Lap Joint of A5083 Aluminum Alloy and SS400 Steel by Friction Stir Welding

Materials Transactions, Vol. 46, No. 4 (2005) pp. 835 to 841 #2005 The Japan Institute of Metals Lap Joint of A5083 Aluminum Alloy and SS400 Steel by Friction Stir Welding Kittipong Kimapong* and Takehiko Watanabe Department of Mechanical Engineering, Niigata University, Niigata 950-2181, Japan A lap joint of A5083 aluminum alloy and SS400 steel was produced by Friction Stir Welding (FSW) using the various process parameters such as rotational speed, traverse speed and pin depth. The variations of welding parameters produced various characteristic interfaces and had conspicuous influences on the joint properties. Increasing the rotational speed decreased the shear load of the joint due to producing a thick FeAl 3 intermetallic compound (IMC) at the interface. When traverse speed increased, the shear load increased because IMC thickness at the interface decreased, however, when the speed was so high, an incomplete interface was formed. Increasing the pin depth produced a thick FeAl 3 IMC phase and an incomplete interface that directly deteriorated the shear load of the joint. (Received October 6, 2004; Accepted February 18, 2005; Published April 15, 2005) Keywords: lap joint, aluminum alloy, steel, friction stir welding, tensile shear strength 1. Introduction Reducing the weight of vehicles is an efficient measure for energy saving in the automobile industry. 1) Lightweight metals such as aluminum have widely replaced steel parts in automobile structures. Necessity of a joint between aluminum and steel has increased. However, joining aluminum to steel by conventional fusion welding is difficult because hard and brittle phases are inevitably formed at the weld. It is known that friction stir welding (FSW) is a solid state welding process that can produce a butt joint between aluminum alloys 2 5) and dissimilar materials. 6 8) Recently, a few preliminary studies have reported on FSW of aluminum and steel. Watanabe et al. 9) reported a butt joint of A5083 aluminum alloy to SS400 steel was successfully produced by FSW and indicated the maximum tensile strength was 86% that of the aluminum base material. The IMC was not formed at the interface of the joint, however, at the upper region of the weld near the shoulder of the rotating tool, the FeAl 3 IMC was slightly formed and tended to decrease the tensile strength of the joint. Chen and Kovacevic 10) welded Al 6061 alloy to AISI 1018 steel and showed the Fe 4 Al 13 and Fe 2 Al 5 IMC were formed in the nugget zone but the report did not explain the relation between the IMC and the strength of the joint. Yoshikawa and Hirano 11) studied on a numerically controlled FSW in layered dissimilar metal of A1050-0 aluminum alloy to SPCC cold rolled steel and showed the joint could be produced. The compressive shear strength at the welded area was found to be equal to the shear strength of aluminum base materials. However, the detail of microstructural investigation was not discussed. Elrefaey et al. 12) also reported the application of FSW to a lap joint of A1100H24 aluminum alloy to SPCC low carbon steel and reported the detail of microstructure of the joint. Interface structure observation showed FeAl and FeAl 3 IMC formed at the interface and indicated the brittle morphology of the joint. However, the lap joint of aluminum alloy that contains high amount of magnesium such as 5000 series aluminum alloy and SS400 has not been reported yet. In this study, we applied FSW to produce a lap joint of A5083 aluminum alloy and *Graduate Student, Niigata University SS400 steel using various process parameters such as rotational speed, traverse speed and pin depth. We examined the interface microstructure and correlated to the tensile shear load of the joint. 2. Experimental Procedure Plates of A5083 aluminum alloy and SS400 mild steel, each 3 mm thick, were used in this study. The chemical compositions of the materials are listed in Table 1. The plates were cut to 100 mm in length and 55 mm in width. The surfaces of both plates were polished with 240-grit emery paper and cleaned with acetone, and then were mounted in a jig to make a lap joint. The A5083 plate overlapped the SS400 plate by 30 mm as shown in Fig. 1. The rotating tool made of JIS-SKH57 tool steel had a shoulder (20 mm diameter) with an unthreaded pin (5 mm diameter and 3.0 3.3 mm long). The tilt angle of the rotating tool with respect to Z-axis of the milling machine was 1 degree as shown in Fig. 1. The rotating tool was plunged into the A5083 surface at a speed of 0.01 mm/s. The pin depth of a tool inserted into the surface of the SS400 was varied from þ0:1 mm to þ0:4 mm. Zero defines the position where the pin end is located at the interface between A5083 and SS400. The offset is defined in Fig. 2 (The pin depth of 0.0 mm was Table 1 Chemical composition of base materials (mass%). Element A5083 SS400 Mg 4.60 Si 0.08 0.02 Mn 0.62 0.42 Cu 0.02 Cr 0.06 Zn 0.01 Ti 0.02 C 0.15 P 0.21 S 0.06 Al Balance Fe 0.21 Balance

836 K. Kimapong and T. Watanabe Fig. 4 Specimen for tensile shear test. Fig. 1 Schematic of lap joint. Fig. 2 Zero offset of pin depth. Fig. 5 Relationship between rotational speeds, shear load and intermetallic compound (IMC) thickness. 3. Results and Discussion Fig. 3 Thermocouple location to measure interface temperature. ignored in this study because the preliminary experiments showed the lap joint produced under this pin depth was very weak. Most of the joint was failed during the specimen preparation process, therefore, the pin depth was selected between 0.1 to 0.4 mm for this investigation.). The holding time at the given depth was 30 s before traversing the tool along the X-axis. The temperatures during welding were measured by K-type thermocouples. The thermocouples were welded on the steel plate, as shown in Fig. 3. Temperatures were measured at 3 points along the welding line at the distances of 30, 50 and 70 mm from the end of the plate. A tensile shear test was conducted and the tensile shear specimens were prepared perpendicular to the X-axis from the welds. The welded area was located in the center of the tensile shear specimen. The dimension of the tensile shear specimen is shown in Fig. 4. Samples for metallographic examination were produced from the welds and mechanically polished. The polished samples were observed and analyzed using a scanning electron microscope (SEM) with X-ray energy-dispersive spectroscopy (EDS). 3.1 The effect of rotational speed on the shear load of a joint The lap joints were produced by rotational speeds of 3.75, 7.33, 14.67 and 20.67 s 1 and a traverse speed of 0.40 mm/s. The pin depth was þ0:1 mm. The relationship between rotational speed, shear load and intermetallic compound thickness is shown in Fig. 5. The shear load of the joint decreased as rotational speed increased. Under these welding conditions, 3.75 s 1 was the condition that gave maximum shear load of 4331 N. The cross sectional structures of the joints made under various rotational speeds are shown in Fig. 6. The rectangular area in the schematic was examined. The quantitative analysis by EDS revealed that the phases observed at the interfaces, indicated by FA and FA3, were identified as FeAl and FeAl 3 intermetallic compounds, respectively. The FeAl phase was formed at the interface with only 3.75 s 1. The FeAl 3 was formed on interface of 7.33, 14.67, and 20.67 s 1 conditions. The typical chemical compositions of those intermetallic compounds phases analyzed by EDS are listed in Table 2. The thickness of the intermetallic compound that was obtained by averaging over 20 points at the interfaces of retreating and advancing sides became thicker with increasing rotational speed. The reason of the increase in the intermetallic compound layer seemed to be because larger rotational speed gave rise to higher temperature at the

Lap Joint of A5083 Aluminum Alloy and SS400 Steel by Friction Stir Welding 837 Fig. 7 Relationship between traverse speeds, shear load and intermetallic compound (IMC) thickness. Fig. 6 Interfaces with traverse speed of 0.40 mm/s and various rotational speeds. (a) 3.75 s 1, (b) 7.33 s 1, (c) 14.67 s 1 and (d) 20.67 s 1. Table 2 Chemical composition of intermetallic compound (IMC) formed at the interface with various rotational speeds (mass%). Rotational speed, s 1 IMC Fe Al Mg O 3.75 FeAl 64.21 32.87 0.54 2.38 FeAl 3 38.63 56.90 1.47 3.00 7.33 FeAl 3 39.69 55.32 0.54 4.46 14.67 FeAl 66.39 30.48 3.12 0.00 FeAl 3 37.30 54.47 7.26 0.96 20.67 FeAl 75.62 20.70 1.38 2.29 FeAl 3 37.10 52.50 7.71 2.69 interface. The average peak temperatures at the interfaces of the joints made under the rotational speeds of 3.75 to 20.67 s 1 were 740, 765, 789 and 799 K, respectively. The increasing in the thickness of IMC with slightly increased average peak temperature of about 50 K is still unclear. Since the FeAl 3 phase is a hard and brittle intermetallic compound, 13,14) it appears that the thick intermetallic compound of FeAl 3 decreased the shear load of the joint. 3.2 The effect of traverse speed on the shear load of a joint The lap joint was made under the conditions of a rotational speed of 3.75 s 1 and the traverse speeds of 0.40, 0.73 and 1.43 mm/s. The pin depth was þ0:1 mm. The relationship between traverse speed, shear load and intermetallic compound thickness is shown in Fig. 7. The shear load increased with increasing the traverse speed from 0.40 to 1.43 mm/s, and it slightly decreased at the traverse speed of 1.43 mm/s. Under these welding conditions, 0.73 mm/s was the condition that gave the maximum shear load of 5591 N. The cross-sectional structures of the joints with 0.73 and 1.43 mm/s traverse speeds are shown in Fig. 8. Voids appreared at the interface of the joints with 0.40 mm/s as shown in Fig. 6(a) and 1.43 mm/s as shown in Fig. 8(b), however, the voids are not observed in the joint with 0.73 mm/s as shown in Fig. 8(a). EDS analysis identified the phases designated by FA and FA3 as FeAl and FeAl 3, respectively. The FeAl phase was formed along the interface of the joints with 0.40 and 0.73 mm/s. The FeAl 3 phase obsevered in Fig. 6(a) was formed only around the Fe particles scattered in the Al alloy matrix. However, these phases were not found at the interface with 0.73 and 1.43 mm/s. It may be attributed to the fact that the traverse speed of a rotating tool is so fast that the time and temperature are insufficient to produce these phases. The average peak temperatures at the interfaces with 0.40 to 1.43 mm/s were 740, 728 and 705 K, respectively, and the interface temperature with 1.43 mm/s was the lowest. The Fe-rich phase was formed only at the interface with 1.43 mm/ s. The typical chemical compositions of those intermetallic compounds phases analyzed by EDS are listed in Table 3. This phase was very thin and of about 0.20 mm in thickness. The thickness of the hard and brittle intermetallic compound of FeAl 3 also decreased with increasing traverse speed as shown in Fig. 7. When the tool traverse speed changed from 0.73 to 1.43 mm/s, the thickness of the intermetallic compound abruptly changed from 1.11 to 0.20 mm, however, the shear load of the joint just decreased slightly from 5591 to

838 K. Kimapong and T. Watanabe Fig. 9 Relationship between pin depth, shear load and intermetallic compound (IMC) thickness. Fig. 8 Interfaces perpendicular to welding direction of rotational speed of 3.75 s 1 at traverse speeds of 0.73 mm/s (a), 1.43 mm/s (b) and interface parallel to welding direction of 0.73 mm/s (c). Table 3 Chemical composition of intermetallic compound (IMC) formed at the interface with various traverse speeds (mass%). Traverse speed, mm/s IMC Fe Al Mg O 0.40 FeAl 64.21 32.87 0.54 2.38 FeAl 3 38.63 56.90 1.47 3.00 0.73 FeAl 62.72 34.23 2.26 0.78 FeAl 3 48.50 46.68 1.30 3.52 1.43 Fe-rich 89.28 6.25 0.69 3.78 5424 N. It seems beacause the defects such as a void is easier to occur at the interface, as seen in Fig. 8(b). Figure 8(c) indicates the cross section parallel to welding direction of the joint produced by 0.73 mm/s and it shows the fin-like shape on the interface of the joint. This fin-like shape was formed in all welding conditions and it will be explained later in Section 3.4. 3.3 The effect of the pin depth on the load of a joint In this section, the influence of the pin depth of a rotating tool was investigated. The lap joint was made under the conditions of a rotational speed of 3.75 s 1, and a traverse speed of 0.73 mm/s. The pin depth was varied from þ0:1 to þ0:4 mm. The relationship between the pin depth, shear load and the thickness of intermetallic compound is shown in Fig. 9. The shear load decreased with increasing of the pin depth. With these welding conditions, the maximum shear load could be obtained at þ0:1 mm pin depth with the shear load of 5591 N. Fig. 10 Interfaces with various pin depths. (a) 0.2 mm, (b) 0.3 mm and (c) 0.4 mm. Interface structures of the joints made with the pin depths of 0.2 to þ0:4 mm are shown in Fig. 10. The joints with the pin depth of þ0:2 to þ0:4 mm showed incomplete welded interfaces. A few voids were observed along the interfaces and they occurred readily at the region close to the retreating side, and increased in size with increasing the pin depth. The intermetallic compounds and the voids were observed along the interface of þ0:2 to þ0:3 as shown in Fig. 10. The EDS analysis identified the phases as FeAl 3 (indicated by FA3) and the analysis results are listed in Table 4. The thickness of the FeAl 3 intermetallic compound increased with increasing the pin depth because of the greater heat generation. The average peak temperatures of the welded interface were 740, 746, 755 and 779 K for the pin depths of

Lap Joint of A5083 Aluminum Alloy and SS400 Steel by Friction Stir Welding 839 Fig. 11 Fracture surfaces of shear test specimen. (a) aluminum side of 3.75 s 1, 0.73 mm/s, (b) steel side of 3.75 s 1, 0.73 mm/s, (c) aluminum side of 20.67 s 1, 0.40 mm/s and (d) steel side of 20.67 s 1, 0.40 mm/s. Table 4 Chemical composition of intermetallic compound (IMC) formed at the interface with various pin depths (mass%). Pin depth, mm IMC Fe Al Mg O 0.1 mm FeAl 64.21 32.87 0.54 2.38 FeAl 3 38.63 56.90 1.47 3.00 0.2 mm FeAl 3 38.62 54.62 2.91 3.85 0.3 mm FeAl 3 39.64 54.63 2.47 3.26 0.4 mm FeAl 3 40.64 54.13 2.24 2.99 0.1, 0.2, 0.3 and 0.4 mm, respectively. However, the relation between the IMC formation and slightly increased average peak temperature of about 39 K is still unclear in this study. 3.4 Fracture surface of tensile shear test specimen In the previous sections, the results suggested that the thickness variation of the intermetallic compound affected the shear load of the joints. In this section, the fracture surfaces of the shear specimens that were produced by the optimum condition (3.75 s 1, 0.73 mm/s and þ0:1 mm of pin depth) and the condition that gave the thickest intermetallic compound (20.67 s 1, 0.40 mm/s and þ0:1 mm of pin depth) were examined. SEM photographs of the fracture surfaces of the joints that were produced by 3.75 s 1, 0.73 mm/s and 20.67 s 1, 0.40 mm/s are shown in Fig. 11. Figures 11(a) and (b) are the fracture surfaces of the aluminum and steel side of the joint that was produced by 3.75 s 1, 0.73 mm/s condition. The aluminum and steel fracture surfaces show dimple pattern that indicates ductile feature of the joint. The chemical compositions of point I to V in Figs. 11(a) and (b) were analyzed by EDS to investigate the location where Table 5 Chemical composition of point I to X in Fig. 11 (mass%). Point IMC Fe Al Mg O I Al-rich 2.54 92.42 4.82 0.23 II Al-rich 1.16 93.95 4.89 0.00 III Al-rich 2.46 91.90 4.67 0.97 IV Al-rich 1.88 93.72 4.31 0.09 V Al-rich 1.98 91.56 5.27 1.20 VI FeAl 3 39.10 55.55 2.55 2.80 VII FeAl 3 48.28 52.16 2.05 2.52 VII FeAl 3 42.05 54.03 0.72 3.20 IX FeAl 3 38.84 57.33 1.19 2.64 X FeAl 3 43.79 52.38 0.73 3.10 the fracture occurred and the results are shown in Table 5 that all of the points contained high amount of Al. Comparing the results with the aluminum base materials in Table 1, it showed that the chemical compositions of these points were same as the base Al alloy used in this study. Therefore, it could be said that the ductile feature shown in Figs. 11(a) and (b) occurred in aluminum base material because if the fracture occurred at the interface of the joint, the FeAl phase should be observed, which was formed at the interface of this joint as reviewed in Section 3.2. Figures 11(c) and (d) are the fracture surfaces of the aluminum and steel sides of the joint that was produced by 20.67 s 1, 0.40 mm/s condition. The photographs show cleavage pattern on the fracture surfaces that indicates brittle feature of the joint. The chemical compositions of point VI to X in Figs. 11(c) and (d) were analyzed by EDS and the results in Table 5 show the phases contained the chemical compositions that were close FeAl 3 IMC in Fe Al phase diagram. 15) However, some measurement technique that can

840 K. Kimapong and T. Watanabe Fig. 12 Bonding process. (a) steel surface removal step at start point, (b) cross section parallel to welding direction shows fin-like shape, (c) bonding step parallel to welding direction and (d) bonding step at interface (Top view). confirm the certain IMC such as XRD or TEM is recommend for further examination. Therefore, it could be said that the fracture in Figs. 11(c) and (d) occurred in the interface of the joint where FeAl 3 existed because of the above reason. The observation of the fracture surfaces confirms that the formation and the thickness increasing of the brittleness IMC formed along the interface deteriorated the shear load of the joint. In order to increase the shear load of the joint, the IMC formed along the interface between aluminum and steel must be controlled to minimize the thickness of IMC. 3.5 Bonding process in the aluminum/steel lap joint produced by FSW The experimental results in above sections suggest the following process to bond aluminum to steel in the lap joint by friction stir welding. The first step: A rotating pin is plunged into the aluminum until the pin end reaches the interface between aluminum and steel. Frictional heat generated between the tool and materials leads the aluminum into fluid-like plastic state and the fluid-like aluminum is transferred around the rotating pin. The second step: The steel surface is rubbed and mechanically removed by the rotating pin as shown in Fig. 12(a). A part of steel is pushed by the rotating pin to the aluminum side and produced a fin-like shape inside the aluminum as shown in Fig. 12(b). The steel surface that touches the rotating pin becomes activated as shown in Figs. 12(a) and (c). The third step: When the rotating pin starts moving toward the traverse direction by a pitch of 1 revolution, aluminum in plastic flow is transferred around the pin in the rotating direction from the leading edge of the tool to the tailing edge and fills the space between the rotating pin and the fin-like shape surface, as schematically shown in Fig. 12(c). The direction of the aluminum that fills the space between an activated surface and the rotating pin is shown in Fig. 12(d). The activated steel surface attracts the aluminum that moves over it and produces a metallic bond between both metals. The fourth step: The rotating pin produces another fin-like shape of the steel and pushes it to the aluminum side as shown in Figs. 12(c) and (d). When a fin-like shape of the steel is produced, the second fin-like shape bonds to the aluminum immediately because the steel of the second finlike shape has an activated surface. As the rotating pin continuously moves toward the traverse direction, the aluminum transferred around the rotating pin bonds repeatedly to the activated steel, resulting in achieving the weld. 4. Conclusion We employed friction stir welding in order to make a lap joint between aluminum alloy and steel. In this study, the effects of rotational speed, traverse speed and pin depth of a tool on the shear load and the interface structure of the joint were investigated. The following results were obtained. (1) The lap joint between A5083 aluminum alloy and SS400 steel was successfully produced by FSW. (2) Increasing the rotational speed of a tool decreased the shear load of the joint because the higher rotational speed formed a thick FeAl 3 intermetallic compound at the interface between aluminum and steel. The fracture feature and location depended on the welding parameters. The fracture part of the joints at lower rotational speed was located in aluminum side and showed ductile fracture. The fracture part of the joints at higher rotational speed was located in intermetallic compound and showed brittle feature.

Lap Joint of A5083 Aluminum Alloy and SS400 Steel by Friction Stir Welding 841 (3) The formation of the FeAl 3 intermetallic compound around the interface conspicuously deteriorated the mechanical properties of the joint. (4) Increasing the traverse speed of a tool increased the shear load of joints. However, extremely high traverse speeds such as of 1.43 mm/s produced the joint with incompletely welded interface. (5) Increasing the pin depth made the intermetallic compound thickness thicker and gave rise to an incomplete joint. Acknowledgments The authors would like to thank Mr. Atsushi Yanagisawa and Mr. Kengo Oohara, with the Department of Mechanical Engineering, Niigata University for their technical support. REFERENCES 1) T. A. Branes and I. R. Pashby: J. Mater. Proc. Technol. 99 (2000) 62 71. 2) W. M. Thomas and E. D. Nicholas: Mater. Des. 18 (1997) 269 273. 3) G. Liu, L. E. Murr, C.-S. Niou, J. C. McClure and F. R. Vega: Scr. Mater. 37 (1997) 355 361. 4) C. G. Rhodes, M. W. Mahoney, W. H. Bingel, R. A. Spurling and C. C. Bampton: Scr. Mater. 36 (1997) 69 75. 5) S. Benavides, Y. Li, L. E. Murr, D. Brown and J. C. McClure: Scr. Mater. 41 (1999) 809 815. 6) K. V. Jata and S. L. Semiatin: Scr. Mater. 43 (2000) 743 749. 7) Y. Li, L. E. Murr and J. C. McClure: Mater. Sci. Eng. A 271 (1999) 213 223. 8) A. P. Reynolds, W. Tang, T. Gnaupel-Herold and H. Prask: Scr. Mater. 48 (2003) 1289 1294. 9) T. Watanabe, A. Yanagisawa and H. Takayama: Q. J. Japan Weld. Society 22 (2004) 466 467. 10) C. M. Chen and R. Kovacevic: Mach. Tool Manuf. 44 (2004) 1205 1214. 11) K. Yoshikawa and T. Hirano: Proc. Third Int. Symp. Friction Stir Welding, Kobe, Japan, September 27 28, 2001. 12) A. Elrefaey, A. Takahashi and K. Ikeuchi: Preprints of the national meeting of JWS. 73 (2003) 66 67. 13) M. Yilmaz, M. Col and M. Acet: Mater. Charact. 49 (2003) 421 429. 14) S. Kobayashi and T. Yakou: Mater. Sci. Eng. A 338 (2002) 44 53. 15) ASM international: Metal Handbook Vol. 9 Binarry Phase Diagram, (ASM international, Materials Park, OH, 1990) p. 147.