Microstructure and Mechanical Properties

of Friction Stir Spot Welded Galvanized Steel

Seung-Wook Baek1;*1, Don-Hyun Choi1, Chang-Yong Lee2, Byung-Wook Ahn1,Yun-Mo Yeon3, Keun Song3 and Seung-Boo Jung1;*2

1School of Advanced Materials Science and Engineering, Sungkyunkwan University,300 Cheoncheon-dong, Jangan-gu, Gyeonggi-do 440-746, Korea2Automotive Applications Research Team, Technical Research Center, Hyundai Steel,167-32 Kodae-Ri, Songak-Myon, Dangjin, Chungchungnam-do 343-823, Korea3Department of Advanced Materials Application, Suwon Science College 9-10,Botong-li, Jeongnam-myeon, Hwasung, Gyeonggi-do 445-742, Korea

Joints of galvanized steel were obtained by friction stir spot welding (FSSW) with lap configuration using CPS design tool. Nomechanically mixed layer was formed between the top and bottom plates at the weld nugget due to the limited tool penetration and the lower pinheight of the welding tool than the steel plate thickness. A deformed region, in which ZnO particles were detected, was observed in the joint. Theformation of this deformed region was attributed to the explosion of the Zn coating layer due to friction heating and tool compression. Withincreasing tool penetration depth, the tensile shear strength of the joint increased to a maximum value of 3.07 kN at a tool penetration depth of0.52mm. [doi:10.2320/matertrans.M2009337]

(Received October 5, 2009; Accepted March 4, 2010; Published April 25, 2010)

Keywords: friction stir spot welding, galvanized steel, WC-Co alloy tool, explosion, tensile shear strength

1. Introduction

Friction stir spot welding (FSSW) is a new process that hasrecently received considerable attention from the automotiveand other industries.1) A novel variant of the friction stirwelding (FSW) process, FSSW creates a spot, lap-jointwithout material melting. The appearance of the resultingjoint resembles that of an electric resistance spot weldcommonly used for auto-body assembly. The solid-statebonding and other features of the process make it inherentlyattractive for body assembly and other similar applications.The primary welding process for auto body structureassembly—the electric resistance spot welding (ERSW)process—can be problematic for many new, high perform-ance, structural materials such as Al and Mg alloys, advancedhigh-strength steels.2,3)

So far, the majority of the research and developmentefforts on FSW have been on aluminum alloys.4–12) BecauseAl alloys are easy to deform at relatively low temperatures(below about 550�C), they undergo FSW easily.

Nevertheless, steel remains the primary material for bodystructures of all high-volume, mass-produced cars. Thestrong, customer-driven emphasis on safety and corrosionprotection has been driving the increased use of galvanizedsteel in automobile body construction. However, the weldingof galvanized steel presents some unique technical challeng-es to both the steel suppliers and the auto end-users due to themarked differences between welding galvanized steels andwelding the bar or uncoated versions of the same steelproducts. The different physical and electrical properties ofthe coating, compare to those of low-carbon steel, lead toits quite different weld formation and welding behavior.

The plating of the coating onto the copper alloy electrodes,in addition to its alloying with these electrodes, greatlyaccelerates the electrode wear.13–16)

Recently, Research about FSSW of steel alloys is in-creased and shown successful result including low carbonsteel,17,18) DP600,19) DP80020) and M-190,21) an effort wascommenced to evaluate this joining process for moreproblematic steel alloys. However, these steel alloys haduncoated surface and FSSW of galvanized steel has notbeen examined. Also, in general, a tool with a threaded pin isused for severe plastic deformation in FSSW. In a previousreport,22) Y. Tozaki et al. indicated that the length ofthreaded probe strongly affected on the strength of FSSWjoints. However, there have been reports of the threaded pintool causing a decrease in the joint’s volume and upperplate’s thickness, owing to the deep tool penetration.23,24)

Moreover, using the conventional tool, pin hole inevitablyremains at the centre of the weld nugget. It is believedthat corrosion could take place preferentially at the pin holebecause rainwater remains in the hole, where body paintbarely reaches the bottom. And Uematsu et al. reported thatshear strength and fatigue strength of joint were improvedusing the re-filling process due to wide joint area and pin holewas not remained after welding.25) Considering these results,it seems that reliability of the joint will be increased byreducing of the pin hole depth and enlarge the joint area.

The present study applies FSSW to galvanized steel plateusing WC-Co alloy tool as CPS (Cylindrical Pin withShoulder) design with the following two goals. The first is toinvestigate the influence of one parameter—tool penetrationdepth—on the mechanical shear strength and failure modeof the welded joints. We chose to vary tool penetrationdepth because it has been previously suggested that thishas an important effect on joint properties. The variationof the tool penetration depth also results in the variation

*1Graduate Student, Sungkyunkwan University*2Corresponding author, E-mail: sbjung@skku.ac.kr

Materials Transactions, Vol. 51, No. 5 (2010) pp. 1044 to 1050#2010 The Japan Institute of Metals EXPRESS REGULAR ARTICLE

of the depth with which the rotating tool shoulder pressesonto the top sheet, thereby affecting that region’s micro-structure. The second goal of this work is to analyze indetail the microstructure of a representative joint (0.52mminsertion depth), in order to explain the mechanisms of thisnew joining process on a more fundamental and generallevel.

2. Experimental Procedure

The base material used for welding in this study was a0.6mm-thick steel plate with 0.025mm-thick zinc coat. Thechemical composition was Fe-0.002C-0.005Si-0.127Mn-0.0099P-0.008S (in mass%), with an ultimate tensile strengthof 314MPa and elongation of 43%. The shear joints weremade in a lap configuration, with each plate being 30mmwide by 100mm long. Two sheets were overlapped by30mm, and then affixed into the welding instrument usinga fixture in preparation for joining. The WC-Co alloy CPStool was shown in Fig. 1. Tool had a shoulder diameter of13.5mm and pin diameter of 8.5mm and height of the pinwas 0.5mm. FSSW was performed at a tool rotation speedof 1600 rpm and over a range of tool penetration depth of0:24�0:52mm. Temperature profile during FSSW wasmeasured using the thermo graphic camera. Optical mi-croscopy (OM), Hitachi S-3000H scanning electron mi-croscopy (SEM) equipped with energy dispersive spectrom-etry (EDS) and JEOL JXA-8500F electron probe microanalyzer (EPMA) observations were carried out for joint. Themetallurgical inspections were performed on a cross-sectionof the joint after polishing and etching with a nital solution.

Detailed microstructural observations were performed byJEOL 300 kV transmission electron microscopy (TEM). TheTEM samples were prepared by electrolytically polishingwith a 10% nitric acid + 90% ethanol solution. Temperatureprofile during FSSW was measured using the thermo graphiccamera. The mechanical properties of the spot welds werecharacterized using tensile shear test. Tensile shear test wasconducted at room temperature at a crosshead speed of1mm�min�1.

3. Result and Discussion

Macroscopic overviews of the cross-section of each jointare shown in Fig. 2. In all samples, the top and bottomsteel sheets were compressed together to form a jointinterface. The gap at the joint edge region was decreasedwith increasing tool penetration depth, suggesting that the

Fig. 1 WC-Co alloy tool with CPS design used in this study.

Fig. 2 Optical macro cross-sectional images at the following tool penetration depths: (a) 0.24mm, (b) 0.30mm, (c) 0.48mm, and

(d) 0.52mm.

Microstructure and Mechanical Properties of Friction Stir Spot Welded Galvanized Steel 1045

decreasing gap at the joint edge region affected the jointstrength and fracture location. A white layer, with a shapeof bilateral symmetry, was observed at the top sheet in allsamples. It was assumed that these layers were relatedwith the phase transformation of the steel sheets or werea result of the reaction between the tool and steel sheets.A line was shown across the middle region of the joint, andwas inferred to be the interface between the top and bottomsteel sheets.

Figure 3 shows the representative grain distribution in thebase metal (a), side region under the pin (b), and centerregion under the pin (c)–(d). The material analyzed in Fig. 3was the 0.52mm-tool penetration depth sample. In the case ofthe base metal, microstructure was comprised of elongated

ferrite phase grains, with an average grain size of 25 mm. Inthe side region under the pin, fine and discolored grains wereobserved, which were attributed to the effect of the zinc-coating layer on this region. In the center region under pin,the grains were larger than those of the base metal, weassumed that recrystalization occurred in this region dueto stirring, and grain growth occurred due to the frictionheating induced during the welding.17) In the middleregion (Fig. 3(d)), a fully metallurgical, bonded region wasobserved, while very fine microstructures were revealedbetween the top and bottom sheets (Fig. 4). In this study, asthe CPS tool did not penetrate the bottom sheet, the twosheets were not bonded by mechanical stirring, in contrastwith conventional tool. However diffusion was occurred

Fig. 3 Optical microstructure of the FSSW joint: (a) base metal, (b) discolored region, (c) top region, (d) middle region.

Fig. 4 Microstructure of the middle region in the FSSW joint: (a) OM micrograph and (b) high magnification SEM micrograph.

1046 S.-W. Baek et al.

between the top and bottom sheets by friction heating andtool compression, they were bonded like diffusion bonding.Moreover this diffusion between top and bottom sheet wasoccurred during short time, it seems that interface betweentop and bottom sheet was disappeared, but grain growth wasnot fully occurred due to rapid cooling.

Figure 5 shows a SEM and EDS analysis results of themicrostructure of the joint. An interface was clearly evidentbetween the top and bottom sheets, and defects were

observed in the bottom sheet. With its trimmed shape, thetop sheet presented a different shape from the bottom sheet.Figure 5(b) shows a high magnification SEM micrographof the trimmed shape region. In the EDS results, Zn wasdetected in this region but was not detected in the otherregions. Also EPMA result was shown that Zn was existedin trimmed shape region (Fig. 6). Considering these result,it seems that Zn was assumed to have come from the Zn-coated layer.

mass% at% mass% at%

Fig. 5 SEM micrographs of the discolored region in the FSSW joints: (a) low magnification, (b) high magnification.

Fig. 6 Result of EPMA analysis in trimmed region.

Microstructure and Mechanical Properties of Friction Stir Spot Welded Galvanized Steel 1047

To analyze the Zn behavior in the joint, the jointmicrostructure was analyzed with TEM. Figure 7 showsTEM micrographs of the Zn-affected and unaffected regions.About 20 nm-sized black particles were dispersed in themicrostructure in the Zn-affected zone (Fig. 7(a)), but werenot observed in the unaffected zone (Fig. 7(b)). In the EDSanalysis results, Zn and O were detected in this region,suggesting that these black particles were ZnO.

Maximum peak temperature during FSSW was shown inFig. 8. It shows that maximum peak temperature betweentool and workipiece was 1293K. Considering this result, Fewas not melted during FSSW, but it suggested that thewelding temperature exceeded the melting and boiling pointof Zn during the welding, it was known that the melting andthe boiling points of Zn are 693 and 1180K, respectively.These result supported the following explanation for the Znbehavior during the welding. The friction heat generatedbetween the tool and workpiece during the welding. This

result indicated that the Zn was melted and evaporated due tothe high temperature. However, we assumed that evaporatedZn was not out of joint, because the outflow of any gas wasblocked by the penetrated tool. This evaporated Zn thatcould not escape was exploded by the friction heat and toolcompression, and this explosion easily deformed the joint,which was already softened by the friction heating (Fig. 5).As the joint was cooled after welding, the evaporated Znturned to liquid and finally solidified in the joint. However,the center region in the joint was not affected by Zn,suggesting that the evaporated and melted Zn was spreadfrom the center region to the side region due to the centrifugalforce of the tool rotation.

Figure 9 shows the effect of tool penetration depth on theshear strength of the joints. The error bars in the figurerepresent the standard deviation of the test data. The shearstrength was maximized at a penetration depth of 0.52mm.The shear strength peaks at the 0.24mm penetration depth

mass% at% mass% at%

Fig. 7 TEM micrographs of the FSSW joint: (a) Zn-affected zone, (b) unaffected zones.

Fig. 8 Result of measured maximum peak temperature during FSSW.

1048 S.-W. Baek et al.

and goes up for the 0.30mm insertion depth. However overthe 0.30mm insertion depth, shear strength goes slightly goesup for the 0.52mm insertion depth.

Figure 10 shows photographs of the failure surfaces andcross section of the lap shear specimens joined at fourdifferent tool penetration depths. The figures also show theexact shear strength of the samples. While the difference inthe lap-shear strength between these specimens is relativelysmall, there is a marked change in the failure mode withincreasing tool penetration depth. In the 0.24mm-toolpenetration depth sample, fracture occurred at the interfacebetween the top and bottom surfaces. However, as the toolpenetration depth increased beyond 0.30mm, the failurelocation shifted further away from the base metal. Failure

occurred in the top sheet, midway through the section whichwas under the tool shoulder. The bonded area of the joint wasthe same at both 0.24mm- and 0.30mm-tool penetrationdepths. However, as the tool penetration depth exceeded0.30mm, the bonding area was slightly increased to amaximum bonding area of 176.15mm2 at the 0.52mm-toolpenetration depth. These results suggested that the increase ofthe lap-shear strength between these samples is related withthe bonded area under the tool shoulder region.

4. Conclusion

FSSW of galvanized steel was carried out as a function oftool penetration depth with a welding tool made fromWC-Coalloy. The study results are summarized as follows:(1) Sound FSSW joints were obtained without any meltingof the workpiece and the mechanically mixed zone, underthe all welding conditions.(2) The deformation region that was observed in the jointand ZnO was detected in this region. This result implied thatdeformation region was related with behavior Zn duringFSSW.(3) Maximum temperature during FSSW was about 1293K,this temperature was lower than melting point of Fe buthigher than boiling point of Zn. It suggested that Zn wasevaporated and exploded due to high temperature and toolcompression and this explosion affected on joint.(4) With increasing tool penetration depth, the tensile shearfracture load gradually increased to a maximum of 3.07 kNat the deepest tool penetration depth of 0.52mm.

Fig. 9 Tensile shear test results.

Fig. 10 Fracture images after tensile shear testing at the following tool penetration depths: (a) 0.24mm, (b) 0.30mm, (c) 0.48mm, and

(d) 0.52mm.

Microstructure and Mechanical Properties of Friction Stir Spot Welded Galvanized Steel 1049

Acknowledgment

This work was supported by the WCU program ofKorea Science & Engineering Foundation funded by theKorean government (MOEHRD) (Grant No. R32-2008-000-10124-0).

REFERENCES

1) R. Hancock: Weld. J. 83 (2004) 40.

2) H. Oikawa, G. Murayama, T. Sakiyama, Y. Takahashi and T. Ishikawa:

NIPPON STEEL TECHNICAL REPORT No. 95 January (2007).

3) D. J. Spinella, J. R. Brockenbrough and J. M. Fridy: Weld. J. 84 (2005)

34–40.

4) C. G. Rhodes, W. M. Mahoney and W. H. Bingel: Scr. Metall. 30

(1997) 69–75.

5) K. E. Knipstrom and B. Pekkari: Weld. J. 76 (1997) 55–57.

6) W. M. Mahoney, C. G. Rodes, J. G. Filntoff, R. A. Spurling and W. H.

Bingell: Metall. Mater. Trans. A 20A (1998) 1955–1964.

7) L. E. Murr, G. Liu and J. C. McClure: J. Mater. Sci. 33 (1998) 1243–

1251.

8) L. E. Murr, G. Liu and J. C. McLure: J. Mater. Sci. Lett. 16 (1997)

1801–1803.

9) K. V. Jata, K. K. Sankaran and J. J. Rushau: Metall. Mater. Trans. A

31A (2000) 2181–2192.

10) J. C. Dawes and W. M. Thomas: Weld. J. 75 (1996) 41–45.

11) R. W. Fonda and J. F. Bingert: Metall. Mater. Trans. A 35A (2004)

1487–1499.

12) Y. S. Sato, H. Kokawa, K. Ikeda, M. Enomoto, S. Jorgan and T.

Hashimoto: Metall. Mater. Trans. A 32A (2001) 941–948.

13) Y. Hovanski, M. L. Santellab and G. J. Grant: Scr. Mater. 57 (2007)

873–876.

14) P. S. James, H. W. Chandler and J. T. Evans: Mat. Sci. Eng. A 230

(1997) 194–201.

15) I. H. Kwon, H. H. Kim, S. S. Baek, S. M. Yang and H. S. Yu: Trans.

KSAE. 13 (2005) 42–49.

16) J. S. Lee and E. Y. Chin: J. Kor. Inst. Met. Mater. 34 (1996) 261–

270.

17) S. W. Baek, D. H. Choi, C. Y. Lee, B. W. Ahn, Y. M. Yeon, K. Song

and S. B. Jung: Mater. Trans. 51 (2010) 399–403.

18) D. H. Choi, C. Y. Lee, B. W. Ahn, J. H. Choi, Y. M. Yeon, K. Song,

H. S. Park, Y. J. Kim, C. D. Yoo and S. B. Jung: Int. J. Ref. Met. Hard

Mater. 27 (2009) 931–936.

19) M. I. Khan, M. L. Kuntz, P. Su, A. Gerlich, T. North and Y. Zhou: Sci.

Technol. Weld. Join. 12 (2007) 175–182.

20) Z. Feng, M. L. Santella, S. A. David, R. J. Steel, S. M. Packer, T. Pan,

M. Kuo and R. S. Bhatnagar: SAE 2005 Trans. J. Mat. Manuf. 114

(2005) 592–598.

21) M. P. Miles, J. Sederstron, K. Kohkonen, R. Steel, S. Packer, T. Pan,

W. J. Schwartz and C. Jiang: MS&T 2 (2006).

22) H. Badarinarayan, Q. Yang and S. Zhu: Int. J. Mach. Tool. Manu. 49

(2009) 142–148.

23) D. Mitlin, V. Radmilovic, T. Pan, J. Chen, Z. Feng and M. L. Santella:

Mater. Sci. Eng. A 441 (2006) 79–96.

24) C. Y. Lee, W. B. Lee, Y. M. Yeon, K. Song, J. H. Moon, J. G. Kim and

S. B. Jung: Adv. Mater. Res. 15–17 (2007) 345–350.

25) Y. Uematsu, K. Tokaji, Y. Tozaki, T. Kurita and S. Murata: Int. J.

Fatigue 30 (2008) 1956–1966.

1050 S.-W. Baek et al.