Journal of Materials Processing Technology 225 (2015) 347–356

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

journa l homepage: www.e lsev ier .com/ locate / jmatprotec

Influence of process parameters on friction stir spot weldedaluminum joints by various threaded tools

Yuan-Ching Lin ∗, Jiun-Nan ChenDepartment of Mechanical Engineering, National Taiwan University of Science and Technology, 43, Keelung Road, Section 4, Taipei 10673 Taiwan, ROC

a r t i c l e i n f o

Article history:Received 28 March 2015Received in revised form 18 June 2015Accepted 20 June 2015Available online 24 June 2015

Keywords:Friction stir spot weldingAluminum alloyMaterial flowJoint strength

a b s t r a c t

The flow behavior and strength of joints created using three stirring tools, with right-handed threadsat different locations on their cylindrical pins, were investigated. The threading locations on the toolssignificantly influenced the evolution of the stir zone in the welds and caused material intermixturein various degrees at varying plunging depths. Failure joints were observed in the welds without toolrotation. Force measurements indicated that the highest torque was found in the welds at 900 rpm. Notmuch difference was observed in the axial force variation between the welds. The peak temperature wasobtained at the measured point nearest to the tool shoulder. The maximum joint strength of over 6.5 kNwas obtained in the 900 rpm weld with a 9 s dwell time.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

The conventional method for joining aluminum alloys, resis-tance spot welding (RSW), is quite complex because of their highthermal conductivity, consuming power and current (Florea et al.,2012). Welding defects such as pores and cracks may also beinduced, decreasing the weld strength. Therefore, FSSW seems tobe a favorable spot welding process for joining aluminum alloys.

Badarinarayan et al. (2009b) reported that hook geometryplayed an important role in determining the static strength ofwelds. In addition, tool-pin geometry significantly affected thehook shape. Babu et al. (2013) proposed that the mechanical per-formance of FSSW is mainly governed by its geometrical features(hook height and bond width). Zhao et al. (2014) indicated thatthe tensile shear failure load of the FSSW joint is dependent on thehook geometry. Material flow morphology affects the joint strengthbecause of the hook shape, and hook formation would be detrimen-tal to weld strength, as reported by Badarinarayan et al. (2009a).Thus, hook formation is worthy of investigation by observation ofthe evolution of material flow during the plunging period.

Su et al. (2007) proposed that dissimilar intermixing during thedwell period results from the incorporation of upper (Al 5754) andlower (Al 6111) sheet materials at the top of the thread on the rotat-ing pin and its rapid transfer downward into the SZ via threads onthe pin. Fujimoto et al. (2008) indicated that the material near the

∗ Corresponding author. Fax: +886 27376460.E-mail address: yclin@mail.ntust.edu.tw (Y.-C. Lin).

pin was driven downward by the threads, and the material origi-nally from the lower sheet was pushed outward as well as upward.Similarly, Lin et al. (2013) reported that threads on a cylindricalpin could increase the axial material flow during FSSW. The mate-rial flows downward along threads while the pin is rotated in theappropriate direction and then discharges at the pin tip to form acirculating material flow. The pin’s bottom and top are the materialentrance and exit, respectively. Fujimoto et al. (2009) indicated thatthe threads on the probe surface are important for producing plas-tic flow in the thickness direction during FSSW. Tozaki et al. (2007)performed FSSW using tools with three different probe lengths,suggesting that joint strength is proportional to probe length. Yuanet al. (2011) investigated the effect of distinct tool rotational speedsand directions on lap-shear separation load. The highest load wasachieved in the upper sheet fracture (USF) mode.

The studies mentioned in the above paragraphs showed that toolgeometry is a crucial characteristic for controlling the material flowevolution in FSSW. The effect of threading locations on cylindricalpins has not yet been discussed. Thus, tools with three threadinglocations on the cylindrical pins were applied to the execution ofFSSW, trying to alter the material flow characteristics such as SZheight and hook morphology. Later, tension-shear tests were car-ried out to evaluate the joint strength. Moreover, the evolution ofmaterial flow and hook formation during the plunging period wastraced step by step.

http://dx.doi.org/10.1016/j.jmatprotec.2015.06.0240924-0136/© 2015 Elsevier B.V. All rights reserved.

348 Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356

Table 1Chemical composition of A5052-H32 and A6061-T6.

Element(wt.%)

Si Fe Cu Mn Mg Cr Zn Ti Al

A5052-H32 0.10 0.27 0.02 0.06 2.38 0.19 0.01 0.01 Rem.A6061-T6 0.72 0.38 0.21 0.10 1.00 0.14 0.09 0.03 Rem.

Table 2Mechanical property of A5052-H32 and A6061-T6.

Modulus of elasticity,E (GPa)

Yield strength,�y (MPa)

Ultimate tensilestrength, �u (MPa)

Elongation,d (%)

A5052-H32 70.3 193 228 12A6061-T6 68.9 276 310 12

2. Experimental

Workpieces of Al–Mg (A5052-H32) and Al–Mg–Si (A6061-T6)aluminum sheets were chosen as weld specimens in this study.Tables 1 and 2list the chemical composition and mechanical prop-erties of A5052-H32 and A6061-T6 sheets, respectively. Prior toFSSW, the aluminum sheets were machined to dimensions of70(L) × 35(W) × 1.6(H) mm3. Ultrasonic cleaning with an acetonebath was conducted to remove contamination from the surface ofthe sheets.

FSSW tools were made of JIS SKD61 hardened steel with a hard-ness of HRc 50. The tool had a 10◦ concave shoulder 12 mm indiameter, and a cylindrical pin 5 mm in diameter and 2.8 mm inlength. On cylindrical pins, external M5 right-handed threads weremachined at distinct locations on three parts, as shown in Fig. 1.After the B-tool reaches a set plunging depth of 2.8 mm, the sinkand source of material are above the interface, whereas they arebelow the interface for the T-tool; for the M-tool, the sink is abovethe interface and the source is below.

Two FSSW experiments were conducted. The first experimentapplied overlapped dissimilar sheets of A5052-H32 (upper) andA6061-T6 (lower) to observe the material flow of welds. The weldswere cut perpendicular to the rolling direction of the sheets toobserve the cross-sectional material flow. The cut cross-sectionalspecimens were cold-mounted and then grounded by wet paperand polished with Al2O3 suspension. Keller’s reagent (380 mLH2O + 10 mL HNO3 + 6 mL HCl + 4 mL HF) was used to etch thespecimens to reveal the microstructure of the welds. The secondexperiment used overlapped A5052-H32 sheets to investigate force

and temperature variations during FSSW. The specimens were fixedwith a steel anvil and FSSW experiments were performed with aCNC machine. Force measurement was carried out using a four-axisKistler 9272 dynamometer associated with a four-channel Kistler5070 charge amplifier. Simultaneously, using an NI 9213 device fordata acquisition and LabVIEW software for data collection, the tem-perature variation was sensed and recorded at a sampling rate of100 Hz. Prior to temperature measurement, thermocouples werecovered with ceramic bars and inserted into specific locations toprevent contact between the sheets and thermocouples. K-typethermocouples as small as 0.5 mm in diameter were used to acquirerapid signal responses in a short processing time. The measur-ing locations in the welds are shown in Fig. 2. Following FSSW,tension-shear tests were performed using the MTS InsightTM test-ing apparatus at room temperature to evaluate the joint strengthof the welded specimens with a mm2 overlapped area. The pneu-matic cross-head speed was fixed at 1 mm/min. The joint strengthsof welds under various welding parameters were obtained byaveraging the results from at least three individual replicates forensuring experimental repeatability. Later, fractographic cross sec-tions were observed to examine the fracture characteristics andfracture modes.

3. Results

Figs. 3–5 show the cross sections of welds made with B-, M-,and T-tools at various pin plunging depths, individually. Tool dwelltime was set to zero to eliminate the material flow disturbanceduring dwell periods. The welds at the 0.6 mm and 1.2 mm plung-

Fig. 1. Geometry and screw threads location of tools.

Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356 349

Fig. 2. Schematic diagram of temperature measurement position (Unit: mm).

Fig. 3. Weld morphology by B-tool (Dwell time: 0 s).

ing depths show no difference by different tools. The material wasbackward extruded to form burrs on the upper sheet’s surface in the900 rpm welds with a 0.6 mm plunging depth. At 1.2 mm, the weldburr height was similar to that at 0.6 mm, and the interface gapbetween the sheets became larger. At 1.7 mm, the tool tip contin-ually extruded nugget material downward, and the sheet interfacecould still be distinguished. At 2.2 mm, the lower sheet extrudedinto the upper sheet to bond them together, and different toolscaused distinct degrees of bonding. As the plunging depth increasedfrom 2.2 mm to 2.8 mm, complete contact between the shoulderand the sheet surface occurred. In the vicinity of the nugget, theupper sheet material flowed downward from the upper side and

the lower sheet material was extruded upward to penetrate intothe upper sheet.

Note that the morphology of the 1800 rpm welds was dif-ferent from that of the 900 rpm welds after reaching a 1.7 mmplunging depth. In Figs. 3–5(b), the interfaces in the 1800 rpmwelds were eliminated by more frictional work input, showingpreliminary bonding behavior that occurred earlier than those inthe 900 rpm welds. Increased penetration depth made massiveextruded material flow upward with greater height. Tools withvarious threading locations exhibited different material flow mor-phologies at a 2.8 mm plunging depth. Lower sheet material wenthighest into the upper sheet in the B-tool weld, as shown in Fig. 3(b).

350 Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356

Fig. 4. Weld morphology by M-tool (Dwell time: 0 s).

Fig. 5. Weld morphology by T-tool (Dwell time: 0 s).

Fig. 6 shows the T-tool welds without tool rotation. Burrs werenot produced on the upper sheet surface. The extruded uppersheet material moved toward the sheet interface with increas-ing penetration depth. At 2.8 mm, the upper sheet material atthe pin tip was severely extruded and broken, indicating a failurejoint.

Fig. 7 shows the morphology of aluminum bulk welds with theB-tool at 1800 rpm. The bulk welds exhibited a morphology dis-tinct from that of the overlapped sheet welds. Here, the materialflow during the plunging process was similar to a piercing process.The friction-heat-affected zones were shown that expanded fromthe pin periphery as the plunging depth increased. An inverse trian-gular zone was generated while the shoulder completely contactedthe upper sheet.

Fig. 8 shows the variations in axial force (Fz) and torque (Tz) ofthe B-tool welds. The Fz and Tz variations exhibited three steps.

First, during plunging, as the pin contacted the sheet, Fz and Tzincreased rapidly and then achieved their peak values when thetool reached a set plunging depth of 2.8 mm. The maximum Fz wasnearly the same at 900 rpm and 1800 rpm. However, the maximumTz in these two welds differed significantly. These phenomena alsocan be recognized in the M- and T-tool welds (not shown here). Sec-ond, during dwelling, Fz and Tz decreased gradually and exhibitedsimilar variation. Both Fz and Tz seemed to approach stable statesand thereby become smoother. Third, Fz and Tz reverted sharply tozero after retracting the tool from the welds.

Figs. 9(a) and (b) shows the temperature variation of themeasured points in the B-tool welds at 900 rpm and 1800 rpm,respectively. During plunging, the temperature increased rapidly.The 1800 rpm weld exhibited a faster temperature increase thanthat the 900 rpm weld. Subsequently, temperature increased rela-tively slowly during the dwell period, and the temperature gradient

Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356 351

Fig. 6. Weld morphology by T-tool without tool rotation.

of the 1800 rpm weld showed less significant variation. The peaktemperature at each measured points occurred before tool retrac-tion.

Fig. 10 shows the tension-shear strengths of the six set of welds,wherein each set comprised of at least three welds that were pro-duced by the three tools. On average, the welds with 900 rpmexhibited joint strengths of over 6.5 kN, which were higher thanthose with 1800 rpm. The joint strength increased with increasingdwell time and seemed to approach saturated values. The effect oftool threading locations on the joint strength becomes weaker withincreasing dwell time.

According to the fractograph of the tension-shear test, two frac-ture modes with distinct welding parameters were observed in thetension-shear specimens. Fig. 11(a) shows the fracture cross sec-tion that necking behavior happened in the upper sheet. The hookmorphology at the periphery of the nugget went up slightly andthen had a turndown. The hook was nearly horizontal and dis-appeared in the vicinity of the keyhole. Therefore, the thinningbehavior of the upper sheet was not found and excellent bondingwas achieved in this region. Fig. 11(b) shows the severe bend-ing phenomena that the upper and lower sheets bent upward anddownward, respectively. Fig. 11(c) exhibits a weld nugget stuck onthe lower sheet surface, indicating an USF mode and superior joint

Fig. 7. Bulk weld morphology by B-tool at 1800 rpm (Dwell time: 0 s).

strength. Fig. 11(d) indicates the characteristic of equiaxed dimpleson the fracture surface. Fig. 12(a) shows a typical interfacial fail-ure characteristic. At the right compressive side, the fracture pathwent through the SZ, which was directly shear-fractured. Figs. 12(b)and (c) shows flat fracture surfaces, which are dissimilar to thosein Figs. 11(b) and (c). In Fig. 12(d), the dimples are deformed andtheir size is smaller than those in Fig. 11(d). Moreover, the dim-ple morphology suggests a combination of shear and tensile failuremodes.

4. Discussion

Fig. 13 shows the effect of rotational speeds on fluctuation ofthe plunging depth without tool dwell. Material deformation in theplunging period involves elasticity and plasticity. The elastic recov-ery of the material resulted in the difference between the set andexperimental plunging depths. The 1800 rpm welds exhibited lessdifference in plunging depth than the 900 rpm welds. The frictionalheat produced in the 1800 rpm welds was higher, thereby leadingto more significant material softening and less elastic recovery fol-lowing tool retraction. Furthermore, the threading locations on thecylindrical pins caused similar recovery trends in the welds.

As shown in Fig. 3, the material flow at the pin peripherychanged from back extrusion to forward extrusion as the plung-ing depth increased to 1.2 mm, which is beyond the half-thicknessof the upper sheet. The material easily flowed toward the sheetinterface because of the pressure gradient. However, as shown in

352 Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356

Fig. 8. Force and torque variation during FSSW (B-tool weld, rotational speed: 900 rpm and 1800 rpm, dwell time: 9 s).

Fig. 9. Temperature variation during FSSW (B-tool weld, dwell time: 9 s).

Fig. 10. Tension-shear strength of welds under various parameters (Plunging depth:2.8 mm).

Fig. 7, the lack of a sheet interface in the bulk welds led to the backextrusion and an accompanying viscous flow of the heated mate-rial. The material flow in the bulk welds could not be observedobviously. At a depth of 1.7 mm, the higher frictional heat, as com-pared with the 900 rpm weld, decreased the flow stress, therebygenerating a material intermixture of the sheets in the 1800 rpmwelds. However, such an intermixture was not generated in the900 rpm welds. Subsequently, at a depth of 2.8 mm, the materialflow became fully developed when the shoulder contacted theupper sheet (Lin et al., 2013). External frictional heat due to shoul-der rubbing resulted in inverse triangular etching markers on thecross section. The extrusion effect and material softening in theupper sheet led to its penetration by the lower sheet material. Notethat the threading locations on the cylindrical pins affected thedegree of material intermixture. The B-tool threads were nearer tothe shoulder than the M- or T-tool threads. Thus, the upper sheetmaterial was transferred downward immediately by the threads inthe 1800 rpm B-tool weld. As shown in Fig. 3(b), sufficient mate-

Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356 353

Fig. 11. Tension-shear testing specimen. The arrows indicate the loading direction (B-tool weld, rotational speed: 900 rpm, dwell time: 9 s).

rial softening and appropriate threading locations caused a greateramount of upper sheet material, as compared with the 900 rpmweld, to flow downward, thereby resulting in a more obvious mate-rial intermixture.

Note that the welds in Figs. 3–5 reveal different etching marksin the vicinity of keyholes. The etching marks showed the locusof plastic flow and the frictional heat zone. Prior to complete con-tact between a shoulder and an upper sheet, the 900 rpm weldsexhibited obvious etching marks in the bottom of the keyholes, inwhich material work hardening was more severe. In contrast, the1800 rpm welds exhibited uniform etching marks, suggesting thetool pin produced significant frictional heat to cause recrystalliza-tion.

The joining mechanisms of friction-based welding comprisecold-pressure welding, diffusion welding, and mechanical alloy-

ing at the same time (Groche et al., 2014). Frictional heat is a keyparameter for obtaining successful FSSW joints. Inspection of theexisting sheet interfaces shown in Fig. 6 suggested the presenceof failure joints because of negligible frictional heat generated by anon-rotational tool. This behavior is similar to an extrusion process.Thus, it seems that the generation of frictional heat by tool rotationis necessary for obtaining a material intermixture and successfuljoints.

Fig. 14 shows the weld morphologies with 1800 rpm and a 9 sdwell time produced by individual threading tools. The SZ morphol-ogy was similar, but it shifted with different threading locations ontools, revealing an effective method for controlling material flowby varying the threading locations on the cylindrical pins.

In Fig. 8, the maximum Tz reached 32.4 N m and 16.5 N m in the900 rpm and 1800 rpm welds, respectively. The maximum torque

354 Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356

Fig. 12. Tension-shear testing specimen. The arrows indicate the loading direction (M-tool weld, rotational speed: 1800 rpm, dwell time: 1.5 s).

in the 900 rpm weld was nearly twice as that in the 1800 rpm weld.As the tool rotates at higher rate, more heat energy is transferredto the workpiece (Mishra et al., 2014). Thus the higher rotationalspeed, i.e., 1800 rpm, generated more frictional heat to cause moresignificant material softening. It is also found that the peak temper-atures in the 1800 rpm welds at the individual measuring pointswere higher than those in the 900 rpm welds (Table 3). Thus, lesstorque was needed to cause material flow during FSSW. In contrast,the maximum Fz was identical. Bilici and Yükler (2012) reportedthat the maximum Fz varied with tool geometry. Cox et al. (2011)

found that Fz was largely dependent on the plunging depth. Theused tools in this study comprised a cylindrical pin with a con-cave shoulder, and the plunging depth was fixed at 2.8 mm. Hence,the threading locations on pins seemed to have little influence onFz variation. In Fig. 8, the Fz variation during the plunging periodshows two substeps resulting from the tool geometry is shown inFig. 8. The first substep was the plunge of the cylindrical pin intothe sheets, leading to a rapid force increase. The second substepwas the contact between the shoulder and upper sheet, resultingin the maximum Fz. During the dwell period, tool friction continu-

Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356 355

Fig. 13. Difference between expected and actual penetrating depth (Dwell time: 0 s).

Fig. 14. Weld morphology using B-tool, M-tool, and T-tool (Rotational speed: 1800 rpm, dwell time: 9 s).

Table 3Peak temperature of different positions (Dwell time: 9 s) (A: 1.2 mm, B: 2.0 mm, C:2.7 mm from upper sheet surface).

Measured point 900 rpm 1800 rpm

B-tool A 464.3 487.3B 444.1 465.3C 442.7 464. 1

M-tool A 470.8 495.5B 444.9 469.5C 451.6 454.6

T-tool A 458.6 487.1B 426.1 469.7C 427.2 475.4

ously produced heat to soften the material. As dwell time increased,the gradients of Fz and Tz became smoother and approached satu-rated values. Simultaneously, the temperature gradient in Fig. 9 alsodecreased because of the gradual decrease in torque, suggesting adecrease in material plasticization.

The shoulder rubbing against the sheets was the primary con-tributor to external frictional heat during FSSW. As indicated byTang et al. (1998), the rotational shoulder generates most heat inFSW. Zhang et al. (2014) also indicated that the shoulder-plate con-tact surface is the main factor of heat generation. Thus, in Table 3,point A, which was nearest measuring point to the shoulder, showsthe highest peak temperature. In addition, Fig. 9 shows that the

356 Y.-C. Lin, J.-N. Chen / Journal of Materials Processing Technology 225 (2015) 347–356

peak temperature was achieved by the accumulation of frictionalheat before the tools were retracted from welds.

In Fig. 11(a), the flat hook morphology is similar to the resultfrom Zhang et al. (2011). In the 900 rpm welds, smaller frictionalheat induced by the shoulder resulted in higher material flow resis-tance. Thus, the material flowing from the lower sheet toward theupper sheet was less significant, suggesting weak material inter-mixture. Although the 900 rpm welds exhibited an inferior materialintermixture effect, the joint strength was superior to that of the1800 rpm welds (Fig. 10). The maximum joint strengths of over6.5 kN were much higher than those A5052-H112 welds joined withthe threaded tool (Zhang et al., 2011). In addition, it seems that rota-tional speed had more significant influence on joint strength thandwell time, which was different from the result reported by Tutaret al. (2014). According to the failure criterion, a weld fails from theweakest location. Tool plunging and stirring led to material forgingand dispersion of surface oxides from the interface, respectively,forming strong and complete bonding in the vicinity of the keyhole.Complete bonding was favorable for enhancing the joint strengthin the 900 rpm welds. When suffering from external force duringtension-shear tests, Fig. 11(a) shows that the crack was initiatedat the notch tip of interface and propagated along the peripheryof the nugget, following fracturing at the circumference of theshoulder indentation. Similarly, in Fig. 12(a), the crack was alsoinitiated at the notch tip of the interface. However, a lower dwelltime of 1.5 s caused inferior bonding in the vicinity of the keyhole.Hence, the crack propagated along the hook and interfacial fractureoccurred.

Referring to the SEM images, the ductile fracture behaviorwas the dominant mechanism in the welds. Ductile fracture wasderived from microvoid nucleation, growth and coalescence. Theequiaxed dimples on the fracture surface indicate the ductilecharacteristic and suffered from tension in this region result-ing from the bending effect of the upper sheet in the breakingstage (Fig. 11(d)). In contrast, the deformed dimples suffered fromshear force, inducing high local stress and plastic deformationin the local area (Fig. 12(d)). Moreover, the dimple size sug-gests the difference in ductility. Fig. 11(d) shows a larger dimplesize, indicating superior ductility and causing a severe bendingphenomenon.

5. Conclusions

This study applied various threading locations on cylindricalpins to execute FSSW. According to the experimental results it canbe concluded as follows:

1. The SZ height in the welds shifted with threading locations oncylindrical pins, thereby demonstrating an effective method forcontrolling material flow.

2. During a plunging period, the welds created by the variousthreading tools resulted in various degrees of material intermix-ture. The FSSW joint created by the B-tool at 1800 rpm revealedan earlier intermixing phenomenon as compared with the jointcreated by the M- and T-tool at the same rotational speed.

3. Tool rotation was beneficial for developing material flow by gen-erating frictional heat in the welds. The lack of tool stirring ledto failure FSSW joints.

4. The welding torque varied with tool rotational speed. The high-est torque occurred at the lowest rotational speed of 900 rpmbecause of minor material softening in welds. Not much differ-ence was found in the axial force variation.

5. During FSSW, the shoulder contributed the major externalfrictional heat, leading the peak temperature to occur at themeasured point nearest to the shoulder.

References

Babu, S., Sankar, V.S., Janaki Ram, G.D., Venkitakrishnan, P.V., Madhusudhan Reddy,G., Prasad Rao, K., 2013. Microstructures and mechanical properties of frictionstir spot welded aluminum alloy AA2014. J. Mater. Eng. Perform. 22, 71–84.

Badarinarayan, H., Shi, Y., Li, X., Okamoto, K., 2009a. Effect of tool geometry onhook formation and static strength of friction stir spot welded aluminum5754-O sheets. Int. J. Machine Tools Manuf. 49, 814–823.

Badarinarayan, H., Yang, Q., Zhu, S., 2009b. Effect of tool geometry on staticstrength of friction stir spot-welded aluminum alloy. Int. J. Machine ToolsManuf. 49, 142–148.

Bilici, M.K., Yükler, A.I., 2012. Influence of tool geometry and process parameterson macrostructure and static strength in friction stir spot welded polyethylenesheets. Mater. Des. 33, 145–152.

Cox, C.D., Gibson, B.T., Strauss, A.M., Cook, G.E., 2011. Effect of pin length androtation rate on the tensile strength of a friction stir spot-welded al alloy: acontribution to automated production. Mater. Manuf. Process. 27, 472–478.

Florea, R.S., Solanki, K.N., Bammann, D.J., Baird, J.C., Jordon, J.B., Castanier, M.P.,2012. Resistance spot welding of 6061-T6 aluminum: failure loads anddeformation. Mater. Des. 34, 624–630.

Fujimoto, M., Koga, S., Abe, N., Sato, S.Y., Kokawa, H., 2008. Analysis of plastic flowof the al alloy joint produced by friction stir spot welding. Yosetsu GakkaiRonbunshu/Quarterly J. Japan Weld. Soc. 26, 67–73.

Fujimoto, M., Koga, S., Abe, N., Sato, S.Y., Kokawa, H., 2009. Analysis of plastic flowof the Al alloy joint produced by friction stir spot welding. Weld. Int. 23,589–596.

Groche, P., Wohletz, S., Brenneis, M., Pabst, C., Resch, F., 2014. Joining by forming–areview on joint mechanisms, applications and future trends. J. Mater. Process.Technol. 214, 1972–1994.

Lin, Y.C., Liu, J.J., Chen, J.N., 2013. Material flow tracking for various tool geometriesduring the friction stir spot welding process. J. Mater. Eng. Perform. 22,3674–3683.

Mishra, R.S., De, P.S., Kumar, N., 2014. Friction Stir Welding and Processing.Springer.

Su, P., Gerlich, A., North, T.H., Bendzsak, G.J., 2007. Intermixing in dissimilar frictionstir spot welds. Metall. Mat. Trans. A 38, 584–595.

Tang, W., Guo, X., McClure, J.C., Murr, L.E., Nunes, A., 1998. Heat input andtemperature distribution in friction stir welding. J. Mater. Process. Manuf. Sci.7, 163–172.

Tozaki, Y., Uematsu, Y., Tokaji, K., 2007. Effect of tool geometry on microstructureand static strength in friction stir spot welded aluminium alloys. Int. J. MachineTools Manuf. 47, 2230–2236.

Tutar, M., Aydin, H., Yuce, C., Yavuz, N., Bayram, A., 2014. The optimisation ofprocess parameters for friction stir spot-welded AA3003-H12 aluminium alloyusing a Taguchi orthogonal array. Mater. Des. 63, 789–797.

Yuan, W., Mishra, R.S., Webb, S., Chen, Y.L., Carlson, B., Herling, D.R., Grant, G.J.,2011. Effect of tool design and process parameters on properties of Al alloy6016 friction stir spot welds. J. Mater. Process. Technol. 211, 972–977.

Zhang, Z., Yang, X., Zhang, J., Zhou, G., Xu, X., Zou, B., 2011. Effect of weldingparameters on microstructure and mechanical properties of friction stir spotwelded 5052 aluminum alloy. Mater. Des. 32, 4461–4470.

Zhang, Z., Wu, Q., Zhang, H.W., 2014. Numerical studies of effect of tool sizes andpin shapes on friction stir welding of AA2024-T3 alloy. Trans. Nonferrous Met.Soc. China 24, 3293–3301.

Zhao, Y.Q., Liu, H.J., Chen, S.X., Lin, Z., Hou, J.C., 2014. Effects of sleeve plunge depthon microstructures and mechanical properties of friction spot welded alclad7B04-T74 aluminum alloy. Mater. Des. 62, 40–46.