mach. Tools Manuf. 45 (2005) 1188–1194.

[19] G. Buffa, L. Fratini, R. Shivpuri, CDRXmodelling in friction stir welding of AA7075-T6aluminum alloy: analytical approaches, J. Mater. Process. Technol. 191 (2007)356–359.

[20] G.E. Totten, D.S. MacKenzie, Handbook of Aluminum: Physical Metallurgy andProcesses, CRC Press, 2003.

[21] G.W. Lorimer, R. Nicholson, Further results on the nucleation of precipitates in theAl–Zn–Mg system, Acta Metall. 14 (1966) 1009–1013.

[22] C.G. Rhodes, M.W. Mahoney, W.H. Bingel, R.A. Spurling, C.C. Bampton, Effects of fric-tion stir welding onmicrostructure of 7075 aluminum, Scr. Mater. 36 (1997) 69–75.

[23] J.Q. Su, T.W. Nelson, R. Mishra, M. Mahoney, Microstructural investigation of frictionstir welded 7050-T651 aluminium, Acta Mater. 51 (2003) 713–729.

[24] G.A. Edwards, K. Stiller, G.L. Dunlop, M.J. Couper, The precipitation sequence in Al–Mg–Si alloys, Acta Mater. 46 (1998) 3893–3904.

[25] C.A.W. Olea, L. Roldo, J.F. dos Santos, T.R. Strohaecker, A sub-structural analysis offriction stir welded joints in an AA6056 Al-alloy in T4 and T6 temper conditions,Mater. Sci. Eng. A 454–455 (2007) 52–62.

[26] Y.S. Sato, H. Kokawa, M. Enomoto, S. Jogan, Microstructural evolution of 6063 alumi-num during friction-stir welding, Metall. Mater. Trans. A 30 (1999) 2429–2437.

[27] Y. Ohmori, L.C. Doan, K. Nakai, Ageing processes in Al–Mg–Si alloys during continu-ous heating, Mater. Trans. 43 (2002) 246–255.

[28] W. Woo, Z. Feng, X.-L. Wang, C.R. Hubbard, Neutron diffraction measurements oftime-dependent residual stresses generated by severe thermomechanical deforma-tion, Scr. Mater. 61 (2009) 624–627.

[29] G.Z. Quan, G.S. Li, Y. Wang, W.Q. Lv, C.T. Yu, J. Zhou, A characterization for the flowbehavior of as-extruded 7075 aluminum alloy by the improved Arrhenius modelwith variable parameters, Mater. Res. 16 (2013) 19–27.